WO2024030346A1 - Ai-based cell classification method and system - Google Patents

Ai-based cell classification method and system Download PDF

Info

Publication number
WO2024030346A1
WO2024030346A1 PCT/US2023/029019 US2023029019W WO2024030346A1 WO 2024030346 A1 WO2024030346 A1 WO 2024030346A1 US 2023029019 W US2023029019 W US 2023029019W WO 2024030346 A1 WO2024030346 A1 WO 2024030346A1
Authority
WO
WIPO (PCT)
Prior art keywords
cell
cells
features
bec
abnormal
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/US2023/029019
Other languages
English (en)
French (fr)
Inventor
Michael G. Meyer
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
VisionGate Inc
Original Assignee
VisionGate Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by VisionGate Inc filed Critical VisionGate Inc
Priority to AU2023317489A priority Critical patent/AU2023317489B2/en
Priority to CA3263775A priority patent/CA3263775A1/en
Priority to IL318743A priority patent/IL318743A/en
Priority to CN202380067171.3A priority patent/CN119866434A/zh
Priority to EP23850630.7A priority patent/EP4565868A4/en
Priority to JP2025505966A priority patent/JP2025525906A/ja
Publication of WO2024030346A1 publication Critical patent/WO2024030346A1/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
    • G06V20/698—Matching; Classification
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
    • G01N15/10—Investigating individual particles
    • G01N15/14—Optical investigation techniques, e.g. flow cytometry
    • G01N15/1429—Signal processing
    • G01N15/1433—Signal processing using image recognition
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/20—Ensemble learning
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three-dimensional [3D] modelling for computer graphics
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
    • G06V20/695—Preprocessing, e.g. image segmentation
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
    • G01N15/10—Investigating individual particles
    • G01N2015/1006—Investigating individual particles for cytology
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
    • G01N15/10—Investigating individual particles
    • G01N15/14—Optical investigation techniques, e.g. flow cytometry
    • G01N15/1434—Optical arrangements
    • G01N2015/144—Imaging characterised by its optical setup
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
    • G01N15/10—Investigating individual particles
    • G01N15/14—Optical investigation techniques, e.g. flow cytometry
    • G01N15/1434—Optical arrangements
    • G01N2015/144—Imaging characterised by its optical setup
    • G01N2015/1445—Three-dimensional imaging, imaging in different image planes, e.g. under different angles or at different depths, e.g. by a relative motion of sample and detector, for instance by tomography
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
    • G01N15/10—Investigating individual particles
    • G01N15/14—Optical investigation techniques, e.g. flow cytometry
    • G01N2015/1497—Particle shape
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G06T2207/10101—Optical tomography; Optical coherence tomography [OCT]
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30024—Cell structures in vitro; Tissue sections in vitro
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30096—Tumor; Lesion

Definitions

  • the present disclosure relates to a system and method for artificial-intelligence (Al)-based 2D classification of cells to prior to 3D analysis to detect cells indicative of cancer or with features of a normal bronchial epithelial cell (BEC), as well as a method of training a 2D cell classifier for use in an Al-based cell classification system and method.
  • Al artificial-intelligence
  • Lung cancer is the most lethal cancer in the United States and over 31 million patients are at a high risk of developing lung cancer. Early detection is the most reliable means of reducing lung cancer deaths, but many detection methods have poor sensitivity and specificity, leading to missed diagnoses and, as a result, higher death rates, as well as increased costs and needless suffering caused by invasive procedures.
  • the present disclosure provides an artificial intelligence (Al)-based cell classification method comprising: a) generating, by an optical tomography system, a representative 2D image of a cell from a patient sample comprising a plurality of cells; b) evaluating the representative 2D image using an abnormal cell 2D classifier to determine if the cell has abnormal features; c) evaluating the representative 2D image using a BEC 2D classifier to determine if the cell has BEC -like features; and d) generating, by the optical tomography system, a 3D image of the cell if the cell is determined to have abnormal features or to have BEC-like features.
  • the method may further include one or more of the following additional features, in any combinations: a 3D image of the cell is not generated if the cell is determined to not have abnormal features and to not have BEC-like features; e) repeating a) to d) for a subset of cells within the plurality of cells to generate patient sample data that reflects a total number of enumerated analyzed cells having abnormal features, a total number of enumerated analyzed cells having BEC-like features, or a total number of enumerated analyzed cells; f) comparing the total number of enumerated analyzed cells having BEC-like features to a threshold value to determine if the total number of enumerated analyzed cells having BEC-like feature is adequate for an accurate assay because the total number is at or above the threshold value, or if the total number of enumerated analyzed cells having BEC-like features is inadequate for an accurate assay because the total number is below the threshold value; repeating a) to f)
  • the present disclosure also provides a method of training an Al-based cell classification system comprising: a) operating an optical tomography system to generate representative 2D images of cells; b) designating a known identifier for each of a plurality of the cells using the representative 2D image of such cell and a known cell identification method, wherein the known identifier is i) abnormal or having abnormal features; ii) BEC or having BEC-like features, or iii) normal or non-BEC or not having abnormal features or BEC-like features; c) computing a plurality of cell feature measurements for each of the plurality of cells from the representative 2D image of such cell; d) performing a regression between the cell feature measurements and the known identifier of each cell to designate an abnormal 2D cell classification score and a BEC 2D cell classification score for each cell; e) designating a test identifier for each cell of i) having abnormal features; ii) having BEC-like features, or iii) not having abnormal features or
  • the method may further include one or more of the following additional features, in any combinations: comparing the accuracy of both of the 2D cell classifiers to pre-selected accuracy requirements and, if the pre-selected accuracy requirements are met by a 2D cell classifier, designating the 2D cell classifier as trained; comparing the accuracy of both of the 2D cell classifiers to pre-selected accuracy requirements and, if the pre-selected accuracy requirements are not met by a 2D cell classifier, adjusting a parameter of the regression for the affected 2D cell classifier and repeating steps d)- f); the pre-selected accuracy requirement is a 2D abnormal cell sensitivity value of at least 90%; the pre-selected accuracy requirement is a 2D abnormal cell specificity value of at least 65%; the pre-selected accuracy requirement is a 2D BEC sensitivity value of at least 85%; the pre-selected accuracy requirement is a 2D BEC specificity value for of at least 60%; the regression comprises Adaptively Boosted Logistic Regression, Random Forest
  • training method may be used to train the Al and described above or otherwise herein.
  • the disclosure further provides a cell classification system comprising: an optical tomography system operable to: generate a representative 2D image of a cell from a patient sample; and generate a 3D image of the cell if the cell has abnormal features or BEC-like features; and a processor operable to: compare the representative 2D image of the cell to an abnormal cell 2D classifier to determine if the cell has abnormal features; and compare the representative 2D image of the cell to a BEC 2D classifier to determine if the cell has BEC-like features.
  • an optical tomography system operable to: generate a representative 2D image of a cell from a patient sample; and generate a 3D image of the cell if the cell has abnormal features or BEC-like features
  • a processor operable to: compare the representative 2D image of the cell to an abnormal cell 2D classifier to determine if the cell has abnormal features; and compare the representative 2D image of the cell to a BEC 2D classifier to determine if the cell has BEC-like features.
  • the system may be used in either the cell classification method or the training method described above or otherwise herein, or in both.
  • Fig- 1 is a schematic representation of an optical tomography system that may be used in the present disclosure
  • Fig- 2 is a schematic representation of the optical tomography system as operated to acquire a plurality of 2D images of a cell that may be used in the present disclosure
  • Fig- 3 is a flow chart for a lung cancer detection method of the present disclosure
  • Fig. 4 is a flow chart of an Al-based cell classification method of the present disclosure for use in the lung cancer detection method of Fig. 3;
  • Fig. 5 is a flow chart of a 2D classifier training method of the present disclosure
  • Fig. 6 is an example 2D section of a 3D image of cells that may be obtained using an optical tomography system of the present disclosure
  • Fig- 7 is a Cumulative Probability Density Curve (CDF) plot, for example abnormal cell 2D classifier results according to the present disclosure, in which the green dashed line indicates the threshold for classifier operation, the red line shows the trend for cells classified as having abnormal features (“abnormal”), the green line shows the trend for cells classified as having BEC-like features (“epithelial”), the light blue line shows the trend for miscellaneous cells (“rest”), and the dark blue line shows the trend for cells having BEC-like features and miscellaneous cells (“combined”);
  • CDF Cumulative Probability Density Curve
  • Fig- 8 is a CDF plot, for example BEC 2D classifier results according to the present disclosure, in which the green dashed line indicates the threshold for classifier operation, the green line shows the trend for cells classified as having BEC-like features, and the light blue line shows the trend for miscellaneous cells (normal).
  • the present disclosure relates to an Al-based cell classification method and system, which may be used to detect cells indicative of cancer among a plurality of cells in a lung-related patient sample, such as sputum.
  • the cells indicative of cancer may be cancerous, but they may also include non-cancerous cells having abnormal features.
  • a sample may be designated as containing cells indicative of cancer (and thus positive for lung cancer) based solely on the detection of non-cancerous cells having abnormal features.
  • the patient may be referred for further testing.
  • the system and method use optical tomography, such as current optical tomography systems, to detect cells and generate 3D images of the cells.
  • optical tomography such as current optical tomography systems
  • the system and method of the present disclosure prior to generating a 3D image of a cell, first generate a representative 2D image of the cell and analyze the representative 2D image using Al-based 2D cell classifiers to determine if a 3D image should be generated.
  • Patient samples contain a substantial number of diverse cells that are not likely indicative of lung cancer and, therefore, are not useful in detecting lung cancer.
  • a 3D image of such cells is not needed, and generating it is a waste of resources.
  • the disclosure also provides a method of training the Al of a 2D cell classifier for use in an Al-based cell classification system or method using cells having known identifiers.
  • the lung cancer may be non-small cell lung cancer (NSCLC), particularly squamous carcinoma and adenocarcinoma.
  • NSCLC non-small cell lung cancer
  • SCLC small cell lung cancer
  • cancer refers to a hyperproliferation of cells that results in unregulated growth, lack of differentiation, local tissue invasion, or metastasis.
  • cells having abnormal features may include cells exhibiting atypia, dysplastic cells, pre-cancerous cells, pleomorphic parakeratosis, type II pneumocyte abnormal squamous cells, adenocarcinoma cells, bronchioloalveolar carcinoma cells, abnormal neuroendocrine cells, small cell carcinoma cells, non-small cell carcinoma cells, tumor cells, neoplastic cells, bronchioloalveolar carcinoma cells, and any combination thereof.
  • cells having BEC-like features may include cells that typically line the airway lumen in the lungs of a patient.
  • the optical tomography system may be a CELL-CT® system (VisionGate, Inc., Washington, USA).
  • an Al-based cell classification system of the present disclosure may include, or an Al-based cell classification method of the present disclosure may be carried out using an optical tomography system 100, which may be used to produce both 3D images and 2D images of a cell 10 (such as cell 10a or cell 10b).
  • an optical tomography system 100 which may be used to produce both 3D images and 2D images of a cell 10 (such as cell 10a or cell 10b).
  • the operation of the optical tomography system is described for acquiring images of one cell 10, in a volume of optical medium in the optical path of a high-magnification microscope, images of multiple cells 10 within the same volume of optical medium may be acquired.
  • the optical tomography system 100 may include a cell imaging system 110, which includes an illumination source 120 optically coupled to an objective lens 130, such that illumination passes through the micro-capillary tube 30 and any intervening cell 10 before reaching the objective lens 130.
  • the illumination then passes through the objective lens 130 to a beam-splitter 140, which causes part of the illumination to be deflected to a mirror 150 and reflected back to the beam-splitter 140 before being transmitted to a high-speed camera 160, and another part of the illumination to be transmitted directly through the beam-splitter 140 to the high-speed camera 160, to generate pseudo-projection images 40 of the cell 10 contained in an optical medium 20 in a micro-capillary tube 30.
  • At least one pseudoprojection image 40 of the cell 10 is generated by scanning the volume occupied by the cell 10 by vibrating the mirror 150 in direction 60 (typically using an actuator, such as a piezo-electric motor, not shown), thus sweeping the plane of focus 50 through the cell 10 and then integrating the image to create the pseudo-projection image from a single perspective. Additional pseudoprojection images are obtained by rotating the micro-capillary tube 30.
  • the pseudo-projection images are each a single image that represents a sampled volume that has an extent greater than the depth of field of the objective lens 130.
  • the high-speed camera 160 generates, for each cell 10, a plurality of pseudo-projection images 40 that correspond to a plurality of axial microcapillary tube rotation positions, examples of which are illustrated as 40a, 40b, and 40c in Fig. 1.
  • 500 pseudoprojection images are generated as the micro-capillary tube 30 is rotated through 360°.
  • optical tomography system 100 is communicatively coupled to a processor 170 operable to receive the plurality of pseudo-projection images 40 from the high-speed camera 160 and use the pseudo-projection images 40 to generate a 3D image (not shown) of the cell 10. Images before or after manipulation by the processor 170 may be stored in communicatively coupled memory 180. The processor 170 may send data regarding the cell 10, or data relating to or derived from a plurality of cells 10 contained in a patient sample to a communicatively coupled output 190.
  • Data sent to the output 190 may include 2D or 3D images of one or more cells 10, or a summary of cells in a patient sample analyzed by the optical tomography system 100, such as a graph or a table, an index, such as an abnormality index, reflecting the likelihood that the patient has or is at high risk for developing lung cancer, or even a simple indication that the sample is positive for cells indicative of lung cancer.
  • the optical tomography system 100 may also be operated to generate a representative 2D image of the cell 10.
  • the objective lens 130 has a focal plane 50 that moves as the objective lens 130 sweeps across the micro-capillary tube 30 and any cell 10 in a back-and-forth direction 60 to produce a plurality of 2D images (not shown).
  • This method of generating 2D images by moving the objective lens 130 is different than the method of poroducing 3D images using pseudo-projection images that are generated by vibrating the mirror 150.
  • the processor 170 is operable to receive the plurality of 2D images and determine a representative 2D image of the cell 10.
  • the representative 2D image of the cell 10 is an image of the central portion, such as the center) of the cell.
  • the processor 170 is then further operable to perform Al-based cell classification of the representative 2D image using 2D cell classifiers as described herein to determine if the cell 10 has abnormal features or BEC-like features, or does not have abnormal features or BEC-like features. Only when the cell 10 is determined to have abnormal features or to have BEC-like features does the processor 170 direct the cell imaging system 110 to generate pseudo-projection images 40 of the cell 10.
  • a substantial number of cells 10 in a patient sputum sample or other samples are not likely to be indicative of lung cancer, such that producing a 3D image of these cells is not useful in detecting lung cancer. Many of these cells do not have the morphology of normal or abnormal cells of the bronchial epithelium so they may be eliminated using only a 2D image. It is difficult to remove such cells from the patient sample prior to Al-based cell classification, so in prior cell classification methods these are simply imaged along with cells that are likely to be indicative of lung cancer.
  • the capability of using 2D images to first assess whether a cell is likely to be indicative of lung cancer before capturing pseudo-projection images 40 allows the Al-based cell classification to focus on cells 10 that provide meaningful information while not wasting processing and analysis time on cells that are not useful in detecting lung cancer.
  • the cell imaging system 110 includes the illumination source 120, the objective lens 130, the beam-splitter 140, mirror 150, and the high-speed camera 160.
  • the optical tomography system 100 further includes the processor 170, any communicatively coupled memory 180, and the communicatively coupled output 190.
  • the optical tomography system 100 further includes the micro-capillary tube 30, the optical medium 20, or one or more cells 10, but in other embodiments, the optical tomography system 100 does not include one or more of these potential components, although they may be supplied for operation of the system.
  • the Al-based cell classification system 100 is operable to generate images in at least two distinct modes.
  • One mode includes a cell search mode, during which 2D images of a cell 10 are generated and processed to generate a representative 2D image, which is then used to determine whether the cell 10 has abnormal features or BEC-like features. 2D images in this mode are generated by moving the objective lens 130 in direction 60 to sweep the focal plane 50 through the cell.
  • a second mode includes a projection image capture mode, during which pseudo-projection images 40 are generated and used to produce a 3D image of the cell 10 if the cell 10 has been determined to have abnormal features or BEC-like features during the cell search mode. 3D images in this mode are generated by vibrating the mirror 150 in direction 60 to create the pseudo-projection images 40.
  • Fig. 3 and Fig. 4 describe a lung cancer detection method 200, which may be performed using an optical tomography system, such as the optical tomography system 100. Elements of the optical tomography system 100 are referenced in this description of the lung cancer detection method 200 as examples. Similar components of different optical tomography systems may also be used in connection with the lung cancer detection method 200.
  • the lung cancer detection method 200 includes a step 210 in which a lung cell sample is collected from a patient.
  • the lung cell sample may be sputum, although other sample types, such as samples obtained by bronchoalveolar lavage (BAL), nasal swab, or biopsy may also be used.
  • samples may include urine or blood.
  • Samples may undergo an Al-based cell classification method 300 at any time during which cells are not to have degraded to the point where limited cellular content remains. The duration of such time may depend on the storage conditions, e.g., whether the sample is refrigerated or how or if the sample is processed. In some embodiments, the sample may undergo the Al-based cell classification method 300 within 30 minutes, one hour, two hours, 6 hours, 12 hours, one day, two days, one week, or two weeks of collection.
  • the patient sample may include many cell types that are not likely to be indicative of lung cancer, such as white blood cells, including polymorphonuclear leukocytes, monocytes, and lymphocytes, cell clusters, oral squamous cells, and squamous intermediate cells (pictured in Fig. 6, upper panel).
  • the sample also typically will contain a substantial amount of non-cellular debris or cell fragments.
  • the sample contains abnormal cells, BECs, squamous cells, monocytes, lymphocytes, polymorphonuclear leukocytes, other white blood cells, debris, cell fragments, cell clusters and any combinations thereof.
  • step 220 the sample is processed for analysis. Processing may optionally include staining to render any of the plurality of cells 10 or any features, such as the nucleus of any of the plurality of cells 10, easier to detect using the optical tomography system 100, or otherwise staining or treating the cells with an agent that facilitates generating the representative 2D image, evaluating the representative 2D image using an abnormal cell 2D classifier, evaluating the representative 2D image using a BEC 2D classifier, generating a 3D image of the cell, or analyzing a 3D image of the cell.
  • the cells may be stained with hematoxylin.
  • Processing may optionally include, alone or in combination with staining, enrichment of the sample for cells of interest.
  • the sample is enriched for BECs.
  • the sample may be enriched for BECs by staining cytoskeleton proteins and using fluorescently activated cell sorting (FACS)-based enrichment.
  • FACS fluorescently activated cell sorting
  • BEC enrichment may include treating the sample with at least one antibody, and typically a plurality of antibodies, having fluorescent conjugates that may be used for FACS-based enrichment.
  • the antibodies may bind to BECs or they may bind contaminating inflammatory cells, such a neutrophils and macrophages.
  • Antibodies that bind contaminating inflammatory cells may include anti-CD45 antibodies.
  • the sample is treated with a combination of antibodies that bind to BECs and antibodies that bind contaminating inflammatory cells, with the antibodies having distinct fluorescent conjugates.
  • the cells may be stained with 4',6-diamidino-2-phenylindole (DAPI) alone or in combination with antibodies for FACS-based cell enrichment purposes as well.
  • DAPI 4',6-diamidino-2-phenylindole
  • Cells may be enriched by FACS in which gating is used to exclude DAPI-positive materials (which tend to be doublet cells or debris), high side-scatter objects, objects bound by anti-inflammatory cell antibodies, or any combinations thereof. Cells may also be enriched by FACS in which gating is used to select cells that are bound by anti-BEC antibodies. In some embodiments, both exclusive and inclusive gating may be used and may be implemented concurrently or in series. [0041] Following any optional staining or enrichment, the sample processing 220 includes placing the cells 10 contained in the sample in an optical medium 20.
  • the optical medium 20 may be any medium reasonably expected to maintain the cells 10 intact during the expected duration of time prior to and during the Al-based cell classification 300.
  • the optical medium 20 may also have a viscosity that allows movement of the optical medium 20 through a micro-capillary tube 30.
  • the optical medium 20 may also not interfere with image generation by the optical tomography system 100.
  • the optical medium may have an optical index that matches the optical index of other components of the optical tomography system 100 and the micro-capillary tube 30 through which light passes during image acquisition.
  • the optical tomography system 100 components through which light passes and the micro-capillary tube 30 also have a matching index. For optimal optical tomography operation, any changes in light movement should be due to encountering the object to be imaged, not changes in the optical index of other components or objects in the light path.
  • the optical medium 20 containing a plurality of cells 10 from the patient sample is injected into a micro-capillary tube 30.
  • the microcapillary tube 30 may have an outer diameter of 500 pm or less, for example, between 30 pm and 500 pm.
  • the micro-capillary tube 30 may have an inner diameter of 400 pm or less, for example, between 30 pm and 400 pM, such as 50 pm.
  • the entire portion of the patient sample to be analyzed is placed in one micro-capillary tube 30.
  • the portion of the patient sample to be analyzed is placed in a plurality of micro-capillary tubes 30, which may undergo the Al-based cell classification method 300 sequentially.
  • the sample may be pumped through the microcapillary tube 30 from a sample reservoir.
  • the micro-capillary tube 30 is loaded into the optical tomography system 100, so that the micro-capillary tube 30 is between the illumination source 120, and the objective lens 130.
  • step 250 the optical medium 20 and any cells 10 contained within it are advanced into (prior to the initial step 240) or through the micro-capillary tube 30 by applying pressure at one end of the micro-capillary tube, such that a different volume of the optical medium 20 carrying a different portion of the patient sample is in the optical path of the optical tomography system 100, between the illumination source 120 and the objective lens 130.
  • a plunger (not shown) is used to advance the optical medium 20.
  • a plunger may be applied to a reservoir (not shown) of the optical medium 20 and patient sample that is connected to the micro-capillary tube 30, forcing additional optical medium 20 from the reservoir into the micro-capillary tube 30
  • the method proceeds to the Al-based cell classification 300, which includes a cell search mode 310 and a projection image capture mode 320, both of which are described in further detail in Fig. 4.
  • the optical tomography system 100 In the cell search mode 310, the optical tomography system 100 generates 2D images of a cell 10 that are further processed to generate a representative 2D image that is used to determine whether the cell 10 has abnormal features or BEC-like features. In some embodiments, an algorithm selects an image of a central portion, such as the center, of the cell 10 as the representative 2D image. In some embodiments, in the cell search mode 310, the optical tomography system 100 sweeps the focal plane 50 through the cell 10 in the direction 60 at 1 pm intervals to capture a series of 2D images. Multiple cells 10 may be identified in the same volume of optical medium 20 in the optical path of the optical tomography system 100.
  • each of cells 10 identified as having abnormal features or having BEC-like features is further imaged to generate a plurality of pseudo-projection images 40, which are used to generate a 3D image of the cell 10.
  • step 260 it is determined whether a pre-selected number of cells having BEC- like features have been 3D imaged. If the pre-selected number has been reached, the process moves on to step 270. Otherwise, the process returns to step 250 and advances the sample in the micro-capillary tube.
  • results of cells selected for 3D imaging are provided to a user, for example, using the output 190.
  • Results may include results specific for an individual cell 10, such as the 3D image of cell 10.
  • Al-based cell classification results may also include data based upon the analysis of all or a portion of the patient sample or plurality of cells 10, such as an abnormality index and/or a cancer threshold value as explained further in Fig. 4.
  • the Al-based cell classification results may include a list or other indication of individual cells 10 found to have abnormal features or BEC-like features or otherwise marked according to preselected criteria for further review of the 3D image by the user.
  • the Al-based cell classification results may include graphical, statistical, or numerical information, such as the number of cells detected (typically the total enumerated analyzed cells), the number of cells for which a 3D image was generated, the number of enumerated cells having abnormal features, the number of enumerated cell having BEC-like features, or the relative proportions of any of these groups of cells.
  • the Al-based cell classification results may also contain information regarding the predicted accuracy.
  • the Al-based cell classification results may include an index, such as an abnormality index, reflecting the likelihood that the patient has or is at high risk for developing lung cancer, or even a simple indication that the sample is positive for cells indicative of lung cancer.
  • Sample identification data may also be provided with the Al-based cell classification results.
  • the Al-based cell classification method 300 may include a cell search mode 310 and a projection image capture mode 320.
  • the cell search mode 310 may include step 400, in which the optical tomography system 100 generates a plurality of single-focal plane 2D images of a cross-sectional region of the micro-capillary tube 30, each image at a different focal plane 50 by sweeping the focal plane 50 across the micro-capillary tube (and any cell in any of the focal panes 50) back-and-forth in direction 60 by moving the objective lens 130.
  • the single-focal plane 2D images may be taken in 1 pm increments in direction 60.
  • a representative 2D image of a cell 10 is generated. To do so, at least a portion of the plurality of single-focal plane 2D images are compiled and filtered by the processor 170 to determine if the 2D image contains features associated with a cell 10 or other solid object in the optical medium 20, such as being dark as compared to the optical medium 20.
  • the processor 170 determines which should be designated as a representative 2D image based on pre-selected criteria. In one embodiment, the processor determines which 2D image represents a central portion, such as the center, of the cell 10. This image representing the central portion of the cell 10 is designated as the representative 2D image of the cell 10.
  • step 420 the method determines if another cell 10 is present in the volume of optical medium and proceeds to conduct step 410 (and, optionally, step 400 if further 2D images are needed) for the other cell. This may be repeated for all, a set number of, or a set proportion of putative cells 10 in the volume of optical medium 20 in the optical path of the optical tomography system 100.
  • the processor 170 evaluates the representative 2D image using an abnormal cell 2D classifier to determine if the cell 10 has abnormal features.
  • the processor may use an Al-trained or implemented algorithm in step 420.
  • the abnormal cell 2D classifier may be trained as set forth in the training method 500 herein. Specifically the abnormal cell 2D classifier may identify a plurality of features of the cell 10 and make cell feature measurements, then compare at least a subset of such cell feature measurements to trained values for cell feature measurements associated with cells having abnormal features and, optionally, also cells not having abnormal features. The cell may be identified as having abnormal features or not having abnormal features based upon the comparison. To facilitate identifying and measuring the plurality of cell features, the abnormal cell 2D classifier may segment the representative 2D image into a nucleus portion and a non-nucleus cellular portion. The cell feature measurements may include object shape features, cell-shape features, cytoplasm features, cell nucleoli features, distribution of chromatin, nuclear-size features, nuclear-texture features, other morphometric elements, or any combination thereof.
  • the abnormal cell 2D classifier may have a pre-selected abnormal cell 2D accuracy, typically a pre-selected abnormal cell 2D sensitivity or a pre-selected abnormal cell 2D specificity,
  • the method may progress directly to the projection image capture mode 320 as shown. However, in some embodiments, even if the cell 10 is determined in 2D to have abnormal features, the method may still progress to step 430 (not shown) to also determine if the cell 10 has BEC-like features in 2D.
  • step 430 if the cell 10 is determined to not have abnormal features, then the method progresses to step 430.
  • the processor 170 evaluates the representative 2D image using a BEC 2D classifier to determine if the cell 10 has BEC-like features.
  • the processor may use an Al-trained or implemented algorithm in step 430.
  • the BEC 2D classifier may be trained as set forth in the training method 500 herein. Specifically, the BEC 2D classifier may identify a plurality of features of the cell 10 and make cell feature measurements, then compare at least a set of the identified and measured cell features measurements to trained values for cell feature measurement associated with cells having BEC-like features and, optionally, also cells not having BEC-like features. The cell may be identified as having BEC-like features or not having BEC-like features based upon the comparison.
  • the cell features and cell feature measurements may be selected from the same group set forth above for the abnormal cell 2D classifier.
  • cell features identification and measurement may be conducted separately for use with the abnormal cell 2D classifier and the BEC 2D classifier.
  • the cell may be designated having BEC-like features or not having BEC-like features based upon the comparison.
  • the BEC 2D classifier may have a pre-selected 2D BEC accuracy, typically a preselected 2D BEC sensitivity or 2D BEC specificity.
  • the abnormal cell 2D classifier and the BEC 2D classifier may have a pre-selected abnormal cell referral rate or a pre-selected BEC referral rate.
  • the method progresses directly to the projection image capture mode 320 as shown. If the cell 10 is determined to not have BEC-like features, then no 3D image is captured and the method progresses to step 250. In some embodiments (not shown), rather than proceeding directly to step 250, the method determines if another cell 10 is present in the volume of optical medium and proceeds to conduct steps conduct steps 410-430 (and, optionally, step 400 if further 2D images are needed) for the other cell. This may be repeated for all, a set number of, or a set proportion of putative cells 10 in the volume of optical medium 20 in the optical path of the optical tomography system 100.
  • the optical tomography system 100 in step 440, the optical tomography system 100 generates a plurality of pseudo-projection images 40 by vibrating the mirror 150. These images may be taken over 360 degrees of rotation around the cell 10. However, pseudoprojection 40 may be taken over as little as 180 degrees of rotation around the cell 10.
  • the processor 170 uses at least a portion of the plurality of pseudo-projection images 40 to generate a 3D image of the cell 10.
  • all pseudoprojection images 40 are used to generate the 3D image of the cell 10.
  • pseudoprojection images 40 that are determined to be of poor quality or likely to contain errors the 3D imaging of the cell may be discontinued.
  • typically pseudo-projection images 40 that cover 360 degrees of rotation around the cell 10 are used to generate the 3D image, but pseudoprojection images that cover as little as 180 degrees of rotation around the cell 10 may be used.
  • step 460 the processor 170 uses a 3D classifier to determine if the cell has abnormal features or BEC-like features. This information is used to determine if the cell is likely indicative of lung cancer.
  • the 3D classifier may also have pre-selected accuracy values.
  • the process then returns to step 260 to determine if a pre-selected number of BECs have been 3D imaged. If the pre-selected number has been reached, then the process moves on to step 270, which, in some embodiments, may include steps 470 and 480.
  • step 470 data relating to the cell is stored as patient sample data.
  • Patient sample data may be used in determining an abnormality index.
  • Patient sample data may also reflect a total number of enumerated analyzed cells having abnormal features, a total number of enumerated analyzed cells having BEC-like features, or a total number of enumerated analyzed cells and steps 410- 470 and, optionally, step 400 if additional 2D images are needed, may be repeated for additional cells until a pre-selected threshold value of any or all of these total numbers is reached.
  • step 480 the processor compares the total number of enumerated analyzed cells having abnormal features with the total number of enumerated analyzed cells having BEC- like features or the total number of enumerated analyzed cells to generate an abnormality index.
  • the abnormality index may be correlated with the proportions of any of these totals.
  • step 490 the processor compares that abnormality index to a cancer threshold value set to correspond with the patient sample being designated positive for cells indicative of lung cancer.
  • the overall method 300 which includes 2D cell classification and 3D cell classification, may have a pre-selected assay accuracy, such as a pre-selected assay sensitivity or a pre-selected assay specificity, cell classification method 300 as described herein may be substantially quicker than an analogous method that does not use 2D cell classifiers and, instead, generates a 3D image of every cell or putative cell identified.
  • a pre-selected assay accuracy such as a pre-selected assay sensitivity or a pre-selected assay specificity
  • cell classification method 300 as described herein may be substantially quicker than an analogous method that does not use 2D cell classifiers and, instead, generates a 3D image of every cell or putative cell identified.
  • the method 300 may reduce cell classification time, such as for steps 400-460, by at least 25%, at least 35%, at least 40%, at least 45%, at least 50%, or in a range of 25% to 60%, 25% to 50%, 25% to 45%, 25% to 40%, 25% to 35%, 35% to 60%, 35% to 50%, 35% to 45%, 35% to 40%, 40% to 60%, 40% to 50%, 40% to 45%, 45% to 60%, or 45% to 50%.
  • a 3D image is not generated for cells that do not have abnormal features or BEC-like features as determined using 2D classification
  • a 3D image may be generated for at least one cell or a plurality of cells that do not have abnormal features or BEC-like features.
  • Such images may be used, for example, to verify performance or accuracy of the Al-based cell classification system or method and or one or both of the component 2D cell classifiers.
  • the total number of cells that do not have abnormal features or BEC-like features for which 3D images are generated may be limited to 25% or less, 20% or less, 15% or less, 10% or less, 5% or less, 1% or less, or in a range of 0.1% to 25%, 0.1 % to 20%, 0.1% to 15%, 0.1% to 10%, 0.1% to 5% 0.1% to 1%, 1% to 25%, 1% to 20%, 1% to 15%, 1% to 10%, 1% to 5%, 5% to 25%, 5% to 20%, 5% to 15%, 5% to 10%, 10% to 25%, 10% to 20%, 10% to 15%, 15% to 15%, 15% to 20%, or 20% to 25% of the total number of enumerated analyzed cells having abnormal features, the total number of enumerated analyzed cells having BEC-like features, or the total number of enumerated analyzed cells.
  • Cells that do not have abnormal features or BEC-like features that are selected for 3D imaging may be randomly selected or selected based on pre-selected criteria, such
  • the Al-based cell classification method 300 as described herein may also substantially reduce the number of cells for which a 3D image is generated as compared to an analogous method that does not use an abnormal cell 2D classifier and a BEC 2D classifier (collectively, 2D cell classifiers) and, instead, generates a 3D image of every cell or putative cell identified.
  • the method 300 may reduce the number of cells for which a 3D image is generated by at least 30%, at least 40%, at least 50%, at least 55%, at least 60%, or in a range of 30% to 70%, 30% to 60%, 30% to 55%, 30% to 50%, 30% to 40%, 40% to 70%, 40% to 60%, 40% to 55%, 40% to 50%, 50% to 70%, 50% to 60%, 50% to 55%, 55% to 70%, 55% to 60%, or 60% to 70%.
  • the abnormal cell 2D classifier, the BEC 2D classifier, and the overall Al-based cell classification method may have pre-selected accuracies.
  • the “sensitivity” of the assay is defined as the percent of assayed items (such as cells or the sample as a whole) that are actually positive for a property that are also correctly identified as positive by the assay.
  • the “specificity” of the assay is defined as the percent of assayed items (such as cells or the sample as a whole) that are actually negative for a property that are also correctly identified as positive by the assay.
  • sensitivity tends to be a more significant measure of accuracy than specificity because only cells that are positive for abnormal features or BEC-like features are 3D imaged. Accordingly, a poor sensitivity for either 2D cell classifier can result in cells that are likely to be indicative of cancer not being 3D imaged and not included in the lung cancer test results, potentially resulting in false negatives. In contrast, if specificity is poor, then more cells will be 3D imaged than is needed, unnecessarily slowing the Al-based cell classification somewhat, but not resulting in any cells that are likely to be indicative of cancer being missed and false negatives.
  • 2D abnormal cell sensitivity is the percent of cells that are or would be determined to be abnormal using a known identification method, such as cytology (e.g. review by a pathologist), that are identified by the abnormal cell 2D classifier as having abnormal features.
  • 2D abnormal cell specificity is the percent of cells that are or would be determined to be normal using a known identification method, such as cytology, that are identified as not having abnormal features by the abnormal cell 2D classifier.
  • the pre-selected 2D abnormal cell sensitivity may be at least 90%, at least 95%, at least 98%, or in a range of 90% to 100%, 90% to 99.9%, 90% to 99%, 90% to 98%, 90% to 95%, 95% to 100%, 95% to 99.9%, 95% to 98%, 98% to 100%, 98% to 99.9%, or 98% to 99%.
  • the pre-selected 2D abnormal cell specificity may be at least 65%, at least 70%, at least 75%, or in a range of 65% to 100%, 65% to 99%, 65% to 90%, 65% to 80%, 65% to 75%, 65% to 70%, 70% to 100%, 70% to 99%, 70% to 90%, 70% to 80%, 70% to 75%, 75% to 100%, 75% to 99%, 75% to 90%, or 75% to 80%.
  • 2D BEC sensitivity is the percent of cells that are or would be determined to be BECs using a known identification method, such as cytology, that are identified as having BEC-like features by the BEC 2D classifier.
  • 2D BEC specificity is the percent of cells that are or would be determined to not be BECs using a known identification method, such as cytology, that are identified as not having BEC-like features by the BEC 2D classifier.
  • the pre-selected 2D BEC sensitivity may be at least 85%, at least 90%, at least 95%, or in a range of 85% to 100%, 85% to 99%, 85% to 95%, 85% to 90%, 90% to 100%, 90% to 99%, 90% to 95%, 95% to 100%, or 95% to 99%.
  • the pre-selected 2D BEC specificity may be at least 60%, at least 65%, at least 70%, or in a range of 60% to 100%, 60% to 90%, 60% to 80%, 60% to 70%, 65% to 100%, 65% to 90%, 65% to 80%, 65% to 70%, 70% to 100%, 70% to 90%, or 70% to 80%.
  • an abnormal cell referral rate may also be calculated.
  • the abnormal cell referral rate is the percent of cells that are or would be determined to be abnormal using a known identification method, such as cytology, that are identified as having abnormal features or as having BEC-like features and are, therefore, 3D imaged.
  • the abnormal cell referral rate may be higher than the 2D abnormal cell sensitivity because cells that are abnormal, but that are not identified as having abnormal features, may still be identified as having BEC-like features and subject to 3D imaging for that reason.
  • the pre-selected abnormal cell referral rate may be at least 70%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98%, at least 99%, or in a range of 70% to 99.9%, 80% to 99.9%, 85% to 99.9%, 90% to 99.9%, 95% to 99.9%, 98% to 99.9%, 99% to 99.9 %, 70% to 99%, 80% to 99%, 85% to 99%, 90% to 99%, 95% to 99%, 98% to 99%, 70 % to 98%, 80% to 98%, 85% to 98%, 90% to 98%, 95% to 98%, 70% to 95%, 80% to 95%, 85% to 95%, 90% to 95%, 70% to 90%, 80% to 90%, 85% to 90%, 70% to 85%, 80% to 85%, or 70% to 80%.
  • an abnormal cell true positive referral rate may also be calculated.
  • the abnormal cell true positive referral rate is the percent of cells that are or would be determined to be abnormal using a known identification method, such as cytology, that are identified as having abnormal features by the abnormal cell 2D classifier.
  • the abnormal cell true positive referral rate may, in practice, be less significant than the abnormal cell referral rate because accurate results can be obtained regardless of why abnormal cells were 3D imaged.
  • the abnormal cell true positive referral rate may be at least 60%, at least 70%, at least 80%, at least 90%, or in a range of 60% to 99.9%, 70% to 99.9%, 80% to 99.9%, 90% to 99.9%, 60% to 99%, 70% to 99%, 80% to 99%, 90% to 99%, 60% to 90%, 70% to 90%, 80% to 90%, 60% to 80%, 70% to 80%, or 60% to 70%.
  • a BEC referral rate may also be calculated and considered in determining accuracy.
  • the BEC referral rate is the percent of cells that are or would be determined to be BECs using a known identification method, such as cytology, that are identified as having abnormal features or BEC- like features and are, therefore, 3D imaged.
  • the BEC referral rate may be higher than the BEC 2D classifier sensitivity because BECs that are not identified as having BEC-like features may still be identified as having abnormal features and subject to 3D imaging for that reason.
  • a given number of BECs is typically imaged in order to ensure accuracy of a lung cancer assay of this type, so the number of BECs actually 3D imaged can influence how long it takes to conduct a lung cancer assay with a given patient sample or whether the sample is deemed adequate for accurate results.
  • the pre-selected BEC referral rate may be at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 91%, at least 95%, or in a range of 70% to 99.9%, 75% to 99.9%, 80% to 99.9%, 85% to 99.9%, 90% to 99.9%, 91% to 99.9%, 95% to 99.9%, 70% to 99%, 75% to 99%, 80% to 99%, 85% to 99%, 90% to 99%, 91% to 99%, 95% to 99%, 70% to 95%, 75% to 95%, 80% to 95%, 85% to 95%, 90% to 95%, 70% to 91%, 75% to 91%, 80% to 91%, 85% to 91%, 90% to 91%, 70% to 90%, 75% to 90%, 80% to 90%, 85% to 90%, 70% to 85%, 75% to 85%, or 80% to 85%.
  • a BEC true positive referral rate may also be calculated.
  • the BEC true positive referral rate is the percent of cells that are or would be determined to be BECs using a known identification method, such as cytology, that are identified as having BEC-like features by the BEC 2D classifier.
  • the BEC true positive referral rate may, in practice, be less significant than the BEC referral rate because accurate results can be obtained regardless of why BECS were 3D imaged.
  • the BEC true positive referral rate may be at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, or in a range of 50% to 99.9%, 60% to 99.9%, 70% to 99.9%, 80% to 99.9%, 90% to 99.9%, 50% to 99%, 60% to 99%, 70% to 99%, 80% to 99%, 90% to 99%, 50% to 90%, 60% to 90%, 70% to 90%, 80% to 90%, 50% to 80%, 60% to 80%, 70% to 80%, 50% to 70%, 60% to 70%, or 50% to 60%.
  • the Al-based cell classification method 300 may also have a pre-selected 2D rejection rate.
  • the 2D rejection rate is the proportion of the total number of enumerated analyzed cells that are identified as not having abnormal features or BEC-like features and, therefore, are not 3D imaged.
  • the 2D rejection rate may be 70% or less, 60% or less, 55% or less, 50% or less, or in a range of 0% to 70%, 1% to 70%, 10% to 70%, 25% to 70%, 40% to 70%, 50% to 70%, 55% to 70%, 60% to 70%, 0% to 60%, 1% to 60%, 10% to 60%, 25% to 60%, 40% to 60%, 50% to 60%, 55% to 60%, 0% to 55%, 1% to 55%, 10% to 55%, 25% to 55%, 40% to 55%, 50% to 55%, 0% to 50%, 1% to 50%, 10% to 50%, 25% to 50%, or 40% to 50%. [0091] Accuracy values may be calculated for the 3D classifier as well.
  • the Al-based cell classification method may have a 3D abnormal cell sensitivity that is the percent of cells that are or would be determined to be abnormal using a known identification method, such as cytology (e.g. review by a pathologist), that are identified by the 3D classifier as abnormal (or positive as indicative of lung cancer).
  • cytology e.g. review by a pathologist
  • the 3D abnormal cell sensitivity the number of false negative cells not referred for 3D imaging by the 2D classifiers must be subtracted from the known abnormal cell total, as the 3D classifier did not have the opportunity to classify such cells as abnormal.
  • a pre-selected minimum 3D abnormal cell sensitivity helps avoid false negative test results. Similar calculations may be made for other accuracy parameters of the 3D cell classifier.
  • the lung cancer detection method 200 also has pre-selected lung cancer accuracy parameters for the lung cancer test as a whole.
  • the lung cancer sensitivity is the percent of patients who actually have lung cancer who are identified as having lung cancer by the lung cancer detection method (a positive rest result).
  • the Al-based cell classification method 300 also has pre-selected assay accuracy parameters. In some embodiments, the may be the same as the lung cancer accuracy parameters.
  • the sensitivity of any lung cancer detection method that uses 3D imaging of the type used in method 300 depends on the number of BECs that are 3D imaged.
  • the number of BECs 3D imaged from a patient sample may be determined from the total number of enumerated BECs and may be set at a threshold value so that a pre-selected assay accuracy, such as a pre-selected assay sensitivity is statistically likely to have been attained.
  • Method 300 may continue to analyze cells 10 until reaching a pre-selected threshold value that reflects the total number of enumerated BECs and may be at least 250, at least 400, at least 500 at least 600, at least 700, at least 800, at least 900, at least 1000, at least 1100, at least 1200, at least 1300, at least 1400, at least 1500, at least 2000, at least 2500, or in a range of 250 to 2500, 250 to 2000, 400 to 1900, 600 to 1800, 800 to 1700, 1000 to 1600, 1200 to 1600, or 1300 to 1500.
  • Method 300 may reach a pre-selected sensitivity more quickly than an analogous method that does not use an abnormal cell 2D classifier and a BEC 2D classifier and, instead, generates a 3D image of every cell or putative cell identified because time is not spent capturing 3D images of cells that are not likely to be indicative of lung cancer.
  • the assay abnormal cell sensitivity is the percent of cells that are or would be identified as abnormal using known identification methods that are identified as abnormal by the Al-based cell classification method 300.
  • the assay abnormal cell sensitivity may be approximated by multiplying the 2D abnormal cell sensitivity by the 3D abnormal cell sensitivity.
  • the assay abnormal cell specificity is the percent of cells that are or would be identified as normal using known identification methods that are identified as normal by the AI- based cell classification method 300.
  • the assay abnormal cell sensitivity may be approximated by multiplying the 2D abnormal cell specificity by the 3D abnormal cell specificity.
  • the present disclosure further includes a method 500 as set forth in the flow chart of Fig. 5, for training 2D cell classifiers which may be used in method 200, method 300, or method 310.
  • 2D classifier training is governed by a binary ground truth; for the abnormal cell 2D classifier, the cell either has abnormal features or does not have abnormal features, while for the BEC 2D classifier, the cell either has BEC-like features or does not have BEC-like features.
  • Elements of the optical tomography system 100 are referenced in this description of the training method 500 as examples. Similar components of different optical tomography systems may also be used in connection with the training method 500. However, the trained values generated using training method 500, as implemented in the abnormal cell 2D classifier and the BEC 2D classifier, may be most useful when used with the same optical tomography system used in training method 500, or with an optical tomography system with identical or highly analogous components and with performance specifications that are not likely to produce differences in any results that are outside of pre-selected accuracies.
  • an optical tomography system 100 is used, for example, according to steps 400 and 410 of method 310, to generate a set of abnormal cell representative 2D images of a plurality of cells identified as abnormal by a known identification method, such as cytology, a set of BEC representative 2D images of a plurality of cells identified as BECs by a known identification method, and a set of normal/non-BEC representative 2D images of a plurality of lung sample cells that are identified as normal and not BECs by a known identification method.
  • a known identification method such as cytology
  • BEC representative 2D images of a plurality of cells identified as BECs by a known identification method a set of normal/non-BEC representative 2D images of a plurality of lung sample cells that are identified as normal and not BECs by a known identification method.
  • a cell is identified as abnormal by a known identification method if it is confirmed by a cytopathologist to be in one of the following diagnostic categories: 1) atypia, dysplasia, pre-cancerous cell, malignant cell, or 2) pleomorphic parakeratosis, type II pneumocyte.
  • a cell is identified as a BEC by a known identification method if it is confirmed by two cytopathologists independently to be a BEC.
  • a cell is identified as normal if it is randomly acquired from a patient sample and identified as such by a cytopathologist because it does not meet abnormal cell criteria, and a cell is identified as a non-BEC if any cytopathologist identifies the cell as not a BEC.
  • Cells for training purposes may, in some embodiments, be acquired from the same type of patient sample (e.g., sputum) or prepared in the same manner (e.g., placed in the same optical medium or stained with the same stain) as used in connection with the Al-based cell classification methods and systems that will be operated using the trained values.
  • patient sample e.g., sputum
  • prepared in the same manner e.g., placed in the same optical medium or stained with the same stain
  • method 500 illustrates concurrent training of the abnormal cell 2D classifier and the BEC 2D classifier for simplicity
  • the 2D cell classifiers may be trained separately, optionally using separate training cell sets, as is discussed in the Examples.
  • a processor computes a plurality of cell feature measurements for each of the plurality of cells using the representative 2D image of the cell generated in step 510.
  • step 530 the processor performs a regression between the cell feature measurements and the known identifier (i) abnormal, ii) BEC, or iii) normal and non-BEC, or in methods that separately train the 2D cell classifiers iii) normal or non-BEC, whichever is the opposite of the other known identifier used) of the cell to designate an abnormal 2D cell classification score and a BEC 2D cell classification score for each cell.
  • the processor performs a regression between the cell feature measurements and the known identifier (i) abnormal, ii) BEC, or iii) normal and non-BEC, or in methods that separately train the 2D cell classifiers iii) normal or non-BEC, whichever is the opposite of the other known identifier used) of the cell to designate an abnormal 2D cell classification score and a BEC 2D cell classification score for each cell.
  • step 540 the processor designates a test identifier of i) having abnormal features, ii) having BEC-like features, or iii) not having abnormal or BEC-like features based on the abnormal 2D cell classification score and the BEC 2D classification score for each cell.
  • step 550 the respective test identifier is compared to the respective known identifier for each cell to calculate accuracy of both 2D cell classifiers.
  • the accuracy of the 2D cell classifiers may be compared to preselected accuracy requirements and, if accuracy requirements are met, designating the 2D cell classifiers as trained, or, if the accuracy requirements are not net, in step 570 adjusting a parameter of the regression for the 2D cell classifier that does not meet accuracy requirements and repeating steps 530 through 560 for at least the adjusted 2D cell classifier.
  • the processor adjusts the 2D cell classifier algorithm using Adaptively Boosted Logistic Regression, Random Forest, Decision Trees, or any combination thereof.
  • Adaptively Boosted Logistic Regression uses logistic regression in an iterative loop to improve overall classification merit.
  • the pre-selected 2D abnormal cell sensitivity value may be at least 90%, at least 95%, or in a range of 90% to 100%, 90% to 99.9%, 90% to 99%, 90% to 98%, 95% to 100%, 95% to 99.9%, 95% to 99%, or 95% to 98%, or any other 2D abnormal cell sensitivity value recited herein for an abnormal cell 2D classifier.
  • the pre-selected 2D abnormal cell specificity value may be at least 65%, at least 70%, at least 75%, or in a range of 65% to 100%, 65% to 99%, 65% to 90%, 65% to 80%, 65% to 75%, 70% to 100%, 70% to 99%, 70% to 90%, 70% to 80%, 70% to 75%, 75% to 100%, 75% to 90%, or 75% to 80%, or any other 2D abnormal cell specificity value recited herein for an abnormal cell 2D classifier.
  • the pre-selected 2D BEC sensitivity value may be at least 85%, at least 90%, at least 95%, or in a range of 85% to 100%, 85% to 99.9%, 85% to 99%, 95% to 90%, 90% to 100%, 90% to 99.9%, 90% to 95%, 95% to 100%, 95% to 99.9%, or 95% to 99% or any other 2D BEC sensitivity value recited herein for a BEC 2D classifier.
  • the pre-selected 2D BEC specificity value may be at least 60%, at least 65%, at least 70%, or in a range of 60% to 100%, 60% to 99%, 60% to 90%, 60% to 80%, 60% to 75%, 60% to 70%, 60% to 65%, 65% to 100%, 65% to 99%, 65% to 90%, 65% to 80%, 65% to 75%, 65% to 70%, 70% to 100%, 70% to 99%, 70% to 90%, 70% to 80%, or 70% to 75%, or any other 2D BEC specificity value recited herein for a BEC 2D classifier.
  • any concentration range, percentage range, ratio range, or integer range is to be understood to include any values or subranges within the recited range unless otherwise indicated.
  • the term “or” is generally employed in its sense including “or” (/. ⁇ ., to mean either one, both, or any combination thereof of the alternatives) unless the content dictates otherwise.
  • the singular forms “a,” “an,” and “the” include plural referents unless the content dictates otherwise.
  • the terms “include,” and “have” and their variants are used synonymously and are to be construed as non-limiting.
  • the term “a combination thereof’ as used herein refers to all possible combinations of the listed items preceding the term.
  • A, B, C, or a combination thereof is intended to refer to any one of A, B, C, AB, AC, BC, or ABC.
  • the term “combinations thereof’ as used herein refers to all possible combinations of the listed items preceding the term. For instance, “A, B, C, and combinations thereof’ is intended to refer to all of A, B, C, AB, AC, BC, and ABC.
  • a lung cancer detection method was implemented using a CELL-CT® optical tomography system.
  • the lung cancer detection method was designed to detect lung cancer in a pre-invasive stage, when treatment is easier and more likely to be successful.
  • cell search mode was initiated.
  • the CELL-CT® continuously swept its plane of focus, in 1 pm steps, through the lumen of the micro-capillary tube to identify the linear, radial and angular position of cells. Detected dark objects beyond a certain size were identified for potential 3D capture.
  • the 2D image set was filtered to take the central image of a cell. This 2D image was then evaluated using the 2D cell classifiers to determine if the cell had abnormal features or BEC-like features as described below.
  • Each cell was also analyzed in projection image capture mode, during which the tube was rotated so that the optical tomography system generated 500 high-resolution images of the same diagonal cross-section of the capillary tube, taken over 360 degrees of tube rotation.
  • These images were pseudo-projection images, which are simulations of projection images created by integrating the light from the objective lens as the focal plane was swept through the nucleus of any cell contained in the images. The pseudo-projection images thus represented the entire nuclear content in a single image, taken from a single perspective.
  • the corrected pseudo-projection images were processed using filtered back projection to yield a 3D tomographic representation of the cell, also referred to as a 3D image.
  • 3D tomographic representations were created for a substantial number of cells, including multiple cells identified in an optical medium volume within the optical path of the high- magnification microscope before the optical medium is advanced to a different volume by pressure.
  • An example of such a 3D image for both squamous intermediate and dysplastic cells is shown in Fig. 6.
  • the 3D images were evaluated by a cytopathologist to determine if the cell was abnormal, a BEC, or normal, or not a BEC.
  • An abnormal 2D cell classification score designated the likelihood that the object under investigation had abnormal features.
  • the abnormal 2D cell classification score for a number of cells imaged was used to prepare a composite abnormal 2D cell classification score, which determined whether the patient likely has lung cancer.
  • the BEC 2D cell classification score designated the likelihood that the object under investigation had BEC-like features.
  • the BEC classification score for a number of cells imaged was used to prepare a composite BEC 2D cell classification score, which helped determine how many BECs had been analyzed.
  • An abnormal cell 2D classifier and a BEC 2D classifier for use in the lung cancer detection method were trained and tested. Images of cells obtained as described above were used by cytopathologists employing conventional, known cell identification methods to designate a known identifier for each cell also given a test identifier in the training method.
  • abnormal cell 2D classifier development followed a supervised learning process that was governed by binary known identifiers (abnormal/normal or having abnormal features/not having abnormal features).
  • a cell was designated with the known identifier “abnormal” or “having abnormal features” if it was confirmed to be in one of the following diagnostic categories by a cytopathologist:
  • a cell was designated as not having abnormal features or “normal,” if it was randomly acquired and was designated as normal by a cytopathologist (due to the absence of the abnormal criteria identified above).
  • BEC 2D classifier development followed a supervised learning process that was governed by binary known identifiers (BEC/non-BEC, or having BEC-like features/not having BEC-like features).
  • a cell was designated with the known identifier “BEC” or “having BEC-like features” when two cytotechnologists independently agreed that it was a BEC.
  • a cell categorized as non-BEC by any cytopathologist was designated with the known identifier “non-BEC” or “not having BEC-like features.”
  • This aggregate data set formed the training set to create the abnormal cell 2D classifier.
  • Adaptively boosted logistic regression (Ada-boost) was used to refine the abnormal cell 2D classifier.
  • Ada-boost uses logistic regression in an iterative loop to improve overall classification merit.
  • Cross-validation was used to limit the complexity of the model to ensure that it generalized.
  • the trained abnormal cell 2D classifier was used to classify cells that were not a part of the training set. Results showing the known identifier and corresponding test identifier for the trained abnormal cell 2D classifier are presented in Table 1 and Fig. 7.
  • the 2D abnormal cell sensitivity was 95% and the 2D abnormal cell specificity was 74.7%. Sensitivity was required to be greater than 90% and specificity was required to be greater than 65% for the abnormal cell 2D classifier to be considered trained, so the trained abnormal cell 2D classifier exceeded pre-selected accuracy values.
  • Fig. 7 shows the cumulative probability density curve for the trained abnormal cell 2D classifier using the results of Table 1.
  • 93.3% (1-0.067) received a correct test identifier (red line) and were then imaged in 3D.
  • 96.2% of cells having a known identifier of having BEC-like features green line
  • 87.5% of cells having a known identifier of not having BEC-like features light blue line
  • 12.5% of miscellaneous cells were eventually subject to 3D imaging.
  • BEC-like features was designated as 1 for both the known identifier and the test identifier. Not having BEC-like features was designated as 0 for both the known identifier and the test identifier.
  • a training set of 12,572 cells were imaged and provided known identifiers by cytopathologists and test identifiers by the BEC 2D classifier. To provide the test identifiers, the cell and nucleus were segmented from the block of voxels that contained the cell and features were automatically measured by the optical tomography system for each cell then used to designate a test identifier. The test identifier for each cell was then compared to the known identifier to determine if the test identifier was correct.
  • This aggregate data set formed the training set to create the BEC 2D classifier.
  • Adaptively boosted logistic regression (Ada- boost) was used to refine the abnormal cell 2D classifier.
  • Ada-boost uses logistic regression in an iterative loop to improve overall classification merit.
  • Cross-validation was used to limit the complexity of the model to ensure that it generalized.
  • the trained BEC 2D classifier was used to classify cells that were not a part of the training set. Results showing the known identifier and corresponding test identifier for the trained BEC cell classifier are presented in Table 2 and Fig. 8.
  • the 2D BEC sensitivity was 88% and the 2D BEC specificity was 66%.
  • Sensitivity was required to be greater than 85% and specificity was required to be greater than 60% for the BEC 2D classifier to be considered trained, so the trained BEC 2D classifier exceeded pre-selected accuracy values.
  • Fig. 8 shows the cumulative probability density curve for the trained BEC 2D classifier using the results of Table 2. For cells having a known identifier of having BEC -like features, 86.6% (1-0.134) received a correct test identifier (green line) and were then imaged in 3D. For cells having a known identifier of not having BEC-like features, 81% received a correct test identifier (light blue line).
  • the trained 2D cell classifiers were subjected to three independent tests following development to ensure robustness.
  • the lung cancer detection method was performed using the trained 2D classifiers and a trained 3D classifier, but, as during training in Example 1, was set to acquire 3D images of all objects for which 2D images were obtained so that accuracy could be assessed.
  • the trained 2D cell classifiers were embedded in executable code that also implemented the remainder of the standard lung cancer test and tested using cells from five adenocarcinoma and two squamous carcinoma cell lines. Various measures of accuracy were assessed. The abnormal cell referral rate was 98.7%.
  • the BEC 2D classifier exhibited a BEC 2D sensitivity of 90% and a BEC 2D specificity of 44%.
  • the Pilot CAS was conducted using a set of specimens that was like the ones used to train the 2D cell classifiers to evaluate the referral rates of the 2D cell classifiers. 34 specimens represented by 267,659 cells were tested. The abnormal cell referral rate was calculated to be 95%. The BEC referral rate was 91%. The abnormal cell referral rate and BEC referral rate demonstrated that training of the 2D algorithms generalized well to a population of cells that were not a part of the training process. Case level accuracy (whether the patient was positive or negative for cells indicative of lung cancer) was not altered by use of 2D cell classifiers and exclusion of some cells from 3D imaging.
  • the 2D rejection rate was not specified, but it was estimated at 53% during training of the 2D cell classifiers.
  • the measured 2D rejection rate in the CAS-A study was 60%, and the median 2D rejection rate by case was 52%.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Biomedical Technology (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Public Health (AREA)
  • Pathology (AREA)
  • Chemical & Material Sciences (AREA)
  • Quality & Reliability (AREA)
  • Radiology & Medical Imaging (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • Multimedia (AREA)
  • Mathematical Physics (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Primary Health Care (AREA)
  • Epidemiology (AREA)
  • Databases & Information Systems (AREA)
  • Geometry (AREA)
  • Computer Graphics (AREA)
  • Signal Processing (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • Immunology (AREA)
  • Dispersion Chemistry (AREA)
  • Computational Linguistics (AREA)
  • Investigating Or Analysing Biological Materials (AREA)
  • Image Analysis (AREA)
PCT/US2023/029019 2022-08-02 2023-07-28 Ai-based cell classification method and system Ceased WO2024030346A1 (en)

Priority Applications (6)

Application Number Priority Date Filing Date Title
AU2023317489A AU2023317489B2 (en) 2022-08-02 2023-07-28 Ai-based cell classification method and system
CA3263775A CA3263775A1 (en) 2022-08-02 2023-07-28 Ai-based cell classification method and system
IL318743A IL318743A (en) 2022-08-02 2023-07-28 Artificial intelligence-based cell classification method and system
CN202380067171.3A CN119866434A (zh) 2022-08-02 2023-07-28 基于ai的细胞分类方法和系统
EP23850630.7A EP4565868A4 (en) 2022-08-02 2023-07-28 AI-BASED METHOD AND SYSTEM FOR CELL CLASSIFICATION
JP2025505966A JP2025525906A (ja) 2022-08-02 2023-07-28 Aiベースの細胞分類方法およびシステム

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US202263394550P 2022-08-02 2022-08-02
US63/394,550 2022-08-02

Publications (1)

Publication Number Publication Date
WO2024030346A1 true WO2024030346A1 (en) 2024-02-08

Family

ID=89849792

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2023/029019 Ceased WO2024030346A1 (en) 2022-08-02 2023-07-28 Ai-based cell classification method and system

Country Status (8)

Country Link
US (2) US12602781B2 (cg-RX-API-DMAC7.html)
EP (1) EP4565868A4 (cg-RX-API-DMAC7.html)
JP (1) JP2025525906A (cg-RX-API-DMAC7.html)
CN (1) CN119866434A (cg-RX-API-DMAC7.html)
AU (1) AU2023317489B2 (cg-RX-API-DMAC7.html)
CA (1) CA3263775A1 (cg-RX-API-DMAC7.html)
IL (1) IL318743A (cg-RX-API-DMAC7.html)
WO (1) WO2024030346A1 (cg-RX-API-DMAC7.html)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP7783889B2 (ja) * 2021-06-17 2025-12-10 富士フイルム株式会社 医用画像診断システム、医用画像診断システムの作動方法及びプログラム

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040076319A1 (en) * 2002-04-19 2004-04-22 Fauver Mark E. Method and apparatus of shadowgram formation for optical tomography
US20100135566A1 (en) * 2008-12-01 2010-06-03 Olympus Soft Imaging Solutions Gmbh Analysis and classification, in particular of biological or biochemical objects, on the basis of time-lapse images, applicable in cytometric time-lapse cell analysis in image-based cytometry
WO2017053592A1 (en) * 2015-09-23 2017-03-30 The Regents Of The University Of California Deep learning in label-free cell classification and machine vision extraction of particles
US20200167914A1 (en) * 2017-07-19 2020-05-28 Altius Institute For Biomedical Sciences Methods of analyzing microscopy images using machine learning
US20210049425A1 (en) * 2018-02-28 2021-02-18 Visiongate, Inc. Morphometric detection of malignancy associated change

Family Cites Families (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6522775B2 (en) 2001-03-28 2003-02-18 Alan C. Nelson Apparatus and method for imaging small objects in a flow stream using optical tomography
US6519355B2 (en) 2001-03-28 2003-02-11 Alan C. Nelson Optical projection imaging system and method for automatically detecting cells having nuclear and cytoplasmic densitometric features associated with disease
US6591003B2 (en) 2001-03-28 2003-07-08 Visiongate, Inc. Optical tomography of small moving objects using time delay and integration imaging
US7907765B2 (en) 2001-03-28 2011-03-15 University Of Washington Focal plane tracking for optical microtomography
US6636623B2 (en) * 2001-08-10 2003-10-21 Visiongate, Inc. Optical projection imaging system and method for automatically detecting cells with molecular marker compartmentalization associated with malignancy and disease
US7811825B2 (en) 2002-04-19 2010-10-12 University Of Washington System and method for processing specimens and images for optical tomography
US7197355B2 (en) 2002-04-19 2007-03-27 Visiongate, Inc. Variable-motion optical tomography of small objects
US6697508B2 (en) 2002-05-10 2004-02-24 Visiongate, Inc. Tomographic reconstruction of small objects using a priori knowledge
WO2005091203A2 (en) * 2004-03-12 2005-09-29 Aureon Laboratories, Inc. Systems and methods for treating, diagnosing and predicting the occurrence of a medical condition
US7494809B2 (en) 2004-11-09 2009-02-24 Visiongate, Inc. Automated cell sample enrichment preparation method
US7569789B2 (en) 2006-03-16 2009-08-04 Visiongate, Inc. Cantilevered coaxial flow injector apparatus and method for sorting particles
US7867778B2 (en) 2007-02-23 2011-01-11 Visiongate, Inc. Fluid focusing for positional control of a specimen for 3-D imaging
US20100088264A1 (en) * 2007-04-05 2010-04-08 Aureon Laboratories Inc. Systems and methods for treating diagnosing and predicting the occurrence of a medical condition
US7835561B2 (en) 2007-05-18 2010-11-16 Visiongate, Inc. Method for image processing and reconstruction of images for optical tomography
US7787112B2 (en) 2007-10-22 2010-08-31 Visiongate, Inc. Depth of field extension for optical tomography
US8090183B2 (en) 2009-03-12 2012-01-03 Visiongate, Inc. Pattern noise correction for pseudo projections
WO2009155151A2 (en) 2008-06-20 2009-12-23 Visiongate, Inc. Functional imaging of cells optical projection tomography
US8155420B2 (en) 2009-05-21 2012-04-10 Visiongate, Inc System and method for detecting poor quality in 3D reconstructions
US9594072B2 (en) * 2015-06-30 2017-03-14 Visiongate, Inc. System and method for determining cell adequacy in a cytological analysis system
WO2017117210A1 (en) * 2015-12-30 2017-07-06 Visiongate, Inc. System and method for automated detection and monitoring of dysplasia and administration of chemoprevention
US10753857B2 (en) 2017-07-24 2020-08-25 Visiongate Inc. Apparatus and method for measuring microscopic object velocities in flow
US12612663B2 (en) * 2018-06-06 2026-04-28 Visiongate, Inc. Morphometric detection of DNA mismatch repair deficiency
US11428692B2 (en) 2018-07-11 2022-08-30 Visiongate, Inc. Specimen enrichment for optical tomography cell analysis
EP3620983B1 (en) * 2018-09-05 2023-10-25 Sartorius Stedim Data Analytics AB Computer-implemented method, computer program product and system for data analysis
IL301977B2 (en) 2020-11-11 2024-02-01 Imagene Ai Ltd Methods and systems for training a model to diagnose abnormalities in tissue sampling

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040076319A1 (en) * 2002-04-19 2004-04-22 Fauver Mark E. Method and apparatus of shadowgram formation for optical tomography
US20100135566A1 (en) * 2008-12-01 2010-06-03 Olympus Soft Imaging Solutions Gmbh Analysis and classification, in particular of biological or biochemical objects, on the basis of time-lapse images, applicable in cytometric time-lapse cell analysis in image-based cytometry
WO2017053592A1 (en) * 2015-09-23 2017-03-30 The Regents Of The University Of California Deep learning in label-free cell classification and machine vision extraction of particles
US20180286038A1 (en) * 2015-09-23 2018-10-04 The Regents Of The University Of California Deep learning in label-free cell classification and machine vision extraction of particles
US20200167914A1 (en) * 2017-07-19 2020-05-28 Altius Institute For Biomedical Sciences Methods of analyzing microscopy images using machine learning
US20210049425A1 (en) * 2018-02-28 2021-02-18 Visiongate, Inc. Morphometric detection of malignancy associated change

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
LEE ET AL.: "Building predictive in vitro pulmonary toxicity assays using highthroughput imaging and artificial intelligence", ARCH TOXICOL, vol. 92, no. 6, 28 April 2018 (2018-04-28), pages 2055 - 2075, XP036526375, Retrieved from the Internet <URL:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6002469/pdf/204-2018-Article-2213.pdf> [retrieved on 20231019], DOI: 10.1007/s00204-018-2213-0 *
See also references of EP4565868A4 *

Also Published As

Publication number Publication date
US20240212137A1 (en) 2024-06-27
CN119866434A (zh) 2025-04-22
JP2025525906A (ja) 2025-08-07
AU2023317489A1 (en) 2025-02-20
EP4565868A1 (en) 2025-06-11
EP4565868A4 (en) 2026-01-07
US12602781B2 (en) 2026-04-14
AU2023317489B2 (en) 2026-06-04
IL318743A (en) 2025-04-01
CA3263775A1 (en) 2024-02-08
US20260237069A1 (en) 2026-08-13

Similar Documents

Publication Publication Date Title
US12399178B2 (en) Specimen enrichment for optical tomography cell analysis
US7274809B2 (en) Computerized methods and systems related to the detection of malignancy-associated changes (MAC) to detect cancer
US11069054B2 (en) System and method for automated detection and monitoring of dysplasia and administration of immunotherapy and chemotherapy
CN108027364B (zh) 用于确定细胞学分析系统中的细胞充分性的系统及方法
US20250092466A1 (en) Morphometric genotyping of cells using optical tomography for detecting tumor mutational burden
Huang et al. A robust and scalable framework for hallucination detection in virtual tissue staining and digital pathology
US12602781B2 (en) AI-based cell classification method and system
US12612663B2 (en) Morphometric detection of DNA mismatch repair deficiency
US10753857B2 (en) Apparatus and method for measuring microscopic object velocities in flow
US20240296949A1 (en) Digital cell pathology display and classification methods and system
WO2025132537A1 (en) Method of classifying cancer subtypes

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23850630

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2025505966

Country of ref document: JP

WWE Wipo information: entry into national phase

Ref document number: 318743

Country of ref document: IL

WWE Wipo information: entry into national phase

Ref document number: AU2023317489

Country of ref document: AU

ENP Entry into the national phase

Ref document number: 2023317489

Country of ref document: AU

Date of ref document: 20230728

Kind code of ref document: A

WWE Wipo information: entry into national phase

Ref document number: 2023850630

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2023850630

Country of ref document: EP

Effective date: 20250303

WWE Wipo information: entry into national phase

Ref document number: 202380067171.3

Country of ref document: CN

WWP Wipo information: published in national office

Ref document number: 202380067171.3

Country of ref document: CN

WWP Wipo information: published in national office

Ref document number: 2023850630

Country of ref document: EP