BR122020017765B1 - METHODS AND APPARATUS FOR DETECTING AN ENTITY IN A BODY SAMPLE - Google Patents

Abstract

methods and apparatus for detecting an entity in a body sample. This is an apparatus and methods comprising a microscope system (11) configured to acquire one or more microscope images of a body sample, an output device (34) and at least one computer processor (28). the computer processor identifies, in the one or more images, at least one element as being a pathogenic candidate and extracts, from the one or more images, at least one candidate informative feature associated with the pathogenic candidate. the computer processor extracts, from the one or more images, at least one sample informative feature that is indicative of contextual information related to the body sample. the computer processor classifies a probability of the body sample being infected with a pathogenic infection by processing the candidate information resource in combination with the sample information resource and generating an output on the output device in response thereto. other applications are also described.

Description

Split from the application for Invention Patent BR 11 2018005099-7, filed on 09/15/2016. Cross Reference to Related Requests

[001] The present application claims priority to U.S. Provisional Patent Application 62/219,889 by Eshel, filed on September 17, 2015, entitled “Methods of detecting a pathogen in a bodily sample and system thereof. ”

[002] The request referred to above is incorporated into this document by way of reference.

Field of Embodiments of the Invention

[003] Some applications in the currently disclosed matter generally relate to the detection of entities in a body sample, and in particular, to the detection of pathogens automatically using classification and image processing.

Fundamentals

[004] The primary method of detecting certain pathogenic infections within a body sample (e.g., a blood sample) is microscopic examination of the body sample, and visual confirmation of the presence and concentration of the pathogen. Staining a body sample with a dye or dye before microscopic examination is often used to enhance contrast in the microscopic image and to visually highlight cells that have a particular biological composition. In particular, some fluorescent dyes have an affinity for nucleic acid in cells. When excited by fluorescent light at an appropriate wavelength, the nucleic acid will fluoresce. Consequently, fluorescent dyes are sometimes used to differently stain parts of a cell for detection under a microscope. For example, when excited by blue light, the fluorochrome Acridine Orange bound to DNA will emit green light, and when bound to RNA it will emit red light. Bloodborne pathogens such as Anaplasma marginale, Hemobartonella, trypanosomes, Plasmodium spp., Babesia spp. and others were all detected with Acridine Orange.

[005] Although the primary method of pathogen detection maintains visual identification in a microscopic bright field image, fluorescent microscopy has also been used, albeit to a lesser extent. However, in both cases, detection of a pathogenic infection through manual pathogen identification suffers from two main disadvantages: many settings (especially rural) are not equipped to perform the test, and the accuracy of the results depends both on the skill of the person examining the sample for levels of the pathogen in the sample. Consequently, attempts have been made to automate the detection of pathogens in a body sample.

Summary of Modalities

[006] In accordance with some applications of the present invention, one or more microscope images of a body sample (e.g., a blood sample) are acquired, using a microscope of a microscope system. A computer processor identifies at least one element as a pathogenic candidate (i.e., a constituent element within the sample that exhibits characteristics that indicate that it may be a pathogen and therefore is a candidate for being a pathogen) within the images . For example, the images may be images of a blood sample that were acquired while the sample was stained with a dye or dye that is configured to stain DNA and/or RNA within the sample, and the computer processor may identify the candidate by detecting it. if dyed elements (e.g. fluorescent elements) within the images. The computer processor extracts, from the one or more images, at least one candidate informative feature associated with the pathogenic candidate and at least one sample informative feature that is indicative of contextual information related to the body sample. The probability of the body sample being infected with a pathogenic infection is classified by processing the candidate information feature in combination with the sample information feature. An output is typically generated at an output device in response to classification.

[007] For some applications, in response to the candidate information resource, the computer processor performs a first classification, in which a probability of the pathogenic candidate being a pathogen is classified. In response to the first classification, in combination with the sample information feature, the computer processor performs a second classification in which a probability of the body sample containing a pathogenic infection is classified. For some applications, the first classification (in which a probability of the pathogen candidate being a pathogen is classified) is performed in response to the candidate information feature in combination with the sample information feature. For some applications, the computer processor classifies a pathogenic infection in the body sample as a given type of pathogenic infection (e.g., Plasmodium, a given strain of Plasmodium, and/or Plasmodium of a given age or age range), processing the candidate information resource in combination with the sample information resource.

[008] For some applications, the candidate information resource comprises a size of the pathogenic candidate (e.g., dimension, length, circumference, minimum width, maximum width, area and/or relative size of the candidate with respect to other candidates or entities) , a shape of the pathogenic candidate, a movement of the pathogenic candidate, an intensity of the pathogenic candidate, a location of the pathogenic candidate within the body sample (comprising proximity, support and/or overlap of the candidate with respect to other candidates or entities), a property of a cell that overlies the pathogenic candidate, a color of the pathogenic candidate (comprising staining intensity and pattern), a texture (e.g., contour) of the pathogenic candidate, and/or an acuity of a boundary of the pathogenic candidate. Additional examples without limitation of candidate informational features are described, for example, in U.S. Patent Document No. 2012/0169863 to Bachelet and/or U.S. Patent No. 2015/0037806 to Pollak, both of which applications are incorporated herein by way of reference.

[009] For some applications, sample informative features comprise a size of one or more non-pathogenic candidate constituents in the body sample, a shape of one or more non-pathogenic candidate constituents within the body sample, an intensity of one or more constituents of non-pathogenic candidate within the body sample, a quantity of cells of a given cell type within the body sample, a distribution of cells of a given cell type within the body sample and/or a distribution of pathogenic candidates within the body sample.

[0010] Therefore, in accordance with some applications of the present invention, an apparatus comprising:

[0011] a microscope system configured to acquire one or more microscope images of a body sample;

[0012] an output device; It is

[0013] at least one computer processor configured to:

[0014] identify, in the one or more images, at least one element as being a pathogenic candidate,

[0015] extract, from the one or more images, at least one candidate informative feature associated with the pathogenic candidate,

[0016] extract, from the one or more images, at least one informative sample feature that is indicative of contextual information related to the body sample; classify a probability of the body sample being infected with a pathogenic infection by processing the candidate informative feature in combination with the sample informative feature, and

[0017] generate an output on the output device, in response thereto.

[0018] In some applications:

[0019] the microscope system is configured to acquire one or more microscope images of a body sample that is dyed with a dye; It is

[0020] the at least one computer processor is configured to identify the at least one element as being a pathogenic candidate by identifying the at least one element as being a pathogenic candidate by identifying that the at least one element is dyed.

[0021] In some applications, the at least one computer processor is configured to process the candidate information resource in combination with the sample information resource via:

[0022] in response to the candidate information resource, performing a first classification, in which a probability of the pathogenic candidate being a pathogen is classified and

[0023] in response to the first classification in combination with the sample information feature, a second classification is performed in which a probability of the body sample containing a pathogenic infection is classified.

[0024] In some applications, the at least one computer processor is configured to process the candidate information resource in combination with the sample information resource via:

[0025] in response to the candidate information resource in combination with the sample information resource, performing a first classification, in which a probability of the pathogenic candidate being a pathogen is classified; It is

[0026] at least partially in response to the first classification, performing a second classification in which a probability of the body sample containing a pathogenic infection is classified.

[0027] In some applications, the at least one computer processor is configured to extract, from the one or more images, at least one candidate informative feature associated with the pathogenic candidate by extracting, from the one or more images, at least one candidate informative feature associated with the pathogenic candidate, wherein the candidate informative feature is a feature selected from the group consisting of: a size of the pathogenic candidate, a shape of the pathogenic candidate, a movement of the pathogenic candidate, a intensity of the pathogenic candidate, a location of the pathogenic candidate within the body sample, a property of a cell that overlaps the pathogenic candidate, a color of the pathogenic candidate, a texture of the pathogenic candidate, and an acuity of a threshold of the pathogenic candidate.

[0028] In some applications, the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from of the one or more images, at least one sample informative feature selected from the group consisting of: a size of one or more non-pathogenic candidate constituents in the body sample, a shape of one or more non-pathogenic candidate constituents within the body sample, an intensity of one or more non-pathogenic candidate constituents within the body sample, a quantity of cells of a given cell type within the body sample, a distribution of cells of a given cell type within the body sample, and a distribution of pathogenic candidates within the body sample.

[0029] In some applications:

[0030] the microscope system is configured to acquire one or more microscope images of the body sample by acquiring one or more microscope images of a body sample that is dyed with a dye; It is

[0031] the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from the one or more images, at least one sample informative feature that is indicative of a dyeing quality of the body sample through dyeing.

[0032] In some applications, the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from gives one or more images, at least one informative sample feature that is indicative of a foreign object that is present in the body sample.

[0033] In some applications, the body sample comprises a body sample selected from the group consisting of: a blood sample, a diluted blood sample, a sample comprising predominantly red blood cells, and a diluted sample comprising predominantly red blood cells, and the system microscope is configured to acquire one or more images of the selected body sample.

[0034] In some applications, the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from gives one or more images, a size of one or more red blood cells that are present within the body sample.

[0035] In some applications, the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from gives one or more images an indication of a presence of Howell Jolly bodies within the body sample.

[0036] In some applications, the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from gives one or more images, a concentration of platelets within the body sample.

[0037] In some applications, the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from gives one or more images, a relationship between several reticulocytes associated with the candidates and several mature red blood cells associated with the candidates.

[0038] In some applications, the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from gives one or more images, a concentration of reticulocyte bodies within the body sample.

[0039] In some applications, the at least one computer processor is configured to classify the probability of the body sample being infected with the pathogenic infection by setting a threshold for a positive determination of a pathogenic infection, based on the concentration of the bodies. of reticulocytes within the body sample.

[0040] In some applications, the at least one computer processor is configured to classify a pathogenic infection in the body sample as containing one or more given pathogen types by processing the candidate information resource in combination with the sample information resource. .

[0041] In some applications, the at least one computer processor is configured to classify the pathogenic infection in the body sample as containing one or more selected pathogen types. from the group consisting of: Plasmodium, a given strain of Plasmodium, Plasmodium of a given age, and Plasmodium of a given age range.

[0042] In some applications:

[0043] the body sample comprises a body sample selected from the group consisting of: a blood sample, a diluted blood sample, a sample comprising predominantly red blood cells and a diluted sample comprising predominantly red blood cells;

[0044] the at least one computer processor is configured to extract, from the one or more images, at least one informative sample feature that is indicative of contextual information related to the body sample by extracting, from the one or more images, a relationship between several reticulocytes associated with the candidates and several mature red blood cells associated with the candidates; It is

[0045] the at least one computer processor is configured to classify the pathogenic infection in the body sample as containing one or more given types of pathogen by classifying the pathogenic infection in the body sample as containing the given type of pathogen, at least partially with based on the relationship between several reticulocytes associated with the candidates and several mature red blood cells associated with the candidates.

[0046] In some applications:

[0047] the body sample comprises a body sample selected from the group consisting of: a blood sample, a diluted blood sample, a sample comprising predominantly red blood cells and a diluted sample comprising predominantly red blood cells;

[0048] the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from the one or more images, red blood cell formats within the body sample and

[0049] the at least one computer processor is configured to classify the pathogenic infection in the body sample as containing the given type of pathogen, classifying the pathogenic infection in the body sample as the given type of pathogenic infection, at least partially based on the formats of red blood cells within the body sample.

[0050] In some applications:

[0051] the body sample comprises a body sample selected from the group consisting of: a blood sample, a diluted blood sample, a sample comprising predominantly red blood cells and a diluted sample comprising predominantly red blood cells;

[0052] the at least one computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample by extracting, from the one or more images, sizes of red blood cells within the body sample and

[0053] the at least one computer processor is configured to classify the pathogenic infection in the body sample as containing the given type of pathogen, classifying the pathogenic infection in the body sample as the given type of pathogenic infection, at least partially based on the red blood cell sizes within the body sample.

[0054] Additionally, in accordance with some applications of the present invention, a method is provided that comprises:

[0055] acquiring one or more microscope images of a body sample using a microscope;

[0056] with the use of at least one computer processor:

[0057] in the one or more images, identify at least one element as being a pathogenic candidate;

[0058] extract, from the one or more images, at least one candidate informative feature associated with the pathogenic candidate;

[0059] extract, from the one or more images, at least one informative sample feature that is indicative of contextual information related to the body sample; classifying a probability of the body sample being infected with a pathogenic infection by processing the candidate informative feature in combination with the sample informative feature; It is

[0060] generate an output in response to it.

[0061] There is further provided, in accordance with some applications of the present invention, a computer software product, for use with a body sample, an output device, and a microscope system configured to acquire one or more microscope images of a bodily specimen, wherein the computer software product comprises a non-transitory computer-readable medium on which program instructions are stored, wherein the instructions, when read by a computer, cause the computer to perform the steps of: one or more images, identify at least one element as being a pathogenic candidate; extracting, from the one or more images, at least one candidate informative feature associated with the pathogenic candidate; extracting, from the one or more images, at least one informative sample feature that is indicative of contextual information related to the body sample; classifying a probability of the body sample being infected with a pathogenic infection by processing the candidate information feature in combination with the sample information feature; and generating an output on the output device in response thereto.

[0062] Additionally, in accordance with some applications of the present invention, an apparatus comprising:

[0063] a microscope system configured to acquire one or more microscope images of a body sample;

[0064] an output device; It is

[0065] at least one computer processor configured to:

[0066] extract, from the one or more images, at least one sample informational feature that is indicative of contextual information related to the body sample, at least partially based on the extracted sample informational feature: identify that there is a defect associated with the body sample arranged in the sample carrier, and

[0067] classify a source of the defect and

[0068] in response thereto, generate an output in the output device that is indicative of the source of the defect.

[0069] In some applications the at least one computer processor is configured to classify the source of the defect by classifying the source as being at least one source selected from the group consisting of: the sample carrier, a given portion of the sample carrier, the body sample, and a diluent in which the sample was diluted.

[0070] Additionally, in accordance with some applications of the present invention, a method is provided that comprises:

[0071] acquiring one or more microscope images of a body sample arranged in a sample carrier using a microscope;

[0072] with the use of at least one computer processor:

[0073] extract, from the one or more images, at least one sample informational feature that is indicative of contextual information related to the body sample, at least partially based on the extracted sample informational feature: identify that there is a defect associated with the body sample arranged in the sample carrier, and

[0074] classify a source of the defect; It is

[0075] in response thereto, generate an output that is indicative of the source of the defect.

[0076] There is further provided, in accordance with some applications of the present invention, a computer software product for use with a body sample, an output device, and a microscope system configured to acquire one or more microscope images of a body specimen, wherein the computer software product comprises a non-transitory computer-readable medium on which program instructions are stored, wherein the instructions, when read by a computer, cause the computer to perform the steps of: extracting , from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample; at least partially based on the extracted sample information resource: identifying that there is a defect associated with the body sample disposed in the sample carrier and classifying a source of the defect; and, in response thereto, generating an output in the output device that is indicative of the source of the defect.

[0077] Additionally, in accordance with some applications of the present invention, an apparatus for classifying a body sample is provided, the apparatus comprising:

[0078] a microscope system configured to acquire one or more microscope images of the body sample;

[0079] an output device; It is

[0080] at least one computer processor configured to:

[0081] identify, in the one or more images, at least one element as being a candidate of a given entity,

[0082] extract, from the one or more images, at least one candidate information resource associated with the identified element,

[0083] extract, from the one or more images, at least one sample informational feature that is indicative of contextual information related to the body sample, process the candidate informational feature in combination with the sample informational feature, and

[0084] generate an output on the output device, in response thereto.

[0085] In some applications, the body sample comprises a sample that contains blood and the computer processor is configured to identify at least one element as being a candidate of a given entity. given entity within the blood.

[0086] In some applications, the computer processor is configured to identify at least one element as being a candidate of a given entity by identifying at least one element as being a pathogenic candidate.

[0087] Additionally, in accordance with some applications of the present invention, a method for classifying a body sample is provided, the method comprising:

[0088] acquiring one or more microscope images of the body sample using a microscope;

[0089] with the use of at least one computer processor:

[0090] in the one or more images, identify at least one element as being a candidate for a given entity;

[0091] extract, from the one or more images, at least one candidate information resource associated with the identified element;

[0092] extracting, from the one or more images, at least one sample informational feature that is indicative of contextual information related to the body sample, processing the candidate informational feature in combination with the sample informational feature; It is

[0093] generate an output in response to it.

[0094] In some applications, the body sample comprises a sample that contains blood and identifying at least one element as being a candidate of a given entity comprises identifying at least one element as being a candidate of a given entity within the blood.

[0095] In some applications, identifying at least one element as being a candidate of a given entity comprises identifying at least one element as being a pathogenic candidate.

[0096] There is further provided, in accordance with some applications of the present invention, a computer software product, for use with a body sample, an output device, and a microscope system configured to acquire one or more microscope images of a body sample, wherein the computer software product comprises non-transitory computer-readable media,

[0097] in which program instructions are stored, these instructions, when read by a computer, cause the computer to perform the steps of: in the one or more images, identifying at least one element as being a candidate of a given entity; extract, from the one or more images, at least one candidate information resource associated with the identified element; extracting, from the one or more images, at least one informative sample feature that is indicative of contextual information related to the body sample; processing the candidate information resource in combination with the sample information resource; and generating an output at the output device in response thereto.

[0098] Additionally, in accordance with some applications of the present invention, an apparatus comprising:

[0099] a microscope system configured to acquire one or more microscope images of a body sample;

[00100] an output device; It is

[00101] at least one computer processor configured to:

[00102] in the one or more images, identify at least one element as being a candidate of a given entity,

[00103] extract, from the one or more images, at least one candidate information resource associated with the candidate,

[00104] extract, from the one or more images, at least one sample informational feature that is indicative of contextual information related to the body sample, process the candidate informational feature in combination with the sample informational feature, and

[00105] in response thereto, perform an action selected from the group consisting of: generating an output on the output device that indicates that the presence of an infection within the body sample could not be determined with a sufficient degree of reliability, generate an output at the output device that indicates that an image of a portion of the sample should be reimaged, generate an output at the output device that indicates that an image of a portion of the sample should be reimaged using different settings, take the microscope system to reimage a portion of the sample, cause the microscope system to reimage a portion of the sample using different settings, and modulate a frame rate at which the microscope images are acquired by the microscope system.

[00106] Additionally, in accordance with some applications of the present invention, a method is provided that comprises:

[00107] acquiring one or more microscope images of a body sample using a microscope;

[00108] with the use of at least one computer processor:

[00109] in the one or more images, identify at least one element as being a candidate for a given entity;

[00110] extract, from the one or more images, at least one candidate information resource associated with the candidate;

[00111] extract, from the one or more images, at least one informative sample feature that is indicative of contextual information related to the body sample; processing the candidate information resource in combination with the sample information resource; It is

[00112] in response thereto, perform an action selected from the group consisting of: generating an output that indicates that the presence of an infection within the body sample could not be determined with a sufficient degree of reliability, generating an output that indicates that a portion of the sample should be reimaged, generate an output that indicates that a portion of the sample should be reimaged using different settings, cause the microscope to reimage a portion of the sample, cause the microscope to reimage a portion of the sample with the use of different settings and modulating a frame rate at which microscope images are acquired by the microscope.

[00113] There is further provided, in accordance with some applications of the present invention, a computer software product for use with a body sample, an output device, and a microscope system configured to acquire one or more microscope images of a bodily specimen, wherein the computer software product comprises a non-transitory computer-readable medium on which program instructions are stored, such instructions, when read by a computer cause the computer to perform the steps of: in one or more images, identify at least one element as being a candidate for a given entity; extract, from the one or more images, at least one candidate information resource associated with the candidate; extracting, from the one or more images, at least one informative sample feature that is indicative of contextual information related to the body sample; processing the candidate information resource in combination with the sample information resource; and in response thereto, perform an action selected from the group consisting of: generating an output on the output device that indicates that the presence of an infection within the body sample could not be determined with a sufficient degree of reliability, generating a output on the output device that indicates that a portion of the sample should be reimaged, generate an output on the output device that indicates that a portion of the sample should be reimaged using different settings, cause the microscope system to reimage a portion of the sample, cause the microscope system to reimage a portion of the sample using different settings, and modulate a frame rate at which microscope images are acquired by the microscope system.

[00114] Additionally, in accordance with some applications of the present invention, an apparatus comprising:

[00115] a microscope system configured to acquire one or more microscope images of a body sample;

[00116] an output device; It is

[00117] at least one computer processor configured to:

[00118] identify, within one or more images of the image set, elements as being candidates for one or more given entities,

[00119] extract, from the one or more images, candidate information resources associated with the candidates,

[00120] extract, from the candidate information resources, two or more sample information resources related to the body sample,

[00121] determine a characteristic of the body sample by processing the two or more sample information resources and

[00122] generate an output in response to it.

[00123] In some applications, the body sample comprises a body sample that contains blood and the computer processor is configured to extract candidate informational features associated with the candidates by extracting one or more candidate informational features associated with a pathogenic candidate within from the blood and extracting one or more candidate information resources associated with platelets within the blood.

[00124] In some applications, the body sample comprises a body sample that contains blood and the computer processor is configured to extract candidate informational features associated with the candidates by extracting one or more candidate informational features associated with a pathogenic candidate within from the blood and extracting one or more candidate informational features associated with reticulocytes within the blood.

[00125] In some applications:

[00126] the body sample comprises a body sample that contains blood,

[00127] the computer processor is configured to identify, within one or more images of the image set, elements as being candidates of one or more given entities, identifying elements as being pathogenic candidates and

[00128] the computer processor is configured to extract, from the candidate information resources, two or more sample information resources related to the body sample by extracting, from the candidate information resources, two or more sample information resources selected from the group consisting of: number of pathogenic candidates in the sample, type of pathogenic candidates in the sample, brightness of the candidates relative to the background brightness, a probability that the candidates are pathogens, number of candidates that have a probability of being a pathogen that exceeds a threshold, number of candidates that have a probability of being a given type of pathogen that exceeds a threshold, a number of platelets in the sample, brightness of platelets, several reticulocytes in the sample, several reticulocytes infected by pathogens in the sample, a proximity of candidate RBCs and several RBCs in the sample.

[00129] Additionally, in accordance with some applications of the present invention, a method for classifying a body sample is provided, the method comprising:

[00130] acquiring a set of microscope images of the body sample, using a microscope; with the use of at least one computer processor:

[00131] identify, within one or more images of the image set, elements as being candidates for one or more given entities;

[00132] extract, from the one or more images, candidate information resources associated with the candidates;

[00133] extracting, from the candidate information resources, two or more sample information resources related to the body sample;

[00134] determining a characteristic of the body sample by processing the two or more sample information resources; It is

[00135] generate an output in response to it.

[00136] There is further provided, in accordance with some applications of the present invention, a computer software product for use with a body sample, an output device, and a microscope system configured to acquire one or more microscope images of a bodily specimen, wherein the computer software product comprises a non-transitory computer-readable medium on which program instructions are stored, such instructions, when read by a computer cause the computer to perform the steps of: identifying within a or more images from the set of image elements as being candidates for one or more given entities; extract, from the one or more images, candidate information resources associated with the candidates; extracting, from the candidate information resources, two or more sample information resources related to the body sample; determining a characteristic of the body sample by processing the two or more sample information resources; and generating an output at the output device in response thereto.

[00137] The present invention will be understood more completely from the following detailed description of the embodiments thereof, obtained together with the drawings, in which: Brief description of the Drawings

[00138] Figure 1 is a generalized functional diagram of a pathogen detection system, in accordance with some applications of the present invention;

[00139] Figure 2 is a generalized flowchart of steps that are performed, in accordance with some applications of the present invention;

[00140] Figure 3 is a non-limiting example of imaging information that is analyzed, in accordance with some applications of the present invention; It is

[00141] Figure 4 is an illustration without limitation of a relative location of an RNA dyed region and a DNA dyed region, in accordance with some applications of the present invention. Detailed Description of Modalities

[00142] Reference will now be made to Figure 1, which is a functional diagram of a pathogen detection system 10, in accordance with some applications of the present invention. Pathogen detection system 10 comprises a processor 28 operatively coupled to a memory 30, for example, via a communication bus 31. In certain embodiments, the pathogen detection system 100 may optionally comprise or be operatively coupled to a microscope 11. Microscope system 11 is typically a digital microscope comprising an imaging module 14, a focus shifting module 16, a sample carrier 18, and an autofocus system 20. For some applications, the microscope system 11 is generally similar to the microscope system described in U.S. 2014/0347459 by Greenfield, which is incorporated herein by reference.

[00143] Typically, the imaging module 14 comprises an optical unit 22 and an image sensor unit 24. Optical unit 22 is configured to form a magnified image of a body sample 12 (e.g., a blood sample) by combining a focus plane 36 and an image plane. The image sensor unit 24 typically comprises an image sensor, e.g., a charge coupled device (CCD), complementary metal oxide semiconductor sensor (CMOS), and/or an array sensor, positioned in the plane of image from the optical drive 22 in order to detect the magnified image.

[00144] Computer processor 28 typically receives and processes images. The computer processor communicates with memory 30 and images are received by the processor through the memory. Through a user interface 32, a user (e.g., a laboratory technician) sends instructions to the computer processor. For some applications, the user interface comprises a keyboard, a mouse, a controller, a touch-screen device (such as a smart phone or tablet computer), a touchpad, a trackball, a command interface voice and/or other types of user interfaces that are known in the art. Typically, the computer processor generates an output through an output device 34. Furthermore, typically, the output device comprises a display, such as a monitor, and the output comprises an output that is displayed on the display. For some applications, the processor generates output to a different type of visual, text, graphic, tactile, audio, and/or video output device, for example, speakers, headphones, a smart phone, or a computer tablet type. For some applications, user interface 32 acts as both an input interface and an output interface, that is, it acts as an input/output interface. For some applications, the processor generates output on a computer-readable medium (e.g., a non-transitory computer-readable medium), such as a disk or portable USB drive, and/or generates output on a printer.

[00145] Microscope system 11 may, in certain embodiments, comprise a local processor that controls at least some of the processes of the microscope system 11, e.g., image acquisition and/or communication with other components, including other components of the microscope system 11. pathogen detection 10 and components external to the pathogen detection system 10. In certain other embodiments, the processor 28 may control one or more processes of the microscope system 11, including, for example, image acquisition and/or communication. Optionally, the pathogen detection system 10 may comprise or be operatively coupled to a plurality of digital microscopes. Optionally, each respective digital microscope in the plurality of digital microscopes has its own local processor.

[00146] In certain embodiments, memory 30 may be configured to store imaging information, program data, and/or executable program instructions for detecting a pathogen in a body sample, as will be detailed below with reference to Figure 2. The memory 30 may be, for example, volatile memory or non-volatile memory. In certain embodiments, memory 30 is non-volatile memory, e.g., hard disk drive, flash memory, etc.

[00147] For some applications, the microscope system 11 is configured to capture one or more high-magnification digital images of a body sample. Optionally, the one or more digital images comprise images covering different portions of the body sample. Optionally, the images do not overlap (or overlap by less than 5 percent or less than 1 percent). Optionally, the images comprise images that overlap and are obtained at different depths of focus and/or with different lighting conditions. Optionally, the one or more digital images comprise sets of images that do not overlap (or overlap by less than 5 percent or less than 1 percent), however, each of the sets comprises images from another set, obtained under different conditions of lighting. In certain embodiments, the microscope system 11 is configured to capture images under a plurality of lighting conditions, including, for example, field light, blue light, and ultraviolet light, as will be further detailed below.

[00148] In accordance with some applications, the body sample 12 (e.g., a blood sample) is digitized by the microscope system so that a plurality of portions of the body sample are transformed into an image. For some applications, a plurality of images are acquired of one or more portions of the body sample, each of the plurality of images being acquired under respective imaging conditions. For example, two images of a portion of the body sample can be acquired using, respectively, imaging conditions that allow detection of cells (e.g., bright field) and imaging conditions that allow visualization of dyed bodies (e.g., example, appropriate fluorescent lighting).

[00149] Image sensor unit 24 may output acquired digital images to the output device 34 (which may comprise a viewfinder) and/or to the autofocus system 20. Focus ranging module 16 may be configured to vary a distance between the focus plane 36 of the optical unit 22 and the sample carrier 18. Focus variation module 16 can be operated manually or automatically by means of a mechanical interface that can, for example, modify the position of the sample carrier 18 to the along an optical axis Z of the optical unit 22. Alternatively or additionally, the focus shifting module 16 can be controlled by the autofocus system 20. For example, the focus shifting module 16 can vary the distance between the sample 18 and the focus plane (1) by modifying the position of optical unit 22 along the optical geometric axis Z, (2) by modifying the position of the sample carrier 18 along the position of the optical geometric axis Z ( for example, by moving a stage on which the sample carrier is placed), (3) modifying the position of the focus plane, for example, by changing a focal length of the optical unit 22 or a combination thereof .

[00150] The sample carrier 18 may comprise a plate. The sample carrier 18 may be configured to accommodate the body sample 12. The carrier may be any carrier known in the art to hold a biological sample. Optionally, the lower surface of the carrier is essentially flat, to allow cells in contact with it to be about the same distance from the focal plane of the microscope. Examples include carrier slides, laboratory vessels, dishes, plates, multiwell plates, test tubes (e.g., with a flat bottom), microfluidic cells, cartridges, and the like.

[00151] Autofocus system 20 may comprise an autofocus computing module 38 and an autofocus adaptation module 39. The autofocus computing module may be connected to the image sensor unit 24 so as to receive images acquired by the image sensor module. imaging 14. The autofocus adaptation module may be connected to the focus range module 16 and may be configured to command the focus range module 16, for example, as described above.

[00152] For some applications, the processor 28 comprises one or more functional modules, such as a feature extraction module, a candidate classifier, a sample classifier, and a diagnostics module. For some applications, processor 28 is configured to process imaging information by extracting features contained in the imaging information. Typically, the processor is configured to extract at least one sample informative feature and at least one candidate informative feature. For some applications, the processor is further configured to process the at least one sample informational resource to obtain contextual information, and to process the at least one candidate informational resource to obtain candidate data, as will be further detailed below .

[00153] Typically, the processor is configured to classify a probability of a candidate (i.e., a constituent element within the sample that exhibits characteristics that indicate that it may be a pathogen and therefore is a candidate for being a pathogen). that is a pathogen, at least partially, based on at least one candidate information feature. Furthermore, typically, the processor is configured to classify a probability of the body sample being infected with a pathogenic infection by processing the at least one candidate informative feature in combination with the at least one sample informative feature.

[00154] For some applications, the processor is programmed to classify the probability of a candidate being a pathogen and/or classify a probability of the sample being infected with a pathogenic infection, using classification and/or a machine learning algorithm, e.g. support vector machines, neural networks, naive Bayesian algorithms, etc. Additional examples of types of classification and/or machine learning algorithms that may be used by the processor are described in U.S. 2012/0169863 by Bachelet and/or U.S. 2015/0037806 by Pollak, both of which applications are incorporated herein. for reference. For some applications, the computer processor is trained, before being used to analyze a body sample, using training images of body samples.

[00155] For some applications, if a body sample is determined to be infected with a pathogenic infection (or if it is determined that the probability of the body sample being infected with a pathogenic infection exceeds a threshold), the computer processor is further configured to extract diagnostic information about the pathogenic infection in accordance with at least one sample information resource.

[00156] It is noted that the teachings of the currently disclosed subject matter are not linked to the specific pathogen detection system described with reference to Figure 1. Equivalent and/or modified functionality may be consolidated or otherwise divided and may be deployed in any appropriate combination of software, firmware and hardware. The processor can be implanted with a suitably programmed computer.

[00157] Reference will now be made to Figure 2, which shows a generalized flowchart of a method for detecting a pathogenic infection in a bodily sample (e.g., a blood sample), in accordance with some applications of the present invention.

[00158] In a first step 200, one or more images of the body sample are acquired by the microscope system 11. To the one or more images, informative data from one or more images or data derived from one or more images (collectively referred to as herein “imaging information”) is typically stored in memory 30. The imaging information is then analyzed by processor 28, as described in more detail below herein. It is noted that in the present application, the computer processor is described as extracting resources from the one or more images. This terminology should be interpreted as including extracting the features from informative data from the one or more images, or data derived from the one or more images, and should not be interpreted as being limited to directly extracting the features from the one or more images. images properly.

[00159] For some applications, imaging information is informative of at least one high-magnification microscopic view of the sample. Alternatively or additionally, the imaging information is informative of a plurality of images, including, for example, images of different portions of the sample, images of the same portion of the sample obtained at different focal depths and/or different lighting conditions, and/or at different times.

[00160] The body sample can be from any living creature, however, preferably from warm-blooded animals. Typically, the body sample is a blood sample. The sample may be any blood sample or a portion thereof comprising one or more red blood cells. Optionally, the sample predominantly comprises red blood cells (i.e., a majority of the cells (e.g., at least 60 percent of the cells) in the sample are red blood cells). Optionally, the sample also comprises at least one of platelets and leukocytes. Optionally, the blood sample is diluted. Optionally, dilution is performed or the sample is otherwise prepared so that the concentration of cells on the surface that is imaged is between 3,000 and 30,000 cells (e.g., red blood cells) per square mm. Optionally, the blood sample is diluted with a dye solution.

[00161] Optionally, the sample or dyeing solution comprises one or more suitable dyes or dyes (which optionally comprise one or more fluorescent dyes). In some embodiments, the blood sample is selected from the whole blood sample, a red blood cell sample, a buffy coat sample, a plasma sample, a serum sample, a sample of any other blood fraction, or any combination thereof.

[00162] Optionally, the sample forms a monolayer on the surface of the sample carrier 18. In the context of the present disclosure, when referring to a monolayer of cells, it should be understood as encompassing the distribution of cells on a surface as a layer essentially unique, in which at least 50 percent (sometimes at least 60 percent, 70 percent, 80 percent or even 90 percent) of the cells are in direct contact with the bottom surface of the carrier and no more than 20 percent (sometimes no more than 10 percent or even no more than 5 percent) of the cells overlap each other (i.e., no more than the above-mentioned percentage of cells are partially or completely above each other ). Furthermore, when referring to a monolayer, it should be understood that at least 5 percent (sometimes at least 10 percent or even at least 20 percent) of the cells touch the bottom surface. For some applications, a monolayer is formed in accordance with techniques described in U.S. 9,329,129 to Pollak, which is incorporated herein by reference.

[00163] For some applications, before being transformed into an image, the body sample is dyed with one or more suitable dyes or dyes. Optionally, the one or more suitable dyes or dyes comprise one or more fluorescent dyes or dyes, and the dyed sample is excited under one or more lighting conditions suitable for detecting pathogens. As used herein, the term “suitable dye” should be construed expansively to include any dye or dye useful for detecting a pathogen of interest, including any fluorescent dye or dye. As used herein, a “suitable fluorescent dye or dye” should be expansively construed to comprise a dye or dye that has the ability to selectively bind to one or more types of nucleic acid (e.g., DNA only, RNA only , both DNA and RNA, etc.) and fluoresces under one or more particular lighting conditions, thereby allowing one or more types of nucleic acids to be discerned in a body sample. Suitable fluorescent dyes or dyes may comprise, for example, dyes or dyes that bind to DNA and do not bind to RNA, dyes or dyes that bind to RNA and do not bind to DNA, and dyes or dyes that bind to both DNA as RNA. Examples without limitation of fluorescent dyes or dyes include, for example, Acridine Orange, Hoechst dye, etc.

[00164] The particular lighting condition that causes a particular suitable fluorescent dye or dye to fluoresce is herein referred to as a “suitable lighting condition”, which should be interpreted expansively to comprise a lighting condition which, when used to excite a particular fluorescent dye or dye, causes fluorescence of the fluorescent dye or dye. In certain embodiments, the fluorescence emitted by the excited dye or dye may be discernible through the use of one or more different light filters that allow the discernment of fluorescence within a given wavelength range. Consequently, suitable lighting conditions can be used in view of such filters. Examples without limitation of suitable lighting conditions include, for example, field light, blue light and ultraviolet light. Additional non-limiting examples of suitable fluorescent dyes or dyes and suitable lighting conditions are described in U.S. 2012/0169863 by Bachelet and U.S. 2015/0037806 by Pollak, both applications of which are incorporated herein by reference.

[00165] As detailed above, in certain embodiments, the sample may be dyed with one or more dyes or dyes that allow differentiation between RNA and DNA in the sample (i.e., differential dyeing). Differential dyeing can be achieved, for example, by dyeing the sample with one or more specific target dyes or dyes. As used herein, a specific target dye or dye (e.g., a specific RNA or a specific DNA) is a dye or dye that under selected conditions would detectably dye the target chemical moiety so that it can be detected in the presence of other cellular components. In this context, detectably staining a target may mean that the dye or dye binds to the target with a higher affinity than other cellular components and/or that it provides a stronger signal (e.g., fluorescence) when associated with the target. target. Note that some dyes or dyes may dye more than one target, however, they can be differentiated, for example, based on the wavelength of fluorescence emitted and/or a wavelength used for excitation of the dye or dye. In some embodiments, a specific target dye or dye is a fluorescent dye or dye that upon binding to the target switches its emission wavelength from an original band to a switched band. In such cases, the target can be detected by a system configured to detect emission wavelengths within the switched band.

[00166] Differential dyeing can be used to determine the relative locations of DNA and RNA, as detailed below with reference to Example 1. Optionally, a single dye or dye (e.g., Acridine Orange) can be used with different lighting conditions, to provide differential tinting. Optionally, a combination of dyes or dyes is used, which comprises one or more DNA-specific dyes or dyes (e.g., Hoechst reagent) and one or more other dyes or dyes (e.g., Acridine Orange) configured to detect any nucleic acid (DNA and RNA).

[00167] For some applications, imaging information is informative of one or more fields of the body sample. As used herein, a “field” is a portion of the body sample to be imaged. Typically, this corresponds to an area at the bottom of a sample carrier that holds the sample. When images are captured at high magnification, only a fraction of the entire blood sample can be imaged at a time. Therefore, the pathogen detection system 10 virtually subdivides an area to be imaged into a plurality of fields, and each field is imaged separately, thereby obtaining a plurality of informative images of the body sample, with each image is informative of a respective field. Optionally, the imaged fields do not overlap, or their degree of overlap is less than 5 percent or less than 1 percent of the area. In certain embodiments, each field to be imaged is imaged under one or more different lighting conditions. Optionally, an image of each field is captured a plurality of times under different lighting conditions. For example, the field may be imaged at least once under lighting conditions to detect RNA-related fluorescence, at least once under lighting conditions to detect DNA-related fluorescence, and at least once under illuminated field.

[00168] Reference will now be made to Figure 3, which shows, by way of example without limitation, imaging information 300 consisting of a field of a blood sample stained with one or more suitable fluorescent dyes and excited under a suitable lighting condition. , in accordance with some applications of this application. As can be seen, due to the dye (or dyes), constituent elements 302 fluoresce, thereby appearing brighter (or, in some cases, a different color) than other non-fluorescent constituent elements 304 (which, in this case, comprise red blood cells) in the sample and allow discerning stained regions in the sample, some features thereof may be informative of some specific cell types in the sample.

[00169] In certain embodiments, the imaging information is informative of one or more sample constituent elements, including candidates (i.e., constituent elements that exhibit characteristics that indicate that they may be pathogens and, therefore, are candidates for being pathogens ) and non-candidates. For some applications, an element is identified as a candidate based on whether the element appears fluoresced when the sample is stained with a suitable fluorescent dye or dye and is excited by a suitable lighting condition, e.g., as described in U.S. 2012. /0169863 by Bachelet and/or in U.S. document 2015/0037806 by Pollak, both applications of which are incorporated herein by reference. Alternatively or additionally, an element may be identified as a candidate based on other criteria, such as its size, shape, color, proximity to other elements, etc. As used herein, the term “non-candidate” should be interpreted expansively to cover a sample constituent that is not a candidate.

[00170] Again, referring to Figure 2, in step 201, the processor 28 extracts from the one or more images, from the imaging information and/or a portion thereof, one or more sample information resources. of the body sample that are indicative of contextual information related to the body sample. Typically, a plurality of sample informative features are extracted. As used herein, “sample informational features” comprise features of the body sample that are not directed to a specific candidate and are useful for providing contextual information that can be used to determine the presence, probability of, or characteristics of a pathogenic infection. in the sample, including, in some modalities, the classification of specific candidates. By way of examples without limitation, sample informative features may comprise, for example, features related to non-candidate constituents in the sample, or features related to the quantity and/or distribution of cells of a given type in the sample. Features related to non-candidate constituents in the sample may comprise, for example, properties related to the size of one or more non-candidates (including relative size compared to an expected size, or to an observed size of one or more other cells), properties related to the shape of one or more non-candidates (including relative shape compared to an expected shape or to an observed shape of one or more other elements) and properties related to the intensity of one or more non-candidates (including relative strength compared to a expected or observed intensity of one or more other elements). As used herein, an “expected” value (e.g., of size, shape, and/or intensity) is such a value that can be known prior to analyzing imaging information with respect to a given sample. Such values comprise, for example, population statistical values that are known or can be calculated (e.g., for all human beings and/or any subgroup thereof, based on, for example, age, sex, race, ethnicity, etc.), optionally according to a specific condition (e.g. altitude, body sample treatment, etc.).

[00171] For some applications, sample informative features comprise features related to the distribution of candidates or pathogens within the sample or portions thereof. For example, if the number of candidates or pathogens found in a given image (or part of an image or a group of images covering a continuous portion of the sample) is significantly greater than the number of candidates or pathogens found in other parts of the same sample, this may indicate that the high concentration of candidates or pathogens found in one part of the sample may be a result of a local effect that should not affect the diagnosis of the sample. For example, a high concentration of candidates or pathogens (e.g., a high concentration of candidates overlapping red blood cells) in one part of the sample but not in other parts may be indicative of contamination, for example, from a droplet. of blood from another sample that entered the sample under investigation.

[00172] For some applications, some or all of step 201 is performed in a preprocessing stage in order to determine, for example, whether some of the imaging information is of low quality as measured by predetermined criteria (e.g., brightness, focus, etc.), in which case portions of the imaging information may be excluded from further processing (e.g., as described below herein with reference to Example 6).

[00173] In step 202, the computer processor 28 identifies one or more constituent elements within the sample as being candidates for a pathogen. As described above in this document, an element can be identified as a candidate based on the element that appears fluoresced when the sample is stained with a suitable fluorescent dye and excited by a suitable lighting condition, for example, as described in the document in the U.S. 2012/0169863 by Bachelet and/or in U.S. document 2015/0037806 by Pollak, both applications of which are incorporated herein by reference. Alternatively or additionally, an element may be identified as a candidate based on other criteria, such as shape, size, proximity to other elements (such as RBCs or other candidates), etc.

[00174] In step 203, the computer processor extracts from the one or more images, from the imaging information or/or from a portion thereof, one or more candidate information resources associated with one or more candidates identified. Typically, for each candidate, a plurality of candidate information features are extracted. As used herein, “candidate informational features” comprise candidate features (or, in some cases, constituent elements in close proximity to the candidate, as will be detailed below) useful for providing information to determine the likelihood of the given candidate being a pathogen. or a part of a pathogen.

[00175] By way of example without limitation, candidate information resources may comprise resources related to: a size of a candidate, a shape of a candidate, a movement of a candidate (based, for example, on a comparison of at least two overlapping images at least partially captured in sequence) and/or an intensity of a candidate.

[00176] For some applications, candidate information resources comprise a relative location of a candidate with respect to other sample constituents (e.g., a red blood cell). Alternatively or additionally, candidate informative features comprise a property of a cell (e.g., RBC) that overlaps at least partially with the candidate (and optionally also the amount of overlap), such as a cell size or shape that overlaps. the candidate. For some applications, features related to the size and shape of a candidate overlapping cell comprise a relative size and relative shape of the overlapping cell compared to an expected size or expected shape. As used herein, a cell is considered to overlap with a candidate, at least partially, if, in imaging information, at least a portion of the cell appears to be colocalized with at least a portion of the candidate (e.g., at least 20 percent or at least 25 percent of the candidate).

[00177] Optionally, candidate informative features may comprise features of other constituent elements (e.g., candidate pathogens and/or pathogens) that are found in close proximity to the candidate. In this context, “close proximity” can be predefined according to any suitable metric. For example, constituents in close proximity to the candidate may comprise constituents located within a distance of up to 2X in the opposite direction to the candidate, where X is an expected RBC diameter (e.g., average). Accordingly, in some embodiments, candidate information resources (including resources of the candidate, of a cell that overlaps the candidate, and/or resources of other constituents) may comprise or be limited to resources that are within close proximity to the candidate.

[00178] For some applications, imaging information, or a portion thereof, is processed for candidate informational feature extraction at least partially in a preprocessing stage. In certain embodiments, the preprocessing stage may comprise extracting sample informational features to obtain contextual information, and determining the imaging information that is used to extract candidate informational features in accordance with the obtained contextual information. For some applications, the portion of the imaging information that is used to extract candidate informational features and the portion of the imaging information that is used to extract sample informational features partially or completely overlap.

[00179] It should be noted that steps 201, 202 and 203 can be performed in any order. In accordance with some applications, steps 201, 202 and 203 are performed as a single step and/or are interconnected with each other. For some applications, some or all of steps 201, 202 and 203 are performed as a plurality of distinct steps.

[00180] Typically, based on the candidate informative feature (or candidate informative features) in combination with the sample informative feature (or sample informative features), the computer processor 28 classifies a probability of the body sample being infected with a pathogenic infection. For some applications, pathogenic infection is detected by implementing the additional steps indicated in Figure 2.

[00181] For some applications, once at least some candidate information features are extracted, in step 205, the processor 28 ranks the probabilities of respective candidates being pathogens, in accordance with the candidate data obtained for each respective candidate. As used herein, the term “probability of being a pathogen” should be interpreted expansively to cover a binary determination (e.g., a pathogen or a nonpathogen) or a scalar determination (e.g., a number, the value which reflects the estimated probability of the given candidate being a pathogen). In certain embodiments, processor 28 ranks the probabilities of the respective candidates being pathogens using the extracted sample informational features (e.g., the features extracted in step 201) in combination with the candidate informational features, as will be further detailed below. for example, with reference to Examples 1 and 2. This is indicated by the dashed arrow connecting step 201 to step 205, which indicates that step 201 is an optional entry in step 205.

[00182] Typically, subsequent to candidate classification (i.e., step 205), in step 207, processor 28 classifies a probability of the body sample being infected with a pathogenic infection. As used herein, the term “probability of the body sample being infected” should be interpreted expansively to cover a binary determination (e.g., whether infected or free) or a scalar determination (e.g., a number, the value that reflects the estimated probability of the given sample being infected). For some applications, the processor 28 classifies the sample based on the ranking of the candidates (extracted in step 205), in combination with the sample information resources (extracted in step 201), as will be further detailed below, for example, with reference to the Examples 1 and 3.

[00183] For some applications, in step 209, the processor 28 classifies the pathogenic infection as containing one or more given pathogen types, in accordance with one or more sample informational features and/or candidate informational features extracted.

[00184] Classifying the pathogenic infection as containing one or more given types of pathogen can be accomplished using information and/or resources that were obtained in one or more of steps 201, 203, 205 and 207, and/or performing one or more additional feature extraction and classification steps. For some applications, in order to classify the pathogenic infection, (a) candidates are classified according to the given pathogen types and (b) the total pathogenic infection is classified based on the classifications of the individual candidates. For some applications, sample information features are used to classify individual candidates according to the given pathogen types and/or to classify the total infection as containing the given pathogen types.

[00185] For some applications, classifying the pathogenic infection as containing one or more given types of pathogen comprises, for example, classifying the pathogenic infection so as to determine species or strains of pathogen contained in the sample, for example, as further detailed below with reference to Examples 4 and 5. Such determination may comprise or be limited to classifying the pathogen into a single species or strain, or into a group of several possible species or strains (at least one of which is contained in the sample) and/or ruling out a given species or strain (such as a species that is not contained in the sample). For some applications, the processor 28 classifies the pathogenic infection as containing one or more of Plasmodium, a given strain of Plasmodium, Plasmodium of a given age, and/or Plasmodium of a given age range.

[00186] For some applications, the computer processor generates output to the user (e.g., in the output device) that indicates whether or not the sample is infected with a pathogen, and indicates a classification of the infection. For some applications, the computer processor generates output that indicates that the presence of an infection within the body sample could not be determined with a sufficient degree of reliability, which indicates that a portion of the sample must be reimaged, and/or indicates that a portion of the sample must be reimaged using different settings (e.g., using different lighting, using a different dye, using a different or new sample preparation method, and/or using the use of different microscope settings). For some applications, in response to determining that the presence of an infection within the body sample could not be determined with a sufficient degree of reliability, the computer processor generates output that indicates that the user should take appropriate user action (e.g. , prepare a new sample and/or test the sample using an independent method, etc.). Alternatively or additionally, the computer processor automatically causes the microscope system to reimage a portion of the sample, causes the microscope system to reimage a portion of the sample using different settings (e.g., different focus or different field size) and/or modulating a frame rate at which microscope images are acquired by the microscope system.

[00187] Note that for some applications, sample informative features are not necessarily derived directly from the images. For example, sample informational features may comprise statistical or other information regarding candidates and/or other entities within the sample and/or general characteristics of the sample. Generally speaking, the scope of the present application comprises analyzing a sample at two levels, first at a candidate-by-candidate level and then at a more general level that is indicative of characteristics of the sample as a whole.

[00188] For some applications, based on candidate-level features, two or more sample informative features related to the body sample are extracted and a characteristic of the body sample is determined by processing the two or more sample informative features. Typically, at least some of the candidates are pathogenic candidates and candidate informative features regarding the pathogenic candidates are extracted. For some applications, candidate entities such as reticulocytes and/or platelets are additionally identified, and candidate information features with respect to these candidates are extracted. For some applications, sample information resources comprise several pathogenic candidates in the sample, type of pathogenic candidates in the sample, luminosity of candidates relative to background luminosity, a probability of candidates being pathogens, number of candidates that have a probability of being a pathogen that exceeds a threshold, number of candidates that have a probability of being a given type of pathogen that exceeds a threshold, several platelets in the sample, luminosity of platelets, several reticulocytes in the sample, several pathogen-infected reticulocytes in the sample, a proximity of candidates for RBCs and/or several RBCs in the sample.

[00189] In embodiments of the currently disclosed matter, fewer, more and/or different stages from those shown in Figure 2 may be performed. In embodiments of the currently disclosed subject matter, one or more stages illustrated in Figure 2 may be executed in a different order and/or one or more groups of stages may be executed simultaneously.

[00190] Several examples detailing specific applications without limiting the method detailed above will be provided in order to better understand the matter revealed.

Example 1: Using Reticulocyte Concentration as a Sample Informative Resource to Classify a Candidate and/or to Classify a Pathogenic Infection

[00191] As detailed above, in certain embodiments, the sample is stained to discern respective locations of DNA and RNA in the sample. Such dyeing may comprise, for example, using at least one DNA-specific dye and at least one RNA-specific dye, or at least one target-specific dye (either DNA or RNA) and at least one dye that dyes both DNA and RNA. . For some applications, in order to rank the probability of a candidate being a pathogen (step 205), the respective RNA and DNA staining sites in the sample are used by the processor to determine whether the staining pattern (or patterns) matches (or correspond) to the pattern (or patterns) expected for a pathogen.

[00192] Figure 4 schematically illustrates candidates 400a and 400b, each candidate shows an area stained for RNA (RNA portion 402) and an area stained for DNA (DNA portion 404). RNA portion 402 and DNA portion 404 may be differently dyed, for example, with the use of different dyes and/or different lighting, in order to discern the particular limits of each dyed area. As shown in Figure 4, in candidate 400a the DNA portion 404 completely overlaps the RNA portion 402, while in candidate 400b the DNA portion 404 partially overlaps the RNA portion 402.

[00193] A candidate that appears to have overlapping DNA and RNA, at least partially, may be a pathogen. However, the appearance of overlapping RNA and DNA-stained regions may also be caused by a different entity or entities, including, for example, a different cell type, or two separate bodies (one containing DNA and the other containing RNA). ) supposedly positioned one above the other.

[00194] Mature red blood cells have no detectable DNA or RNA and therefore do not fluoresce when stained with nucleic acids. In contrast, Plasmodium trophozoites (which are a type of pathogen) can be detected as DNA-containing and RNA-containing bodies within red blood cells. Therefore, for some applications, in order to identify red blood cells that contain pathogens, a dye substance that dyes both DNA and RNA (such as Acridine Orange) is used. Alternatively or additionally, a dye that stains only DNA (such as a Hoechst dye) is used.

[00195] Howell Jolly bodies are DNA-containing bodies that can be found in red blood cells in some unsanitary conditions. In some cases, the presence of Howell Jolly bodies in a sample may increase the chance of a false positive determination of a pathogen infection. Even if a DNA-specific dye is used in combination with a dye that stains both DNA and RNA, Howell Jolly bodies can cause a false positive determination of a pathogen infection. Therefore, in some embodiments, differentiating between red blood cells that contain Howell Jolly bodies and red blood cells that contain pathogens may be beneficial.

[00196] Young red blood cells, called reticulocytes, are sometimes found in the blood. These cells contain only RNA bodies. A positive correlation is known between the presence of Howell Jolly bodies in red blood cells and a greater than normal number of reticulocytes. Therefore, for some applications, sample informative features comprise features that are indicative of a reticulocyte concentration in a blood sample. (Note that a Plasmodium infection also raises the reticulocyte count for a patient. However, this increase (e.g., about 5%) is typically much smaller than the typical increase in patients who have a high reticulocyte count. Howell Jolly body (which can be about ten times higher).Consequently, a threshold for determining a high reticulocyte count is typically somewhat higher than the average value for humans (or a given subpopulation thereof).

[00197] Based on the identification of a high concentration of reticulocytes, the probability of pathogenic candidates being Howell Jolly bodies increases. On the other hand, the probability that the candidates are pathogens decreases and the probability that the sample is infected decreases. Therefore, for some applications, the computer processor sets a threshold for a positive determination of an infection, based on the reticulocyte concentration. For example, several reticulocytes detected simultaneously with low parasitemia (e.g., less than 200 parasites/microliter of blood) can be interpreted as being indicative of a high probability of a false positive (i.e., the sample is not infected).

[00198] Alternatively or additionally, based on the reticulocyte concentration, in order to rank the probability of a candidate being a pathogen (step 205 of Figure 2), the processor assigns more weight to the relative positions of DNA and/or RNA within the candidate given RBC, rather than simply the presence of DNA and/or RNA within the RBC. Alternatively or additionally, based on reticulocyte concentration, in order to rank the probability of a sample being infected (step 207 of Figure 2), the processor assigns more weight to extracellular Plasmodium candidates rather than intracellular Plasmodium candidates (which could be Howell Jolly bodies).

Example 2: Using Candidate Distribution within a Sample as a Sample Information Resource to Rank a Candidate

[00199] Candidates within a sample are expected to be evenly distributed. Therefore, for some applications, a distribution of candidates within the sample that differs significantly from an expected uniform distribution is used as an informative sample feature. For example, if there are significant candidate clusters, the clusters may be foreign bodies associated with the sample carrier rather than a portion of the blood sample, or they may indicate that a different infected sample has contaminated the sample being analyzed (e.g., spillage). if from an adjacent chamber on a sample carrier). In response to detecting a non-uniform candidate distribution, candidates that are within localized clusters may be a given lower score (i.e., may be classified as being less likely to be pathogens). For example, if sample informative features are indicative of candidate clustering, the processor may use the distance from the cluster center (or cluster centers) of any given candidate as a feature for classifying the candidate.

Example 3: Platelet Concentration as a Sample Informative Resource to Classify a Sample as Infected

[00200] Platelets typically appear as small extracellular RNA bodies, although some platelets may appear to be overlapping with cells because they are positioned over or under a cell when the sample is imaged. A normal platelet concentration is typically between 150,000 and 400,000 platelets per microliter of whole blood.

[00201] It is known that platelet concentration can be affected by Plasmodium infection, its severity and the species of Plasmodium, as well as by other unrelated conditions (including medical conditions, treatments and medications). Consequently, for some applications, the number and/or concentration of platelets in a sample is used as an informative sample feature, and, for example, may be used as an input in classifying the probability of the sample being infected.

Example 4: Platelet Concentration as an Informational Resource of a Pathogen Species Informational Sample

[00202] As mentioned in the context of Example 3, the number and/or concentration of platelets may be correlated with a specific species of pathogen, for example, a low platelet count has been shown to correlate with a Plasmodium falciparum infection to a significantly greater than Plasmodium vivax infection. For some applications, in accordance with step 209 of Figure 2, the number and/or concentration of platelets in a blood sample is used as an input to classify a pathogenic infection as containing a given type of pathogen.

Example 5: Red Blood Cell Size and Shape as a Specimen Informational Resource

[00203] Some pathogens alter the morphology of infected cells. For example, some pathogens (e.g., relatively mature tropozoites of Plasmodium vivax and Plasmodium ovale) cause an enlargement of infected red blood cells, sometimes by about twice that of uninfected cells. Other pathogens (e.g., Plasmodium malariae) reduce the size of infected red blood cells. Still other pathogens (e.g., Plasmodium falciparum) do not enlarge infected cells or reduce their size. For some applications, the sizes of red blood cells that appear to be infected within a blood sample are used as a sample information feature that is indicative of the sample that is infected (e.g., in step 207 of Figure 2) and/or is indicative of an identity of the pathogen (e.g., in step 209 of Figure 2).

[00204] For example, a blood sample infected by Plasmodium vivax and/or Plasmodium ovale is expected to include infected red blood cells that are significantly increased. A blood sample infected with Plasmodium malariae, on the other hand, would be expected to include infected red blood cells that are significantly decreased in size. Therefore, for some applications, the detection of such increased and/or decreased cells is used as a sample informative feature that is indicative of the sample being infected (e.g., in step 207 of Figure 2) and/or is indicative of a pathogen identity (e.g., in step 209 of Figure 2).

[00205] In another example, Plasmodium ovale can cause infected red blood cells to become more oval than uninfected red blood cells which tend to appear round. Consequently, one or more of the following sample informative features may be interpreted as being indicative of the sample being infected and/or being infected with Plasmodium ovale: the presence of oval red blood cells in the sample, the presence of a larger than expected portion of oval red blood cells and/or the presence and/or number of infected red blood cells that appear to be oval.

[00206] Oval (or oval) features (e.g., height versus width) can be used in a classification in a weighted way. For example, the weight given to an oval feature that is close to an expected value may be greater than if the value is closer to an expected value for an uninfected RBC or values that are significantly more deviated than the expected value for infected red blood cells (for example, if the oval appears to be rod-like).

[00207] It should be noted that a determination that an infected RBC (or a group of infected RBCs), or potentially infected RBC, is different in any given property (e.g., size, shape, etc.) than the general population of red blood cells (or different from uninfected cells), and/or a determination of a degree of such difference, is typically achieved using any acceptable statistic. For example, an average size of two groups may be used and/or an average size of one group may be used in relation to a given percentage of the other group. Optionally, a plurality of statistics is used. Optionally, one or more of the values for RBCs (or for uninfected RBCs) are taken from statistics known from the general population or a subgroup thereof. In some embodiments, one or more statistics of all RBCs in the sample (or a portion thereof) may be used, rather than using only uninfected RBCs. For example, this can be used in cases where the portion of infected RBCs within the sample is small enough.

[00208] Optionally, a determination that an infected or potentially infected RBC is different or relatively different in any given property (e.g., size, shape, etc.) is made by comparing the given property of the infected RBC or potentially infected RBC with properties of one or more clean red blood cells in the sample. Thus, the properties of one or more clean red blood cells can also be used as sample information resources to determine the probability that a candidate is a pathogen, to determine the probability that a sample is infected, and/or to classify the species of a pathogen.

[00209] It should be noted that RBC features (e.g., features related to RBC size and/or shape) may also be used in candidate classification (e.g., as a candidate informative feature used in step 205 of Figure 2 ) compared to an expected value. By way of non-limiting example, candidates that appear to be within (or colocalized with) RBCs that are relatively larger or smaller than an expected value or have a shape that is different from an expected shape (e.g., oval instead of round ) are more likely to be pathogens.

[00210] It should also be noted that features of RBCs in the sample (e.g., features related to RBC size and/or shape) may be used in sample classification (e.g., as a sample informative feature used in step 207 of Figure 2). By way of non-limiting example, such resources comprise a statistic taken for a group of RBCs in the sample (e.g., a statistic for RBCs that appear infected in the sample, a statistic for uninfected RBCs in the sample, and/or a statistic for RBCs in the sample). sample in general). A non-limiting example comprises comparing a statistic of RBCs that appear infected in the sample (e.g., size) with an expected value (e.g., average human RBC size) or with a corresponding statistic for RBCs in the sample overall. When red blood cells that appear infected are found to be larger or smaller than the compared value, this can be used as an indication that the sample is infected.

[00211] Echinocytes are red blood cells that have abnormal cell membranes with many small, symmetrically spaced spiny projections. Acanthocytes also have abnormal spiny projections, but they are irregular and asymmetrical.

[00212] In some cases, a Plasmodium infection causes the appearance of echinocytes or acanthocytes. Other causes for such formats can be other pathologies and even prolonged exposure to some solutions (for example, dye solutions). Additionally, some strains of Plasmodium cause greater deformity than other strains. For example, Plasmodium vivax is typically more misshapen than Plasmodium falciparum, while each of Plasmodium vivax and Plasmodium ovale is typically more misshapen than each of Plasmodium malariae and Plasmodium falciparum. Therefore, for some applications, the presence of such formats in a sample is used as a sample informative feature that is indicative of the sample being infected (e.g., in step 207 of Figure 2) and/or is indicative of an identity of the pathogen (e.g., in step 209 of Figure 2).

[00213] In some strains of Plasmodium, there is a positive correlation between a degree of red blood cell deformity and the age of the infecting pathogens. Therefore, for some applications, the presence of such shapes in a sample is used as an informative sample feature that is indicative of the age of detected pathogens (e.g., in step 209 of Figure 2).

Example 6: Using Dye Quality as a Sample Informational Resource

[00214] Dyeing a biological sample is known to be a time-dependent process. Once cells are exposed to a dye, it takes time for the dye to penetrate the cells and reach its target site (or sites) within the cells. During this time, staining may become sharper and/or more localized with sharper intensity gradients (especially if the sample is not washed before being imaged). Typically, this process follows a saturation curve. In a first phase, dyeing increases relatively quickly until some degree of dyeing is achieved (i.e., a rapid dyeing phase). From now on, quality still increases, but relatively slowly for another period of time (i.e. a slow dyeing phase). At a later time, the dyeing quality may deteriorate, for example, due to photo discoloration (in a fluorescent dye) and/or due to the dye slowly diffusing through the sample, away from the target (or targets).

[00215] For some applications, imaging a plurality of fields (e.g., 200 fields or more, at least some of which may be imaged more than once) from a body sample, such as a blood sample, may take some time. minutes (e.g. 2 to 3 minutes or more). For Hoechst staining (which is used, for example, in detecting Plasmodium infection), the rapid staining phase may take 30 minutes or more.

[00216] Consequently, when imaging a plurality of fields is performed over the time scale described above, there may be significant variation in dyeing quality between fields. This variation can affect the diagnostic result (for example, changing an intensity value, an intensity gradient value and/or a threshold for infection). For some applications, in order to represent variation in dyeing quality, the dyeing quality of the sample is used as a sample information resource in accordance with the techniques described herein.

[00217] For some applications, an average dyeing quality across a plurality of images is determined and a dyeing quality parameter for each image is determined, based on the relative dyeing quality of the image compared to the average value. The staining quality parameter can be used as an informative sample resource, for example, to rule out from being used in further analysis images that were taken too early or too late, when the staining is not informative enough for diagnosis. Alternatively or additionally, the dyeing quality parameter can be used to adjust one or more thresholds for different fields or groups of fields based on the dyeing quality parameter and/or introduce the dyeing quality parameter into the candidate ranking, from so that candidates from a plurality of fields that have different dyeing qualities can be compared. For example, candidate ranking can be normalized using dye quality.

[00218] For some applications, in response to the dyeing quality parameter, a frame rate at which images of the body sample are acquired is modulated. For example, in response to detecting that the dye is of poor quality, images may be acquired at a higher frame rate and vice versa. Alternatively or additionally, in response to the dye quality parameter, the number of times each field is imaged is modulated. In general, the scope of the present invention comprises using sample information resources as an input to rule out some images from being used, to modulate a frame rate at which images are acquired, and/or to modulate the number of times each image field is imaged.

[00219] Optionally, when using two or more dyes, the variation in dyeing quality may be different between the dyes. For example, Acridine Orange dyeing can be completed when Hoechst dyeing is still in the rapid dyeing phase. Consequently, for some applications, the dye quality parameter is treated separately for each dye and/or the relative dye quality between dyes can be used as a dye quality parameter.

Example 7: Using Candidate Affinity for Reticulocytes as a Sample Informative Feature

[00220] Plasmodium vivax and Plasmodium ovale have an affinity for infecting reticulocytes, through mature red blood cells. On the other hand, Plasmodium falciparum infects all red blood cells equally, while Plasmodium malariae has an affinity for mature red blood cells. Consequently, for some applications, a relationship between the number of pathogenic candidates associated with reticulocytes and the number of pathogenic candidates associated with mature red blood cells is determined and this is used as an informative sample resource.

[00221] As described above, for some applications, according to step 209 of Figure 2, the processor classifies a pathogen as containing a given type of pathogen. For some applications, once a sample is determined to be infected, the processor classifies the pathogen as containing a given pathogen type, based on the estimated ages of infected RBCs and/or based on a relationship between the number of pathogenic candidates associated with reticulocytes and the number of pathogenic candidates associated with mature red blood cells. For example, if pathogens have an affinity for reticulocytes, this can be used as an indication that the detected pathogen is Plasmodium vivax. Alternatively, an essentially uniform distribution of the pathogen in RBCs of all ages (proportional to the relative abundance of RBCs of different ages) can be used as an indication that the detected pathogen is Plasmodium falciparum.

Example 8: Using Contamination Detection as a Sample Informational Resource

[00222] An image of a blood sample may contain foreign objects that are irrelevant to the diagnosis and may be due to, for example, dust or flaws in a solution or tool used to prepare or maintain the blood sample (e.g., a solution dilution chamber and/or a sample chamber used to house the sample). Such objects may comprise objects that appear similar to a pathogen and/or objects that are not similar to a pathogen.

[00223] For some applications of the present invention, the computer processor is configured to extract, from the one or more images, at least one sample information feature that is indicative of contextual information related to the body sample. Based at least partially on the extracted sample informational resource, the computer processor identifies that there is a defect associated with the body sample disposed in the sample carrier and classifies a source of the defect (e.g., as the sample carrier, a given portion sample carrier, the sample itself, and/or a diluent in which the sample was diluted). The computer processor generates an output at the output device that is indicative of the source of the defect.

[00224] For example, such output may be generated based on a sample informative feature that is indicative of the presence of foreign objects within the sample. The source of the foreign objects is expected to affect their concentration and distribution in the sample. For example, if the source is the blood sample itself, then the amount of foreign objects that is detected is typically proportional to the size of the sample. For some applications, in response to the detection that foreign objects have this characteristic, a diagnostic threshold is adjusted. For example, the threshold for the number of pathogens within a sample that is sufficient to consider the sample to be infected may be a function of a ratio of the concentration of foreign objects within the sample to that of red blood cells.

[00225] For some applications, the processor is configured to determine that the source of foreign objects is limited to a given chamber, a given set of cameras, a given field, or a given set of fields. For such cases, the computer processor may assign a lower weight to features that are detected in the affected chambers or fields and/or may use data from other cameras or fields as inputs to analyze features detected in the affected chambers and fields.

[00226] For some applications, the computer processor is configured to detect that the source of foreign objects is a diluent in which the sample has been diluted (e.g., in response to the detection of foreign objects with characteristics that are common to a plurality of samples and/or chambers). In such cases, the processor may generate output that indicates the likely source of the foreign objects. For some applications, the computer processor is configured to detect cross-contamination between chambers and to generate output that indicates this is the case.

[00227] It is noted that, although some applications of the present invention have been described in relation to detecting a pathogen infection within a body sample, the scope of the present invention comprises performing similar techniques in relation to the identification of other components or other entities within a body sample. For example, techniques similar to those described herein can be used to detect the concentration of a given entity within a blood sample by (a) extracting one or more candidate informational features associated with an element that is a candidate of the given entity, (b) extracting one or more sample informational features that are indicative of contextual information related to the body sample, and (c) processing the candidate informational feature in combination with the sample informational feature. For some applications, the sample is a sample comprising blood and the candidates are candidate entities within the blood, such as platelets, leukocytes, anomalous leukocytes, circulating tumor cells, red blood cells, reticulocytes, Howell Jolly bodies, etc. For some such applications, a blood count (e.g., a complete blood count) is performed by identifying such candidates and performing techniques as described herein.

[00228] For some applications, the sample is a different body sample and the techniques described herein are used to identify a pathogen infection within the sample. For example, the techniques described herein can be used to identify Mycobacterium tuberculosis within a saliva sample. Alternatively or additionally, the sample is a different body sample and the techniques described herein are used to identify abnormal cells within the sample. For example, the techniques described herein can be used to identify cancer cells in a PAP test or in a urine sample.

[00229] In general, it is noted that, although some applications of the present invention have been described in relation to a blood sample, the scope of the present invention comprises applying the apparatus and methods described herein to a variety of samples.

[00230] For some applications, the sample is a body sample, such as blood, saliva, semen, sweat, sputum, vaginal fluid, feces, breast milk, bronchoalveolar lavage, gastric lavage, tears and/or nasal discharge. The body sample can be from any living creature and is typically from warm-blooded animals. For some applications, the body sample is a sample from a mammal, for example, a human body. For some applications, the sample is taken from any domestic animal, warm-blooded zoo animals, and warm-blooded farm animals, including, but not limited to, dogs, cats, horses, cows, and sheep. Alternatively or additionally, the body sample is taken from warm-blooded animals that act as disease vectors, including deer or rats.

[00231] For some applications, techniques similar to those described above are applied to a non-body sample. For some applications, the sample is an environmental sample, such as a water sample (e.g., groundwater), surface collection, soil sample, air sample, or any combination thereof. In some embodiments, the sample is a food sample, such as a meat sample, a dairy sample, a water sample, a washing liquid sample, a beverage sample, and any combination thereof.

[00232] Applications of the invention described herein may take the form of a computer program product accessible from a computer-usable or computer-readable medium (e.g., a non-transitory computer-readable medium) that provides code for program for use by or in connection with a computer or any instruction execution system, such as a computer processor 28. For purposes of this description, a computer-usable or computer-readable medium may be any device that can comprise, store , communicate, propagate or transport the program for use by or in connection with the instruction executing system, apparatus or device. The media may be an electronic, magnetic, optical, electromagnetic, infrared or semiconductor system (or apparatus or device) or a propagation medium. Typically, computer-usable or computer-readable media is non-transitory computer-usable or computer-readable media.

[00233] Examples of a computer-readable medium comprise a semiconductor or solid-state memory, magnetic tape, a removable computer floppy disk, a random access memory (RAM), a read-only memory (ROM), a hard magnetic disk, and an optical disc. Current examples of optical discs include compact disc read-only memory (CD-ROM), compact disc read/record (CD-R/W), and DVD.

[00234] A data processing system suitable for storing and/or executing program code will include at least one processor (e.g., computer processor 28) coupled directly or indirectly to memory elements (e.g., memory 29) through a system bus. The memory elements may comprise local memory employed during actual execution of the program code, raw storage, and cache memories that provide temporary storage of at least some program code in order to reduce the number of times the code needs to be retrieved from a raw storage during execution. The system may read instructions of the invention from program storage devices and follow those instructions to execute the methodology of embodiments of the invention.

[00235] Network adapters may be coupled to the processor to enable the processor to become coupled to other processors or remote printers or storage devices via private or public intervening networks. Modems, cable modems, and Ethernet cards are just some of the currently available types of network adapters.

[00236] Computer program code for performing operations of the present invention may be written in any combination of one or more programming languages, including an object-oriented programming language, such as Java, Smalltalk, C++ or the like, and programming languages. Conventional procedural programming, such as the C programming language or similar programming languages.

[00237] It will be understood that blocks of the flowchart shown in Figure 2 and combinations of blocks in the flowchart can be implemented by means of computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, a special purpose computer, or other programmable data processing apparatus to produce a machine, so that the instructions, which are executed through of the computer processor (e.g., computer processor 28) or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowcharts and/or algorithms described in this application. Such computer program instructions may also be stored on a computer-readable medium (e.g., a non-transitory computer-readable medium) that can direct a computer or other programmable data processing apparatus to a function in a particular manner, in a so that instructions stored on computer-readable media produce an article of manufacture comprising instruction means that implement the function/act specified in the flowchart blocks and algorithms. Computer program instructions may also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, so that the instructions that are executed on the computer or other programmable device provide processes for implementing the functions/acts specified in the flowcharts and/or algorithms described in this application.

[00238] Computer processor 28 is typically a hardware device programmed with computer program instructions to produce a special purpose computer. For example, when programmed to perform the algorithms described with reference to Figure 2, the computer processor 28 typically acts as a special purpose sample analysis computer processor. Typically, the operations described herein that are performed by means of computer processor 28 transform the physical state of memory 30, which is an actual physical article, to have a different magnetic polarity, electrical charge, or the like, depending on the technology. of the memory that is used.

[00239] Unless specifically stated otherwise, as is evident from the discussions herein, by all discussions of the specification that use terms such as “which processes”, “which executes”, “which obtains”, “ that determines”, “that classifies”, “that stores”, “that selects” or similar, refer to the action (or actions) and/or process (or processes) of a computer that manipulates and/or transforms data into other data, wherein said data is represented as physical quantities, such as electronic, and/or said data represents physical objects. The terms “computer” and “processor” should be expansively interpreted to cover any type of electronic device with data processing capabilities comprising, by way of non-limiting example, the system disclosed in this application.

[00240] It should be understood that the term “non-transient” is used herein to exclude transient propagation signals, but to otherwise understand any volatile or non-volatile computer memory technology suitable for the subject matter disclosed herein.

[00241] Typically, the computer processor generates output onto an output device 34. The output may be provided in any acceptable form, including a graph, a graph or text displayed on a monitor of a control unit, a printout, as a voice message or on a user's smart phone display, to accept processed data from the processing utility and display information regarding obtained structural features and/or associated values that determine the presence and, optionally, the identity of an infection pathogenic, with the use of lists, tables, graphs, etc. The output device may comprise a monitor which is connected to a printer to print the output.

[00242] User interface 32 may be used to control the operation of system 10 and/or computer processor 28, including, inter alia, entering data regarding the body sample examined (e.g., source, date, location, etc. .), control system operating conditions, types of dye used, number of images to be taken, time interval between images, etc.

[00243] At times, image analysis by the computer processor may involve adjusting or normalizing image brightness based on the degree of staining of the sample. These may be based, for example, on identifying one or more of the brightest and/or darkest pixel values in the image or set of images (e.g. corresponding to a particular sample), average brightness of the brightest area, and /or darker and/or image histogram. Such features can be extracted from a representative image (not necessarily the one being normalized) or from statistical analysis of multiple images. The features used for normalization can be based on a single image or multiple images, which can be captured using different excitation wavelengths (e.g. Acridine Orange which provide different colors under illumination wavelengths). many different). Image brightness can also be adjusted using other means of control, such as image capture component exposure time and/or lighting brightness.

[00244] The conditions of microscope system 11 may be in order to control the timing of image acquisition, for example, to allow sufficient incubation time with the one or more dyes or dyes, as well as operation with optical configurations other than excitation and/or emission wavelengths in order to image the dyed sample in various colors or fluorescence spectra.

[00245] The components of the pathogen detection system, namely, imaging module 14, computer processor 28, output device 34, etc. they may be directly connected to each other (e.g., directly via a wire) or one or more of the components may be remote from one or more other components. For example, the imaging module may send data to the computer processor 28 via an intranet or via the internet to enable processing at a remote location.

[00246] Examples of systems that can be used to perform the techniques of the present disclosure are described in document in WO 2012/090198 to Bachelet and in document in US 2014/0347459 to Greenfield, both applications of which are incorporated herein by reference title.

[00247] Therefore, the concepts of the following invention are provided, according to some applications of the present invention:

[00248] Concept of the invention 1. A method for detecting a pathogenic infection in a body sample, wherein the method comprises:

[00249] store in a memory imaging information related to the body sample, wherein at least a portion of the imaging information is informative of one or more pathogenic candidates in the sample,

[00250] provide, by means of a processor operatively coupled to memory, a first processing of a first part of the imaging information, wherein the first processing comprises: extracting the at least one sample informative feature and processing the at least one informative feature sample extracted to obtain context data indicative of contextual information related to the sample,

[00251] provide, through the processor, a second processing of a second part of the imaging information, wherein the second processing comprises: identifying at least one pathogenic candidate in the sample; extracting at least one candidate informative feature associated with the identified candidate and processing the at least one extracted candidate informative feature to obtain candidate data indicative of at least one classification property of the candidate,

[00252] providing, by means of the processor, a first classification, wherein the first classification comprises classifying the at least one candidate identified as a pathogen or a non-pathogen according to at least the candidate data obtained,

[00253] provide, through the processor, a second classification, wherein the second classification comprises classifying the sample as infected or clean according to at least the results of the first classification and the context data obtained,

[00254] wherein a pathogenic infection in the body sample is determined based on the results of the second classification.

[00255] Concept of the invention 2. The method, according to the concept of the invention 1, in which the first classification is additionally carried out according to the context data obtained.

[00256] Concept of the invention 3. The method, according to the concept of the invention 1, in which the second classification is additionally carried out according to the candidate data obtained for at least one identified candidate.

[00257] Concept of the invention 4. The method, according to any one of the concepts of the invention 1 to 3, which further comprises providing, by means of the processor, a classification of at least one pathogen in the sample according to at least the context data obtained.

[00258] Concept of the invention 5. The method, according to the concept of the invention 4, in which the classification of the at least one pathogen comprises determining the species of the at least one pathogen.

[00259] Concept of the invention 6. The method, according to any one of the concepts of the invention 1 to 5, wherein the at least one candidate informative feature is selected from the group consisting of a feature related to: a size of the candidate, a candidate shape, a candidate movement, a candidate intensity, a candidate location within the sample, and a property of a cell that overlaps the candidate.

[00260] Concept of invention 7. The method, according to the concept of invention 6, in which the cell is a red blood cell and the property comprises at least one of: a size-related property and a shape-related property.

[00261] Concept of the invention 8. The method, according to any of the concepts of the invention 1 to 7, wherein the at least one sample informative feature is selected from the group consisting of a feature relating to: a size , shape or intensity of one or more non-candidate constituents in the sample, a number of cells of a given cell type, a distribution of cells of a given cell type, and a distribution of candidates.

[00262] Concept of the invention 9. The method, according to any of the concepts of the invention 1 to 8, wherein the body sample is selected from a blood sample, a diluted blood sample, a sample comprising predominantly red blood cells and a diluted sample comprising predominantly red blood cells.

[00263] Concept of the invention 10. A method for detecting a pathogen in a body sample, wherein the method comprises:

[00264] store in a memory imaging information related to the body sample, wherein at least a portion of the imaging information is informative of one or more pathogenic candidates in the sample,

[00265] provide, by means of a processor operatively coupled to memory, a first processing of a first part of the imaging information, wherein the first processing comprises: extracting the at least one sample informative feature and processing the at least one informative feature sample extracted to obtain context data indicative of contextual information related to the sample,

[00266] provide, through the processor, a second processing of a second part of the imaging information, wherein the second processing comprises: identifying at least one pathogenic candidate in the sample; extracting at least one candidate informative feature associated with the identified candidate and processing the at least one extracted candidate informative feature to obtain candidate data indicative of at least one classification property of the candidate,

[00267] providing, by means of the processor, a first classification, wherein the first classification comprises classifying the at least one candidate identified as a pathogen or a non-pathogen according to at least the candidate data and the context data obtained.

[00268] Concept of invention 11. The method, according to the concept of invention 10, which additionally comprises:

[00269] provide, through the processor, a second classification, wherein the second classification comprises classifying at least one pathogen in the sample according to at least the obtained candidate data.

[00270] Concept of the invention 12. The method, according to the concept of the invention 11, wherein the second classification comprises determining the species of the at least one pathogen.

[00271] Concept of the invention 13. The method, according to any of the concepts of the invention 10 to 12, which further comprises: providing, by means of the processor, a pre-processing of the imaging information, wherein the pre-processing comprises determining the imaging information to be included in at least one tooth the first part and the second part, wherein the preprocessing comprises extracting the at least one sample informative feature from the imaging information and processing the at least one sample informative feature. sample extracted to obtain context data indicative of contextual information related to the sample, and wherein the determination is made according to the context data obtained.

[00272] Concept of the invention 14. The method, according to any of the concepts of the invention 10 to 13, wherein the at least one candidate informative feature is selected from the group consisting of a feature related to: a size of the candidate, a shape of the candidate, a movement of the candidate, a intensity of the candidate, a location of the candidate within the sample, and a property of a cell that overlaps the candidate.

[00273] Concept of invention 15. The method, according to the concept of invention 14, wherein the cell is a red blood cell and the property comprises at least one of: a size-related property and a shape-related property.

[00274] Concept of the invention 16. The method, according to any of the concepts of the invention 10 to 15, wherein the at least one sample informative feature is selected from the group consisting of a feature relating to: a size , shape or intensity of one or more non-candidate constituents in the sample, a number of cells of a given cell type, a distribution of cells of a given cell type, and a distribution of candidates.

[00275] Concept of the invention 17. The method, according to any of the concepts of the invention 10 to 16, wherein the body sample is selected from a blood sample, a diluted blood sample, a sample comprising predominantly red blood cells and a diluted sample comprising predominantly red blood cells.

[00276] Concept of the invention 18. A system for detecting a pathogenic infection in a body sample comprising:

[00277] a memory operatively coupled to a digital microscope and configured to store imaging information captured by the digital microscope, imaging information relating to a body sample, wherein at least a portion of the imaging information is informative of one or more candidates pathogens in the sample; It is

[00278] a processor operationally coupled to memory and configured to:

[00279] processing, in a first processing, a first part of the imaging information, wherein the first processing comprises: extracting at least one sample informative feature and processing the at least one sample informative feature

[00280] extracted to obtain context data indicative of contextual information related to the sample,

[00281] processing, in a second processing, a second part of the imaging information, wherein the second processing comprises: identifying at least one pathogenic candidate in the sample; extracting at least one candidate informational feature associated with the identified candidate and processing the at least one extracted candidate informational feature to obtain candidate data indicative of the at least one classification property of the candidate, classifying, in a first classification, the at least one candidate identified as a pathogen or a non-pathogen according to at least the obtained candidate data,

[00282] classify, in a second classification, the sample as infected or clean according, at least, to the results of the first classification and the context data obtained,

[00283] wherein a pathogenic infection in the body sample is determined based on the results of the second classification.

[00284] Concept of the invention 19. The system, according to the concept of the invention 18, in which the first classification is additionally carried out according to the context data obtained.

[00285] Concept of the invention 20. The system, according to any of the concepts of the invention 18 or 19, in which the second classification is additionally carried out according to the candidate data obtained for at least one identified candidate.

[00286] Concept of the invention 21. The system, according to any of the concepts of the invention 18 to 20, wherein the processor is additionally configured to, before the first and second processing:

[00287] pre-processing the imaging information, wherein the pre-processing comprises determining the imaging information to be included in at least one tooth the first part and the second part, wherein the pre-processing comprises extracting at least one sample information resource of the imaging information and processing the at least one extracted sample information resource to obtain context data indicative of contextual information related to the sample, and wherein the determination is made according to the obtained context data.

[00288] Concept of the invention 22. The system, according to any of the concepts of the invention 18 to 21, wherein the processor is further configured to classify a pathogen in the sample according to at least the context data obtained.

[00289] Concept of the invention 23. The system, according to any of the concepts of the invention 18 to 22, wherein the at least one candidate informational resource is selected from the group consisting of a resource related to: a size of the candidate, a candidate shape, a candidate movement, a candidate intensity, a candidate location within the sample, and a property of a cell that overlaps the candidate.

[00290] Concept of invention 24. The system, according to the concept of invention 23, in which the cell is a red blood cell and the property comprises at least one of: a size-related property and a shape-related property.

[00291] Concept of the invention 25. The system, according to any one of the concepts of the invention 18 to 24, wherein the at least one sample informative resource is selected from the group consisting of a resource related to: a size , shape or intensity of one or more non-candidate constituents in the sample, a number of cells of a given cell type, a distribution of cells of a given cell type, and a distribution of candidates.

[00292] Concept of the invention 26. The system, according to any of the concepts of the invention 18 to 25, wherein the body sample is selected from a blood sample, a diluted blood sample, a sample comprising predominantly red blood cells and a diluted sample comprising predominantly red blood cells.

[00293] Concept of the invention 27. A computer program product implanted on a non-transitory computer-usable medium that has computer-readable program code embedded therein to cause the computer to perform a method for detecting a pathogenic infection in a body sample, where the method comprises:

[00294] store in a memory, comprised or operatively coupled to the computer, imaging information relating to the body sample, wherein at least a portion of the imaging information is informative of one or more pathogenic candidates in the sample,

[00295] provide, by means of a processor comprised or operatively coupled to the computer, and operatively coupled to memory, a first processing of a first part of the imaging information, wherein the first processing comprises: extracting at least one informational resource

[00296] sample and processing the at least one extracted sample informative feature to obtain context data indicative of contextual information related to the sample,

[00297] provide, through the processor, a second processing of a second part of the imaging information, wherein the second processing comprises: identifying at least one pathogenic candidate in the sample; extracting at least one candidate informative feature associated with the identified candidate and processing the at least one extracted candidate informative feature to obtain candidate data indicative of at least one classification property of the candidate,

[00298] providing, by means of the processor, a first classification, wherein the first classification comprises classifying the at least one candidate identified as a pathogen or a non-pathogen in accordance with at least the obtained candidate data, and

[00299] provide, through the processor, a second classification, wherein the second classification comprises classifying the sample as infected or clean according to at least the results of the first classification and the context data obtained.

[00300] It will be appreciated by those skilled in the art that the present invention is not limited to what has been particularly shown and described above. Instead, the scope of the present invention comprises both combinations and subcombinations of the various features described above, as well as variations and modifications thereof that are not in the prior art, which may occur to persons skilled in the art upon reading the foregoing description.

Claims (23)

1. Apparatus for analyzing a blood sample stained with one or more dyes, the apparatus comprising: a microscope system (11) configured to acquire a plurality of microscope images of the stained blood sample; and at least one computer processor (28) configured to: identify, in the microscope images, at least one element as being a candidate of a given entity, extract, from the microscope images, at least one associated candidate informative feature to the candidate of the given entity, characterized by the fact that it further comprises: determining dyeing quality parameters for the respective microscope images, the dyeing quality parameters being indicative of a dyeing quality in each of the respective microscope images, and process the candidate informative feature in combination with the dyeing quality parameters of the respective microscope images. 2. Apparatus according to claim 1, characterized by the fact that at least one computer processor (28) is configured to discard at least some of the images from being used in an analysis of the blood sample in response to the determination that the tinting quality of at least some images is not sufficient. 3. Apparatus according to claim 1, characterized by the fact that the at least one computer processor (28) is configured to adjust one or more thresholds that are used to identify entities within respective microscope images, based on the dyeing quality parameters from respective microscope images. 4. Apparatus according to claim 1, characterized by the fact that the at least one computer processor (28) is configured to modulate a number of times that the respective microscope image fields of the samples are imaged by the microscope, based on the dyeing quality parameters of the respective microscope images. 5. Apparatus according to claim 1, characterized by the fact that the at least one computer processor (28) is configured to classify candidates that are identified in respective microscope images based, at least partially, on quality parameters staining of respective microscope images. 6. Apparatus according to claim 5, characterized by the fact that the at least one computer processor (28) is configured to classify candidates that are identified in the respective microscope images by normalizing the candidate classification in the respective microscope images based on the dyeing quality parameters of the respective microscope images. 7. Apparatus according to claim 1, characterized by the fact that the at least one computer processor (28) is configured to modulate an image capture parameter of the microscope images based on the dyeing quality parameters of the respective microscope images. 8. Apparatus according to claim 7, characterized by the fact that the at least one computer processor (28) is configured to modulate the image capture parameter of the microscope images by modulating a frame rate at which the Microscope images of the sample are acquired, based on the dyeing quality parameters of the respective microscope images. 9. Apparatus according to claim 8, characterized by the fact that the at least one computer processor (28) is configured to modulate the frame rate at which microscope images of the sample are acquired based on quality parameters of tinting the respective microscope images by obtaining microscope images at a higher frame rate in response to detecting that the tinting quality of at least some of the microscope images is relatively low, and acquiring microscope images at a low rate of frames in response to detecting that the tinting quality of at least some of the microscope images is relatively high. 10. Apparatus according to claim 1, characterized by the fact that the at least one computer processor (28) is configured to acquire the plurality of microscope images of the stained blood sample by acquiring a plurality of microscope images of the respective fields image of the stained blood sample over a period of time such that there is variation in the staining quality of the microscope images of the respective image fields. 11. Apparatus according to claim 10, characterized by the fact that the at least one computer processor (28) is configured to acquire the plurality of microscope images of respective image fields of the stained blood sample over the period of time that is such that there is variation in staining quality in the microscope images of the respective imaging fields acquiring a plurality of fluoroscopy microscope images of the respective imaging fields of the stained blood sample over a period of time that is such that there is variation in the quality of dyeing in microscope images of the respective image fields due to photobleaching. 12. Apparatus according to claim 10, characterized by the fact that the at least one computer processor (28) is configured to acquire the plurality of microscope images of respective image fields of the stained blood sample over the period of time that is such that there is variation in the quality of staining in the microscope images of the respective image fields acquiring a plurality of microscope images of the respective image fields of the stained blood sample over a period of time that is such that there is variation in the quality of dye in microscope images of the respective image fields due to at least one of the one or more dyes diffusing through the blood sample away from the targets. 13. Method comprising: staining a blood sample with one or more dyes; acquiring one or more microscope images of the stained blood sample using a microscope; with the use of at least one computer processor (28): in the one or more microscope images, identify at least one element as being a candidate of a given entity; extracting, from the microscope images, at least one candidate information feature associated with the candidate of the given entity; characterized by the fact that it comprises: determining dyeing quality parameters for the respective microscope images, the dyeing quality parameters being indicative of a dyeing quality in each of the respective microscope images; and processing the candidate informative feature in combination with the dyeing quality parameters of the respective microscope images. 14. Apparatus for analyzing a blood sample, the apparatus comprising: a microscope system (11) configured to acquire one or more microscope images of the blood sample; an output device (34); and at least one computer processor (28) configured to: identify, in the microscope images, at least one element as being a candidate of a given entity, extract, from the microscope images, at least one associated candidate informative feature to the candidate of the given entity, characterized by the fact that it comprises identifying deformed red blood cells in microscope images, determining a degree of deformation of the red blood cells in the sample, at least partially based on the identified deformed red blood cells, determining an informative parameter of the sample that is indicative of a characteristic of the blood sample as a whole, at least partially based on the degree of deformation of the red blood cells in the sample, process the candidate information feature in combination with the sample information feature, and generate an output in response thereto. 15. Apparatus according to claim 14, characterized by the fact that the at least one computer processor (28) is configured to identify deformed red blood cells in microscope images by identifying echinocytes and/or acanthocytes in microscope images. 16. Apparatus according to claim 14, characterized by the fact that the at least one computer processor (28) is configured to determine the informative parameter of the sample based, at least partially, on the degree of deformation of the red blood cells in the sample by determining that the sample has pathology based, at least partially, on the degree of deformation of the red blood cells in the sample. 17. Apparatus according to claim 14, characterized by the fact that the at least one computer processor (28) is configured to determine the informative parameter of the sample based, at least partially, on the degree of deformation of the red blood cells in the sample by determining that the sample is infected with a pathogen based, at least partially, on the degree of deformation of the red blood cells in the sample. 18. Apparatus according to claim 14, characterized by the fact that the at least one computer processor (28) is configured to detect that the sample is infected with a pathogen, and wherein the at least one computer processor (28) is configured to detect that the sample is infected with a pathogen, and wherein the at least one computer processor (28) 28) is configured to determine the informative parameter of the sample based, at least partially, on the degree of deformation of the red blood cells in the sample. 19. Apparatus according to claim 14, characterized by the fact that the at least one computer processor (28) is configured to determine the informative parameter of the sample based, at least partially, on the degree of deformation of the red blood cells in the sample by determining that the sample has undergone prolonged exposure to a solution based, at least in part, on the degree of deformation of the red blood cells in the sample. 20. Apparatus according to claim 19, characterized by the fact that the at least one computer processor (28) is configured to determine that the sample has undergone prolonged exposure to a solution based, at least partially, on the degree of deformation of the red blood cells in the sample by determining that the sample has undergone prolonged exposure to a dye based, at least in part, on the degree of deformation of the red blood cells in the sample. 21. Apparatus according to claim 14, characterized by the fact that the at least one computer processor (28) is configured to determine the informative parameter of the sample based, at least partially, on the degree of deformation of the red blood cells in the sample by determining that a parameter relates to the age of the sample based, at least partially, on the degree of deformation of the red blood cells in the sample. 22. Apparatus according to claim 21, characterized by the fact that the at least one computer processor (28) is further configured to detect that the sample is infected with a pathogen, and wherein the at least one computer processor (28) is configured to determine the parameter related to the age of the sample based, at least partially, on the degree of deformation of the RBCs in the sample by estimating the age of the pathogen based, at least partially, on the degree of deformation of the RBCs in the sample . 23. Method comprising: acquiring one or more microscope images of a blood sample; with the use of at least one computer processor (28): in the microscope images, identify at least one element as being a candidate for a given entity; extracting, from the microscope images, at least one candidate information feature associated with the candidate of the given entity; characterized by the fact that it includes identifying deformed red blood cells in microscope images; determining a degree of deformation of the red blood cells in the sample, at least partially based on the identified deformed red blood cells; determining an informative sample parameter that is indicative of a characteristic of the blood sample as a whole, at least partially based on the degree of deformation of the red blood cells in the sample; process the candidate information resource in combination with the sample information resource, generate an output in response thereto.