AU2021103689A4 - A Method for Automatic Detection of Infected Erythrocytes - Google Patents
A Method for Automatic Detection of Infected Erythrocytes Download PDFInfo
- Publication number
- AU2021103689A4 AU2021103689A4 AU2021103689A AU2021103689A AU2021103689A4 AU 2021103689 A4 AU2021103689 A4 AU 2021103689A4 AU 2021103689 A AU2021103689 A AU 2021103689A AU 2021103689 A AU2021103689 A AU 2021103689A AU 2021103689 A4 AU2021103689 A4 AU 2021103689A4
- Authority
- AU
- Australia
- Prior art keywords
- clump
- cell
- concavity
- valley
- erythrocytes
- 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
Links
- 210000003743 erythrocyte Anatomy 0.000 title claims abstract description 68
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000001514 detection method Methods 0.000 title claims abstract description 12
- 210000004027 cell Anatomy 0.000 claims abstract description 90
- 210000004369 blood Anatomy 0.000 claims abstract description 27
- 239000008280 blood Substances 0.000 claims abstract description 27
- 230000011218 segmentation Effects 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 244000045947 parasite Species 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000000926 separation method Methods 0.000 claims description 4
- 208000030852 Parasitic disease Diseases 0.000 claims description 3
- 108010077544 Chromatin Proteins 0.000 description 12
- 210000003483 chromatin Anatomy 0.000 description 12
- 230000008901 benefit Effects 0.000 description 9
- 201000004792 malaria Diseases 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000007704 transition Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 3
- 208000015181 infectious disease Diseases 0.000 description 3
- 230000003071 parasitic effect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 230000000877 morphologic effect Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000036281 parasite infection Effects 0.000 description 1
- 210000005259 peripheral blood Anatomy 0.000 description 1
- 239000011886 peripheral blood Substances 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000010186 staining Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 210000003812 trophozoite Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G01N1/00—Sampling; Preparing specimens for investigation
- G01N1/28—Preparing specimens for investigation including physical details of (bio-)chemical methods covered elsewhere, e.g. G01N33/50, C12Q
- G01N1/2813—Producing thin layers of samples on a substrate, e.g. smearing, spinning-on
- G01N2001/282—Producing thin layers of samples on a substrate, e.g. smearing, spinning-on with mapping; Identification of areas; Spatial correlated pattern
-
- 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
- G01N15/01—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials specially adapted for biological cells, e.g. blood cells
- G01N2015/012—Red blood cells
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
- G01N2021/0181—Memory or computer-assisted visual determination
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2333/00—Assays involving biological materials from specific organisms or of a specific nature
- G01N2333/435—Assays involving biological materials from specific organisms or of a specific nature from animals; from humans
- G01N2333/44—Assays involving biological materials from specific organisms or of a specific nature from animals; from humans from protozoa
- G01N2333/445—Plasmodium
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2800/00—Detection or diagnosis of diseases
- G01N2800/26—Infectious diseases, e.g. generalised sepsis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/48—Biological material, e.g. blood, urine; Haemocytometers
- G01N33/483—Physical analysis of biological material
- G01N33/487—Physical analysis of biological material of liquid biological material
- G01N33/49—Blood
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The present disclosure relates to a method for automatic detection of infected
erythrocytes based on concavity point identification and pseudo-valley-based thresholding.
The method comprises: Pre-processing a blood smear image; identifying a clump cell in the
blood smear image; determining a concavity point in the clump cell, the distance from the
concavity point is the maximum; determining a second concavity point, wherein, it will be
geometrically opposite to the first concavity point and among all the points opposite to the
first concavity point, it will have the shortest distance; segmenting the clump cell, wherein
the concavity pair points are joined to divide the clump into individual erythrocytes and re
examining the segmented clump; and determining a pseudo valley, wherein, this valley is
used for segmentation of erythrocytes from a background using Otsu's thresholding.
15
Sto
~414~'B!'
tMY
Description
Sto
~414~'B!'
tMY
A Method for Automatic Detection of Infected Erythrocytes
The present disclosure relates to a method for automatic detection of infected erythrocytes based on concavity point identification and pseudo-valley-based thresholding.
Different techniques are available in the market to diagnose malaria, i.e., rapid diagnosis test (RDTs), quantitative Buffy coat (QBC), and manual microscopic examinations of peripheral blood smear. However, a manual exam is considered the gold standard to validate various examination techniques. Two different blood smears (i.e., thin, and thick smear) are used for manual microscopic examination. Still, a thin smear is considered more efficient, as the malaria species analysis can be done in a thin smear.
To avoid the tedious process of manual microscopic examination, different image processing methodologies are developed for the timely diagnosis of malaria. Due to malaria infection, the homogeneity of the erythrocyte gets distorted. Various computerized techniques are suggested in literature studies for erythrocyte segmentation and erythrocyte analysis to detect malaria infection in blood smear images. In order to make the existing solutions more efficient, there is a need to develop a method for automatic detection of infected erythrocytes based on concavity point identification and pseudo-valley-based thresholding.
The present disclosure relates to a method for automatic detection of infected erythrocytes based on concavity point identification and pseudo-valley-based thresholding. In this disclosure, a framework for erythrocyte analysis has been developed. It entails segmenting erythrocytes into individual cells and subsequently analyzing the erythrocyte to identify malaria-infected cells. Otsu's thresholding and morphological filtering are used to separate the erythrocytes from the complicated backdrop. The suggested method also segments the clump erythrocyte into individual cells, based on the detection of the concavity point with a distance measure between the clump cell's boundary and the boundary of the related convex hull. Finally, this approach is unaffected by any other features that are prone to distortion.
In an embodiment, a method 100 for automatic detection of infected erythrocytes based on concavity point identification and pseudo-valley-based thresholding comprises the following steps: at step 102, pre-processing a blood smear image, wherein pre-processing comprises: converting the RGB blood smear image to grayscale; normalizing the grayscale blood smear image; and processing the grayscale blood smear image using median filtering technique to remove the noise present in smear images; at step 104, identifying a clump cell in the grayscale blood smear image; at step 106, determining a concavity point in the clump cell, wherein, out of all distances, from points on the original image boundary to the points on the convex hull boundary which are on the same straight line, the distance from the concavity point is the maximum; at step 108, determining a second concavity point, wherein, it will be geometrically opposite to the first concavity point and among all the points opposite to the first concavity point, it will have the shortest distance; at step 110, segmenting the clump cell, wherein the concavity pair points are joined to divide the clump into individual erythrocytes and re-examining the segmented clump; and at step 112, determining a pseudo valley threshold, wherein, this valley is used for segmentation of erythrocytes from a background using Otsu's thresholding.
To further clarify advantages and features of the present disclosure, a more particular description of the invention will be rendered by reference to specific embodiments thereof, which is illustrated in the appended drawings. It is appreciated that these drawings depict only typical embodiments of the invention and are therefore not to be considered limiting of its scope. The invention will be described and explained with additional specificity and detail with the accompanying drawings.
These and other features, aspects, and advantages of the present disclosure will become better understood when the following detailed description is read with reference to the accompanying drawings in which like characters represent like parts throughout the drawings, wherein:
Figure 1 illustrates a method for automatic detection of infected erythrocytes based on concavity point identification and pseudo-valley-based thresholding in accordance with an embodiment of the present disclosure.
Figure 2 illustrates (A) Architecture of erythrocyte segmentation; (B) (a) Boundary of the clump cell (Ib) and (b) boundary of the convex hull of the clump cell (Ic); (C) Distance of the points between lb and Ic. White: distance from concavity point, Red: distance from the other point; (D) (a) Boundary pixels with 0; value within 80° to 90°, (b) Pair of concavity point's determined, (c) Clump cell image, (d) Segmented clump by joining the pair of concavity points; and (E) Microscopic blood smear images (a) RGB image, (b) Grayscale image, (c) Histogram of grayscale image, (d) Extracted erythrocyte, (e) Histogram of the extracted erythrocyte in accordance with an embodiment of the present disclosure.
Further, skilled artisans will appreciate that elements in the drawings are illustrated for simplicity and may not have necessarily been drawn to scale. For example, the flow charts illustrate the method in terms of the most prominent steps involved to help to improve understanding of aspects of the present disclosure. Furthermore, in terms of the construction of the device, one or more components of the device may have been represented in the drawings by conventional symbols, and the drawings may show only those specific details that are pertinent to understanding the embodiments of the present disclosure so as not to obscure the drawings with details that will be readily apparent to those of ordinary skill in the art having benefit of the description herein.
For the purpose of promoting an understanding of the principles of the invention, reference will now be made to the embodiment illustrated in the drawings and specific language will be used to describe the same. It will nevertheless be understood that no limitation of the scope of the invention is thereby intended, such alterations and further modifications in the illustrated system, and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates.
It will be understood by those skilled in the art that the foregoing general description and the following detailed description are exemplary and explanatory of the invention and are not intended to be restrictive thereof.
Reference throughout this specification to "an aspect", "another aspect" or similar language means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. Thus, appearances of the phrase "in an embodiment", "in another embodiment" and similar language throughout this specification may, but do not necessarily, all refer to the same embodiment.
The terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process or method that comprises a list of steps does not include only those steps but may include other steps not expressly listed or inherent to such process or method. Similarly, one or more devices or sub-systems or elements or structures or components proceeded by "comprises...a" does not, without more constraints, preclude the existence of other devices or other sub-systems or other elements or other structures or other components or additional devices or additional sub-systems or additional elements or additional structures or additional components.
Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The system, methods, and examples provided herein are illustrative only and not intended to be limiting.
Embodiments of the present disclosure will be described below in detail with reference to the accompanying drawings.
Referring to Figure 1 illustrates a method for automatic detection of infected erythrocytes based on concavity point identification and pseudo-valley-based thresholding in accordance with an embodiment of the present disclosure. The method 100 for automatic detection of infected erythrocytes based on concavity point identification and pseudo-valley based thresholding comprises the following steps: at step 102, pre-processing a blood smear image, wherein pre-processing comprises: converting the RGB blood smear image to grayscale; normalizing the grayscale blood smear image; and processing the grayscale blood smear image using median filtering technique to remove the noise present in smear images; at step 104, identifying a clump cell in the grayscale blood smear image; at step 106, determining a concavity point in the clump cell, wherein, out of all distances, from points on the original image boundary to the points on the convex hull boundary which are on the same straight line, the distance from the concavity point is the maximum; at step 108, determining a second concavity point, wherein, it will be geometrically opposite to the first concavity point and among all the points opposite to the first concavity point, it will have the shortest distance; at step 110, segmenting the clump cell, wherein the concavity pair points are joined to divide the clump into individual erythrocytes and re-examining the segmented clump; and at step 112, determining a pseudo valley threshold, wherein, this valley is used for segmentation of erythrocytes from a background using Otsu's thresholding.
In an embodiment, the method, wherein, for identifying the clump cell in the grayscale blood smear image, features such as cell area, compactness ratio and aspect ratio are extracted from the image to characterize the clump and isolated cell. k-NN classifier is used for the classification of erythrocytes as isolated and clump cell.
In another embodiment, the method, wherein, re-examining the segmented clump, comprises: calculating an average area of the isolated cells by using all the isolated cells separated from the clump cells; dividing each of the individual clump cell area by the average area of the single cells; separating the cell as isolated if the result is less than 1.5; and identifying the cell as clump cell if the result is more than 1.5 and is further processed to determine a pair of concavity points to split the clump.
In yet another embodiment, the method, wherein, once all the single erythrocytes are segmented from background as well as from the clump cells, individual erythrocytes are examined for possible parasitic infection based on the pseudo-valley concept; For the pseudo valley to exist on a histogram there must be a clear separation between the peaks corresponding to the dominant histogram modes; the pseudo-valley lies within the lower intensity region corresponding to lower pixel population on the gray image histogram and it comprises of three distinct and consecutive intensity values; each of this set of intensity values possesses a distinguishable feature and there also exists a unique relation between them, out of these three consecutive intensity values, the first and third intensities correspond to a larger pixel population than that corresponding to the second intensity; the gradient between the first and second intensity values is equal and negative to the gradient between the second and third intensities; parasite segmentation technique based on the pseudo-valley concept comprises: locating the pseudo-valley and sets the intensity threshold onto that valley; examining each of the single erythrocytes individually with the threshold to detect any pixel within an erythrocyte having intensity value less than or equal to the threshold, if any erythrocyte is found to contain any such pixels, then it is recognized as malarial infected erythrocyte.
Figure 2 illustrates (A) Architecture of erythrocyte segmentation; (B) (a) Boundary of the clump cell (Ib) and (b) boundary of the convex hull of the clump cell (Ic); (C) Distance of the points between Ib and Ic. White: distance from concavity point, Red: distance from the other point; (D) (a) Boundary pixels withO; value within 80° to 90°, (b) Pair of concavity point's determined, (c) Clump cell image, (d) Segmented clump by joining the pair of concavity points; and (E) Microscopic blood smear images (a) RGB image, (b) Grayscale image, (c) Histogram of grayscale image, (d) Extracted erythrocyte, (e) Histogram of the extracted erythrocyte in accordance with an embodiment of the present disclosure.
Pre-Processing: (Fig 2A) The proposed method converts the original RGB image of the thin blood smear into a grayscale image for processing. There is an issue of non-uniform illumination in the microscopic images of the thin blood smears due to staining variability and differences in camera light source. An updated grayworld normalization technique was used to overcome this. To reduce the noise in smear photos, the normalized grayscale image is next processed with the median filtering approach.
Erythrocyte Segmentation: Using Otsu's thresholding methodology, the foreground regions, i.e. erythrocytes, are first segregated from the background. In addition, morphological filtering is used to appropriately recreate the foreground region. The isolated and clump erythrocytes were both visible in the filtered binary image of the blood smear. Figure 2A depicts the overall process of erythrocyte segmentation.
Clump Cell Identification: Clump erythrocytes are identified from isolated erythrocytes using a machine learning methodology since they need to be processed further to be segmented into individual cells. To characterise the clump and isolated cell, the filtered binary image is used to extract parameters such as cell area, compactness ratio, and aspect ratio. In addition, the k-NN classifier is employed to distinguish between solitary and clump cell erythrocytes. The binary filtered image provides quantitative details on the separation of single and clump erythrocytes. In addition, the photos of clump erythrocytes are utilised in the processing.
Concavity Point Determination and Clump Cell Segmentation: There are two steps to determining the pair of concavity points. The boundaries of the clump cell and the convex hull of the clump cell are used to identify the initial concavity point in step 1. The basic idea used here in determining the first concavity point is that "the distance between the concavity point and its corresponding convex hull boundary pixel is the maximum among all the distances between the pixels lying on the same straight line, on the original image boundary to the points on the convex hull boundary." The equivalent pixel on the convex hull boundary that lies on the same straight line as a pixel on the clump cell boundary is characterized as the pixel with the smallest Euclidian distance on the convex hull boundary. The corresponding convex hull pixels, i.e. the pixels located on the same straight line on the convex hull boundary, are determined for all pixels on the cluster cell boundary.
The initial concavity point is the clump cell boundary pixel that corresponds to the maximum distance from its convex hull border pixel. Step 2 determines the second concavity point. The data shows that, in general, the concavity pair points are geometrically opposite each other. The following can be said about the concavity pair point of the first concavity point based on this observation:
(a) It will be on the opposite side of the initial concavity point geometrically.
(b)It will be the least distant pixel among all the locations on the opposite side of the first concavity point.
To meet condition (a), all locations on the clumped image's boundary that are geometrically opposing the first concavity points must be acquired. This is accomplished by establishing an angle criterion: points on the border that lie within a specified bound of angle with respect to the first concavity point are geometrically opposed to it. The pixel that corresponds to the least Euclidian distance from the first concavity point out of this set of geometrically opposing pixels is determined once the set of geometrically opposing pixels is determined, as stated in requirement (b). The second concavity point can be defined as the pixel with the smallest distance from the first concavity point. A additional search for the lowest remote point around the second concavity point is done to improve the robustness of the proposed methodology of locating the second concavity point under varied datasets. The purpose of this is to ensure that any inaccuracy in calculating the geometrically opposite location due to the fixed angle criterion does not result in an incorrect concavity point. On the border pixels, an iterative search is done along both sides of the second concavity point for a given distance.
If a lower distant pixel is found along the search path, the second concavity point is updated, and this new location is labelled as the second concavity point. However, if the initial second concavity point stays the point with the smallest distance from the first concavity point, no further updates are required. The clump cell is divided into two cells using this pair of concavity points. These segmented cells are then re-examined to see if there are any clump cells present. If a clump is identified in any of the segmented cells, the cell is put through an iterative clump cell segmentation process until all of the clumps are segmented into single cells.
Figure 2B illustrates the binary pictures of a clump cell and its corresponding convex hull for determining the first concavity point. The canny edge detection filter is used to determine the boundaries of these images. Figure 2C shows that the maximum distance between the concavity point and the point on the convex-hull boundary that is on the same straight line is greater than any other distance. All the points on the convex hull boundary that are on the same x-axis and y-axis as the clumped cell boundary are determined for each point on the clumped cell boundary. The minimum of the x-axis and y-axis distances from the convex-hull boundary is regarded the distance of that pixel from the convex-hull boundary for any pixel on the cluster cell boundary. That is, the distance from a point on the same straight line on the convex-hull border is considered. The process is performed for all the co-ordinates on the binary image of the clumped erythrocyte's boundaries. After all the distances have been calculated, the biggest of all of them is determined. The point on the clumped-image boundary that corresponds to the greatest distance is the first concavity point, which is point 'A' in figure 2C. As a result, point 'A' is designated as the first concavity point.
Determining the Second Concavity Point: The first stage in determining the second concavity point is to find all the pixels on the clumped image's boundaries that are geometrically opposite to 'A.' This is accomplished by establishing an angle criterion. The underlying premise is that the points on the boundary that lie within a specific bound of angle with regard to point 'A' are the geometrically opposite points to 'A.' Initially, a neighbour point on the boundary at a specific pixel distance from 'A' is identified, and the angle in between lines linking 'A (xA, yA)' and 'A (xA + 5, yA + 5)' and joining 'A' and every pixel on the boundary is determined, retaining 'A (x, y)' as the vertex. The angles that meet the angle boundary criterion are filtered out, and the points that correspond to those angles are referred to as the points opposite to 'A.' Out of these points, the point with the least distance is to be found. The point opposite to 'A', which may or may not be the second concavity point, is designated as 'AO.' As a result, an iterative search around 'AO' is carried out to determine the exact concavity pair point of 'A,' with the goal of determining the shortest possible distance from 'A' around 'AO' including 'AO'. Two iterative search methods have been used here, one to the right of 'AO' and the other to the left, with a defined number of increments and decrements in the index of 'AO' across the set of border pixels, respectively. With the concavity pair points in hand, the clump can now be divided into individual erythrocytes. This is shown in Figure 2D(c) and 2D(d).
Using all of the isolated cells separated from the clump cells, the average area of the isolated cells is computed. The average area of the single cells is then divided by the area of each individual clump cell. If the result is less than 1.5, the cell is considered isolated and separated. Otherwise, the cell is classified as a clump cell and is processed further to find the pair of concavity sites that will be used to divide the clump. The reason for choosing 1.5 as the threshold for this decision is that pathologists have observed and confirmed that an infected cell in its advanced stage may inflate and grow larger in size than a healthy red blood cell, but rarely matches the size of a clump cell, which can even be formed by two single cells. If the object is determined to be a clump cell, it is processed once again for additional clump cell splitting; otherwise, it is regarded as a single cell. This method is repeated until each clump split object is identified as a single erythrocyte cell.
Once, all of the single erythrocytes have been segmented from the background as well as the clump cells, the individual erythrocytes are evaluated for suspected parasite infection using the pseudo-valley concept. One of the most fundamental approaches for segmenting parasites in erythrocytes is grey thresholding. However, intensity thresholding segmentation necessitates a valley in the intensity histogram in order to set the threshold onto this valley. The grayscale image (Figure 2E(b)) of the microscopic blood smear image of Figure 2E has a valley, as seen in Figure 2E(c) (a). Using Otsu's thresholding, this valley is utilized to segregate erythrocytes from the backdrop. There must also be a distinct separation between the peaks corresponding to the main histogram modes for a valley to appear on a histogram. The width and depth of the valley separating the histogram modes determine the success of intensity thresholding.
To avoid this error, a closer look at the parasitic image's histogram revealed that there is an infinitesimally small valley-like area between the parasite and non-parasitic parts. This valley-like area does not occur in non-parasitic photos, i.e. photos in which the malarial parasites are not present. This area does not quite resemble a valley in its real form, but it is sufficient to differentiate infected parasite pixels from non-infected pixels, thus the name "pseudo-valley." The development of rings in infected cells during the early stages of malaria is thought to be the major cause of the pseudo-valley in the infected cells' histogram.
The malarial infection in the erythrocyte starts with the development of a ring (trophozoite) that contains a chromatin dot. This chromatin dot has the lowest pixel intensity of all the other parts of the erythrocyte as well as the rest of the ring. The intensity differential between the centre of the chromatin dot and the area surrounding it is unique. Within an infected erythrocyte, the area surrounding the centre of the chromatin has the darkest intensity, whereas the core portion has a higher intensity. This property of the chromatin dot is maintained throughout the malarial parasite's life cycle within an erythrocyte. On the grayscale histogram of an infected cell, the distribution of intensities (between lower and higher intensity) inside this chromatin area has a distinct characteristic.
There are no pixels on the lower intensity portion of the histogram in the non-infected histogram. The infected cell's histogram has a non-zero number of pixels in its lower intensity zone, which signifies infected pixels, especially within the chromatin region. Extensive experimental study has revealed that there is an intensity shift between the region corresponding to the chromatin dot's centre and the darker region around the dot's centre. This intensity transition boundary has a distinguishing feature. The number of pixels in the darker region of the chromatin dot is found to be identical to the number of pixels in the adjacent higher intensity value that indicates the central region at that transition boundary.
There could be multiple transition points or borders on the chromatin dot. The presence of a chromatin dot, and thus the presence of a parasite within a cell, is determined by such a barrier. A threshold must be set precisely onto the bottom portion of the histogram to segment the core region and the surrounding darker region within the chromatin dot in order to define such a transition boundary.
The drawings and the forgoing description give examples of embodiments. Those skilled in the art will appreciate that one or more of the described elements may well be combined into a single functional element. Alternatively, certain elements may be split into multiple functional elements. Elements from one embodiment may be added to another embodiment. For example, orders of processes described herein may be changed and are not limited to the manner described herein. Moreover, the actions of any flow diagram need not be implemented in the order shown; nor do all of the acts necessarily need to be performed. Also, those acts that are not dependent on other acts may be performed in parallel with the other acts. The scope of embodiments is by no means limited by these specific examples. Numerous variations, whether explicitly given in the specification or not, such as differences in structure, dimension, and use of material, are possible. The scope of embodiments is at least as broad as given by the following claims.
Benefits, other advantages, and solutions to problems have been described above with regard to specific embodiments. However, the benefits, advantages, solutions to problems, and any component(s) that may cause any benefit, advantage, or solution to occur or become more pronounced are not to be construed as a critical, required, or essential feature or component of any or all the claims.
Claims (10)
1. A method for automatic detection of infected erythrocytes based on concavity point identification and pseudo-valley-based thresholding, the method comprises: pre-processing a blood smear image, wherein pre-processing comprises: converting the RGB blood smear image to grayscale; normalizing the grayscale blood smear image; and processing the grayscale blood smear image using median filtering technique to remove the noise present in smear images;
identifying a clump cell in the grayscale blood smear image;
determining a concavity point in the clump cell, wherein, out of all distances, from points on the original image boundary to the points on the convex hull boundary which are on the same straight line, the distance from the concavity point is the maximum;
determining a second concavity point, wherein, it will be geometrically opposite to the first concavity point and among all the points opposite to the first concavity point, it will have the shortest distance;
segmenting the clump cell, wherein the concavity pair points are joined to divide the clump into individual erythrocytes and re-examining the segmented clump; and
determining a pseudo valley threshold, wherein, this valley is used for segmentation of erythrocytes from a background using Otsu's thresholding.
2. The method as claimed in claim 1, wherein, for identifying the clump cell in the grayscale blood smear image, features such as cell area, compactness ratio and aspect ratio are extracted from the image to characterize the clump and isolated cell.
3. The method as claimed in claim 2, wherein, k-NN classifier is used for the classification of erythrocytes as isolated and clump cell.
4. The method as claimed in claim 1, wherein, re-examining the segmented clump, comprises: calculating an average area of the isolated cells by using all the isolated cells separated from the clump cells; dividing each of the individual clump cell area by the average area of the single cells; separating the cell as isolated if the result is less than 1.5; and identifying the cell as clump cell if the result is more than 1.5 and is further processed to determine a pair of concavity points to split the clump.
5. The method as claimed in claim 1, wherein, once all the single erythrocytes are segmented from background as well as from the clump cells, individual erythrocytes are examined for possible parasitic infection based on the pseudo-valley concept.
6. The method as claimed in claim 1, wherein, for the pseudo valley to exist on a histogram there must be a clear separation between the peaks corresponding to the dominant histogram modes.
7. The method as claimed in claim 1, wherein, the pseudo-valley lies within the lower intensity region corresponding to lower pixel population on the gray image histogram and it comprises of three distinct and consecutive intensity values.
8. The method as claimed in claim 5, wherein, each of this set of intensity values possesses a distinguishable feature and there also exists a unique relation between them, out of these three consecutive intensity values, the first and third intensities correspond to a larger pixel population than that corresponding to the second intensity.
9. The method as claimed in claim 5, wherein, the gradient between the first and second intensity values is equal and negative to the gradient between the second and third intensities.
10. The method as claimed in claim 1, wherein, parasite segmentation technique based on the pseudo-valley concept comprises:
locating the pseudo-valley and sets the intensity threshold onto that valley; and examining each of the single erythrocytes individually with the threshold to detect any pixel within an erythrocyte having intensity value less than or equal to the threshold, if any erythrocyte is found to contain any such pixels, then it is recognized as malarial infected erythrocyte.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| AU2021103689A AU2021103689A4 (en) | 2021-06-29 | 2021-06-29 | A Method for Automatic Detection of Infected Erythrocytes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| AU2021103689A AU2021103689A4 (en) | 2021-06-29 | 2021-06-29 | A Method for Automatic Detection of Infected Erythrocytes |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| AU2021103689A4 true AU2021103689A4 (en) | 2021-08-26 |
Family
ID=77369661
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| AU2021103689A Ceased AU2021103689A4 (en) | 2021-06-29 | 2021-06-29 | A Method for Automatic Detection of Infected Erythrocytes |
Country Status (1)
| Country | Link |
|---|---|
| AU (1) | AU2021103689A4 (en) |
-
2021
- 2021-06-29 AU AU2021103689A patent/AU2021103689A4/en not_active Ceased
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liao et al. | An accurate segmentation method for white blood cell images | |
| Díaz et al. | A semi-automatic method for quantification and classification of erythrocytes infected with malaria parasites in microscopic images | |
| CN102682305B (en) | Automatic screening system and automatic screening method using thin-prep cytology test | |
| EP2719754B1 (en) | Image processing apparatus, image processing method and image processing system | |
| US8331642B2 (en) | Method and device for determining a contour and a center of an object | |
| Savkare et al. | Automatic system for classification of erythrocytes infected with malaria and identification of parasite's life stage | |
| Ramoser et al. | Leukocyte segmentation and classification in blood-smear images | |
| Shafique et al. | Computer-assisted acute lymphoblastic leukemia detection and diagnosis | |
| Ahirwar et al. | Advanced image analysis based system for automatic detection and classification of malarial parasite in blood images | |
| US9196036B2 (en) | Device and method for determining objects in a color recording | |
| Mohammed et al. | Detection and classification of malaria in thin blood slide images | |
| Razzak | Automatic detection and classification of malarial parasite | |
| WO1997026617A1 (en) | Classifying red blood cells using imaging techniques to analyze color and shape | |
| Molina et al. | Sequential classification system for recognition of malaria infection using peripheral blood cell images | |
| WO1996009594A1 (en) | Apparatus for automated identification of thick cell groupings on a biological specimen | |
| Memeu et al. | Detection of plasmodium parasites from images of thin blood smears | |
| Percannella et al. | A classification-based approach to segment HEp-2 cells | |
| Panchbhai et al. | RBCs and parasites segmentation from thin smear blood cell images | |
| Savkare et al. | Automatic classification of normal and infected blood cells for parasitemia detection | |
| Ghane et al. | Classification of chronic myeloid leukemia cell subtypes based on microscopic image analysis | |
| CN114037868B (en) | Image recognition model generation method and device | |
| Varma et al. | Detection of malaria parasite based on thick and thin blood smear images using local binary pattern | |
| GB2329014A (en) | Automated identification of tubercle bacilli | |
| Mustafa et al. | Malaria parasite diagnosis using computational techniques: a comprehensive review | |
| Aris et al. | Fast k-means clustering algorithm for malaria detection in thick blood smear |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FGI | Letters patent sealed or granted (innovation patent) | ||
| MK22 | Patent ceased section 143a(d), or expired - non payment of renewal fee or expiry |