WO2022124832A1 - Method for determining cell zone of sample image of slide smeared with bone marrow, and method for imaging cell zone at high magnification - Google Patents

Method for determining cell zone of sample image of slide smeared with bone marrow, and method for imaging cell zone at high magnification Download PDF

Info

Publication number
WO2022124832A1
WO2022124832A1 PCT/KR2021/018680 KR2021018680W WO2022124832A1 WO 2022124832 A1 WO2022124832 A1 WO 2022124832A1 KR 2021018680 W KR2021018680 W KR 2021018680W WO 2022124832 A1 WO2022124832 A1 WO 2022124832A1
Authority
WO
WIPO (PCT)
Prior art keywords
zone
cell
particle
image
bone marrow
Prior art date
Application number
PCT/KR2021/018680
Other languages
French (fr)
Korean (ko)
Inventor
최종호
정보원
이주선
이봉기
이중훈
이영득
Original Assignee
(주)유아이엠디
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 (주)유아이엠디 filed Critical (주)유아이엠디
Publication of WO2022124832A1 publication Critical patent/WO2022124832A1/en

Links

Images

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
    • G06V20/695—Preprocessing, e.g. image segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00—Geometric image transformations in the plane of the image
    • G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/90—Dynamic range modification of images or parts thereof
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/54—Extraction of image or video features relating to texture
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/56—Extraction of image or video features relating to colour
    • 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

Definitions

  • the present invention relates to a method for automatic detection of cell zones or determination of cell zones through image processing and analysis of sample images for bone marrow readout.
  • Bone marrow is a soft tissue located in the inner space of bones and is a hematopoietic organ that produces blood cells such as red blood cells, white blood cells, and platelets. Also, if there is a problem in the production of blood cells in the human body, bone marrow aspirate and biopsy the bone marrow from a large bone, make a slide, and if necessary, stain it. It is a test to check for abnormal lesions. These bone marrow tests are used for diagnosis based on normal or abnormal white blood cell counts, whether there are abnormal white blood cells that should not be present, or as data for diagnosing other leukemias.
  • This bone marrow examination or bone marrow reading process takes a long time and accompanies the fatigue of the diagnostician (or analyst) because various parts of the slide must be observed at high magnification. Moreover, the manual operation of finding all the white blood cells present in the bone marrow sample image, determining the type of each white blood cell, and counting the number of each type of white blood cell takes a lot of time and is highly likely to have low accuracy depending on the experience level of the diagnostician.
  • 'Cell Zone' means a large number of single cells ideal for viewing and reading slide specimens at high magnification. That is, the 'cell zone' can be defined as a limited number of single cells with high reliability that can ensure the accuracy of bone marrow reading.
  • Bone marrow reading utilizes sample images obtained from bone marrow specimen slides that have undergone the smearing and staining process, and all areas of the slide specimen are reliable due to differences in the spread thickness and uniformity of the specimen according to the smear and the inability to be stained uniformly. It cannot be an area. Therefore, when all areas are randomly observed at high magnification, the probability of reading failure is high, and repeated reading failures inevitably lead to reduced work efficiency and long hours of labor. Therefore, diagnostic testers with high experience and skill are observing and reading the aforementioned 'cell zone', that is, a limited number of reliable single cells.
  • the conventional 'cell zone' selection method has a disadvantage in that it is difficult to achieve consistency because it involves the subjective judgment of a person, and may lead to different reading results depending on the diagnostician.
  • the present invention has been derived from the above needs, and a first object of the present invention is to automatically detect a highly reliable cell zone in a slide sample image prepared for bone marrow reading and to detect a highly reliable cell zone. to provide a method for determining the cell zone of a plated slide sample image.
  • a second object of the present invention is to provide a method for determining the cell zone of a slide sample image on which the bone marrow is smeared, which can increase speed and accuracy in determining the cell zone in the slide sample image prepared for bone marrow reading.
  • a third object of the present invention is to provide a fast and efficient high-magnification imaging method for a determined cell zone.
  • a confidence zone selection step (S110) of selecting a confidence zone (Confidence Zone) from which the deep zone with thick staining is removed from the slide sample image on which the bone marrow is smeared in one direction; a background subtraction step 120 of removing a background and extracting a cell image based on the color and texture characteristics of the cell image in the selected confidence zone; a particle zone candidate selection step 130 of selecting a plurality of particle candidates based on the integral image method in the confidence zone from which the background is removed; a particle zone correction step (S140) of correcting overlapping particle candidates among a plurality of selected particle candidates into particle zones based on color properties (color, saturation, and brightness); and a cell zone determining step (S150) of determining a cell zone in an area in which the cell population of the extracted cell image is higher than a preset standard or in an area spaced apart by a preset distance in the opposite direction to the smear based on the corrected particle zone
  • the smeared slide sample image may be scaled down and adjusted.
  • the selected confidence zone may be highlighted in which an effective pixel corresponding to an effective area is extracted.
  • the method for determining the cell zone of the slide sample image on which the bone marrow is smeared includes a white balancing step of performing color equalization on effective pixels between the confidence zone selection step (S110) and the background subtraction step (S120) (S112); and a color temperature correction step (S114) of correcting the color temperature of the effective pixel.
  • color and brightness are normalized for effective pixels whose color temperature has been corrected. It may further include a normalization step (S116) to perform.
  • the background subtraction step ( S120 ) may be performed in a manner that primarily removes a blurred area, an area having a brightness equal to or less than a certain level, and an area having a color equal to or less than a certain level.
  • the cell zone determining step S150 may be a step of determining a cell zone by removing the overlapping area of the corrected particle zone from the determined cell zone when the corrected particle zone and the determined cell zone overlap.
  • an object of the present invention is to perform high-magnification imaging at N ⁇ N times in an N ⁇ N mask with respect to the cell zone determined by the cell zone determination method described above, and one autofocusing result performed at the center point of the N ⁇ N mask This can be achieved by providing a high-magnification imaging method of a cell zone characterized by use at different points in an NxN mask. [where N is an odd number greater than or equal to 3]
  • FIG. 1 is a flowchart sequentially illustrating a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 2 is a diagram illustrating an example of resizing among a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 3 is a diagram illustrating an example of selecting a confidence zone (or a confidence map) in a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 4 is a view showing a state in which effective pixels are extracted through highlighting for the confidence zone selected in FIG. 3;
  • FIG. 5 is a flowchart illustrating a series of image processing processes for a confidence zone in a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 6 is a diagram illustrating an example of a confidence zone before (a) and after (b) auto white balancing in a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 7 is a view showing an example of a confidence zone after color temperature correction in a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 8 is a view showing examples of confidence zones before (a) and after (b) color-brightness normalization in a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 9 is a view showing an example (a) and a partial enlarged photograph (b) of a confidence zone after background subtraction in a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 10 is a view showing an example of a confidence zone in which a particle candidate group (overlapping square box) is displayed through an integral image method among the cell zone determination methods according to an embodiment of the present invention
  • FIG. 11 is a view showing an example of a confidence zone in which particle candidates (non-overlapping rectangular boxes) corrected through a Non-Maxima Suppression Method among the cell zone determination method according to an embodiment of the present invention are displayed;
  • FIG. ego
  • FIG. 12 is a view showing a slide sample image in which a cell zone determined according to a method for determining a cell zone according to an embodiment of the present invention is displayed;
  • FIG. 13 is a view showing a state in which particles and cell zones overlap (a) and a cell zone (b) of an effective area serving as a high magnification imaging area in a method for determining a cell zone according to an embodiment of the present invention
  • FIG. 14 is a view for explaining auto-focusing for high magnification imaging with respect to a cell zone determined according to a method for determining a cell zone according to an embodiment of the present invention.
  • FIG. 1 is a flowchart sequentially illustrating a method for determining a cell zone according to an embodiment of the present invention.
  • an embodiment of the present invention selects a confidence zone from a slide sample image (S110), background subtraction (S120), particle zone candidate selection (S130), particle zone correction (S140) and cell zone determination ( S150).
  • high-magnification imaging can be performed by determining a cell zone on the slide sample image on which the bone marrow is smeared, so that the diagnostician can save the effort of separately finding the cell zone.
  • the confidence zone (Confidence Zone or Confidence Map) in which the deep zone with thick staining is removed from the slide sample image on which the bone marrow is smeared in one direction.
  • a zone selection step (S110) is performed.
  • the confidence zone is a highly reliable area with a high probability of distribution of a cell zone, which is an ideal area of a large number of single cells suitable for observing and reading a slide specimen at high magnification.
  • a confidence zone mapped to a mask shape may be selected based on light characteristics such as saturation characteristics of light passing through the slide or transmittance of light in a specific wavelength band.
  • the step (S110) of selecting a confidence zone in the reduced image after downscaling is performed.
  • the confidence zone selected in the confidence zone selection step S110 is preferably a confidence zone filled with effective pixels by performing highlighting for extracting only effective pixels corresponding to the effective area.
  • the effective area corresponds to the part on the slide that is stained, that is, the part with cells, and is an area in which the saturated area that is not stained when light is transmitted through the slide is removed.
  • FIG. 5 is a flowchart illustrating a series of image processing processes for a confidence zone in a method for determining a cell zone according to an embodiment of the present invention.
  • a white balancing step Auto White Balancing; S112 of performing color equalization on effective pixels and a color temperature correction step of correcting color temperature of effective pixels (Color Temperature Correctness; S114), and A normalization step (Color-Brightness Normalization; S116 ) of performing normalization of color and brightness with respect to an effective pixel whose color temperature has been corrected may be sequentially performed.
  • a series of steps such as S112, S114, and S116 is an image preprocessing process for computer vision recognition. Due to the nature of the slide sample, the slide sample image cannot always have a uniform color and uniform brightness, so it is desirable to obtain and use a uniform image through this pre-processing process.
  • the white balancing step (S112) as shown in FIG. 6 , the color balance of the image as a whole is performed.
  • AWB Auto White Balancing
  • the point having color properties of (254, 254, 245) and standard deviation (SD) of 4.242 is shown in Fig. 6 ( As in b), the color properties were (246, 245, 244) and the standard deviation (SD) was 0.816.
  • a specific color temperature is corrected using a Basic LED Reference Example and a corresponding Kelvin Color Temperature Scale Chart. For example, in this embodiment, as shown in FIG. 7 , 3800 K correction was performed.
  • the color-brightness normalization step S116 is a step of lowering or increasing the brightness to a specific brightness so that an even brightness can be always maintained. For example, in the present embodiment, intensity (intensity) is improved by 3% in (b) compared to (a) shown in FIG. 8 .
  • the background subtraction step 120 is a process of removing the background and extracting the cell image based on the color and texture characteristics of the cell image in the selected confidence zone.
  • the background subtraction step ( S120 ) may be performed in a manner that primarily removes a blurred area, an area having a brightness below a certain level, and an area having a color below a certain level. Through this background subtraction, color features (color and texture of the cell image) corresponding to the cell can be extracted. After the background subtraction step ( S120 ), a morphology of correcting image breakage by using neighboring pixel information may be performed.
  • the particle zone candidate selection step 130 is a process of selecting a plurality of particle candidates based on the integral image method in the confidence zone from which the background is removed.
  • the particles which are substances obtained from bone marrow samples, are not mostly normal cells, and generally appear in a shape large enough to be identified with the naked eye.
  • the particle zone correction step ( S140 ) is a process of correcting the particle candidates that overlap each other among a plurality of selected particle candidates as a particle zone based on color properties (color, saturation, and brightness). Correction may be performed by absorbing duplicate particle candidates into a single particle zone having a large value through the Non-Maxima Suppression Method. Therefore, the corrected particle zones are extracted in a form that does not overlap with each other.
  • the cell zone determination step S150 is a process of determining a cell zone in an area spaced apart by a preset distance in the opposite direction to the smearing based on the corrected particle zone.
  • a region in which the cell population of the extracted cell image is higher than a preset standard or a region in which a large number of independent single cells are distributed may be determined as the cell zone.
  • a particle zone is assumed for the unification of the cell zone determination method, and an area spaced apart by a preset distance in the direction opposite to the smearing based on the assumed particle zone may be determined as the cell zone.
  • the cell zone determining step S150 may be a step of determining a cell zone by removing the overlapping area of the corrected particle zone from the determined cell zone when the corrected particle zone and the determined cell zone overlap. By removing the overlapping area of the particle zone, the cell zone can be determined more accurately and the amount of computation can be reduced, which helps in quick and accurate diagnosis.
  • FIG. 14 is a view for explaining auto-focusing for high magnification imaging with respect to a cell zone determined according to a method for determining a cell zone according to an embodiment of the present invention.
  • high magnification imaging for diagnosis is performed on the cell zone determined by the method for determining the cell zone. It is preferable to image the determined cell zone at high magnification at N ⁇ N times in an N ⁇ N mask, and use one autofocusing result performed at the center point of the N ⁇ N mask at another point in the N ⁇ N mask. By doing so, it is possible to reduce the amount of computation due to repeated autofocusing (where N is an odd number equal to or greater than 3).

Landscapes

  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Molecular Biology (AREA)
  • Quality & Reliability (AREA)
  • Radiology & Medical Imaging (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Medical Informatics (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

According to one embodiment of the present invention, provided is a method for determining a cell zone of a sample image of a slide on which bone marrow is smeared, the method comprising: a confidence zone selection step (S110) of selecting a confidence zone from which a deep zone with thick staining is removed from the sample image of the slide on which bone marrow is smeared in one direction; a background subtraction step (S120) of removing a background based on color and texture characteristics of the cell image in the selected confidence zone and extracting a cell image; a particle zone candidate selection step (130) of selecting a plurality of particle candidates based on an integral image method in the confidence zone from which the background is removed; a particle zone correction step (S140) of correcting overlapping particle candidates among the plurality of selected particle candidates into particle zones based on color attributes (color, saturation, and brightness); and a cell zone determination step (S150) of determining a cell zone in an area in which the cell population of the extracted cell image is higher than a preset standard or in an area spaced apart by a preset distance in the opposite direction to the smear based on the corrected particle zones.

Description

골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법 및 셀 존의 고배율 촬상 방법A method for determining the cell zone of a slide sample image on which bone marrow is smeared and a method for imaging the cell zone at high magnification
본 발명은 골수 판독을 위한 샘플 이미지의 이미지 처리 및 분석을 통해 셀 존의 자동 검출 또는 셀 존 결정 방법에 관한 것이다.The present invention relates to a method for automatic detection of cell zones or determination of cell zones through image processing and analysis of sample images for bone marrow readout.
골수는 뼈의 안쪽 공간에 위치한 부드러운 조직으로 적혈구, 백혈구, 혈소판과 같은 혈액세포를 생성하는 조혈기관이다. 또한 골수 검사는 인체에서 혈구 생성에 문제가 생긴 경우 큰 뼈에서 골수를 흡인하고 생검하여 슬라이드로 제작하고 필요한 경우에는 염색한 후, 현미경을 통해 골수 세포를 관찰하고 감별 계산을 수행함으로써 골수의 기능 및 비정상적인 병변을 확인하는 검사이다. 이러한 골수 검사는 정상백혈구 수치나 비정상 백혈구 수치를 통한 진단, 비정상 백혈구 중 있어서는 안되는 세포가 있는지 여부 또는 기타 백혈병 진단의 자료로 이용된다.Bone marrow is a soft tissue located in the inner space of bones and is a hematopoietic organ that produces blood cells such as red blood cells, white blood cells, and platelets. Also, if there is a problem in the production of blood cells in the human body, bone marrow aspirate and biopsy the bone marrow from a large bone, make a slide, and if necessary, stain it. It is a test to check for abnormal lesions. These bone marrow tests are used for diagnosis based on normal or abnormal white blood cell counts, whether there are abnormal white blood cells that should not be present, or as data for diagnosing other leukemias.
이러한 골수 검사 또는 골수 판독 과정은 슬라이드의 여러 부위를 고배율로 관찰해야 하므로 장시간이 소요되고 진단검사의(또는 분석가)의 피로가 수반되는 작업이다. 더군다나 골수 표본 이미지에 존재하는 모든 백혈구를 찾아서 각각의 백혈구의 종류를 판단하고 각 종류의 백혈구의 수를 세는 수작업은 많은 시간이 소요될 뿐만 아니라 진단검사의의 경험 정도에 따라 정확도가 낮을 가능성도 크다.This bone marrow examination or bone marrow reading process takes a long time and accompanies the fatigue of the diagnostician (or analyst) because various parts of the slide must be observed at high magnification. Moreover, the manual operation of finding all the white blood cells present in the bone marrow sample image, determining the type of each white blood cell, and counting the number of each type of white blood cell takes a lot of time and is highly likely to have low accuracy depending on the experience level of the diagnostician.
한편 진단검사의는 골수 판독의 효율성 및 정확도를 높이기 위해 골수 표본 관찰을 위한 '셀 존(Cell-Zone)'을 설정한다. '셀 존'이란 슬라이드 표본을 고배율로 관찰하고 판독하기에 적합한 이상적인 다수의 싱글 세포를 의미한다. 즉 '셀 존'은 골수 판독의 정확성을 기할 수 있는 신뢰도 높은 한정된 개수의 싱글 세포로 정의될 수 있다.Meanwhile, the diagnostician sets up a 'Cell-Zone' for bone marrow specimen observation in order to increase the efficiency and accuracy of bone marrow reading. 'Cell Zone' means a large number of single cells ideal for viewing and reading slide specimens at high magnification. That is, the 'cell zone' can be defined as a limited number of single cells with high reliability that can ensure the accuracy of bone marrow reading.
골수 판독은 도말 및 염색 과정을 거친 골수 표본 슬라이드로부터 획득된 샘플 이미지를 활용하는데, 도말에 따른 표본의 펼쳐지는 두께 및 균일도가 다르고 단일하게 염색되지 못하는 등의 이유로 슬라이드 표본의 모든 영역이 신뢰할 수 있는 영역이 될 수는 없다. 따라서 모든 영역을 임의대로 고배율 관찰하는 경우 판독에 실패할 확률이 높고 반복되는 판독 실패는 작업 효율 저하와 장시간의 노동으로 귀결될 수밖에 없다. 따라서 경험 및 숙련도가 높은 진단 검사의들이 전술한 '셀 존', 즉 신뢰할 수 있는 한정된 개수의 싱글 세포에 대하여 관찰 및 판독을 수행하고 있는 실정이다.Bone marrow reading utilizes sample images obtained from bone marrow specimen slides that have undergone the smearing and staining process, and all areas of the slide specimen are reliable due to differences in the spread thickness and uniformity of the specimen according to the smear and the inability to be stained uniformly. It cannot be an area. Therefore, when all areas are randomly observed at high magnification, the probability of reading failure is high, and repeated reading failures inevitably lead to reduced work efficiency and long hours of labor. Therefore, diagnostic testers with high experience and skill are observing and reading the aforementioned 'cell zone', that is, a limited number of reliable single cells.
그러나 종래 수작업에 의한 '셀 존' 선정 및 이에 따른 골수 판독은 경험 및 숙련도가 낮은 진단검사의들에게는 작업 효율 저하와 장시간 노동이 수반된다는 문제점은 여전히 존재한다. 도제 교육에 따라 숙련도 낮은 진단검사의들의 수준을 높이는 방법을 이용하고 있지만 많은 시간의 소요와 시행 착오가 수반된다.However, the conventional manual selection of a 'cell zone' and thus bone marrow reading still has problems in that it entails reduced work efficiency and long hours of labor for diagnosticians with low experience and skill. Although the method of raising the level of low-skilled diagnosticians according to apprenticeship education is used, it takes a lot of time and involves trial and error.
더욱이 경험 및 숙련도가 높은 진단검사의라 하더라도 종래 '셀 존'의 선정 방법은 사람의 주관적인 판단이 개입되므로 일관성을 갖기가 힘든 단점이 있으며, 진단검사의에 따라 다른 판독 결과로 이어질 수 있다.Moreover, even for a diagnostician with high experience and skill, the conventional 'cell zone' selection method has a disadvantage in that it is difficult to achieve consistency because it involves the subjective judgment of a person, and may lead to different reading results depending on the diagnostician.
따라서 골수 표본 슬라이드의 이미지 처리 및 분석을 통해 셀 존의 자동 검출 및 셀 존 결정에 대한 연구의 필요성이 대두된다.Therefore, there is a need for research on automatic detection of cell zones and determination of cell zones through image processing and analysis of bone marrow specimen slides.
본 발명은 상기와 같은 필요성에 기하여 도출된 것으로서, 본 발명의 제1 목적은, 골수 판독을 위해 준비된 슬라이드 샘플 이미지에서 신뢰성이 높은 셀 존의 자동 검출과 신뢰성 높은 셀 존을 디텍팅할 수 있는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법을 제공하는 데 있다.The present invention has been derived from the above needs, and a first object of the present invention is to automatically detect a highly reliable cell zone in a slide sample image prepared for bone marrow reading and to detect a highly reliable cell zone. to provide a method for determining the cell zone of a plated slide sample image.
본 발명의 제2 목적은, 골수 판독을 위해 준비된 슬라이드 샘플 이미지에서 셀 존을 결정함에 있어서 신속성 및 정확성을 높일 수 있는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법을 제공하는 데 있다.A second object of the present invention is to provide a method for determining the cell zone of a slide sample image on which the bone marrow is smeared, which can increase speed and accuracy in determining the cell zone in the slide sample image prepared for bone marrow reading.
본 발명의 제3 목적은, 결정된 셀 존에 대한 신속하고 효율적인 고배율 촬상 방법을 제공하는 데 있다.A third object of the present invention is to provide a fast and efficient high-magnification imaging method for a determined cell zone.
상기와 같은 본 발명의 목적은, 일 방향으로 골수가 도말된 슬라이드 샘플 이미지에서 염색이 두꺼운 딥 존(Deep Zone)이 제거된 컨피던스 존(Confidence Zone)을 선정하는 컨피던스 존 선정단계(S110); 선정된 컨피던스 존에서 세포 이미지의 컬러와 질감 특징에 기반하여 백그라운드를 제거하고 세포 이미지를 추출하는 백그라운드 서브트랙션 단계(120); 백그라운드가 제거된 컨피던스 존에서 적분 이미지 방식에 기반하여 다수의 파티클 후보를 선정하는 파티클 존 후보 선정단계(130); 선정된 다수의 파티클 후보 중 상호 오버랩되는 파티클 후보에 대하여 컬러 속성(컬러, 채도 및 밝기)을 기준으로 파티클 존으로 보정하는 파티클 존 보정단계(S140); 및 추출된 세포 이미지의 세포 군집도가 기 설정된 기준보다 높은 영역 또는 보정된 파티클 존을 기준으로 도말 반대 방향으로 기 설정된 거리만큼 이격된 영역에서 셀 존을 결정하는 셀 존 결정단계(S150)를 포함하는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법을 제공함으로써 달성될 수 있다.The object of the present invention as described above, a confidence zone selection step (S110) of selecting a confidence zone (Confidence Zone) from which the deep zone with thick staining is removed from the slide sample image on which the bone marrow is smeared in one direction; a background subtraction step 120 of removing a background and extracting a cell image based on the color and texture characteristics of the cell image in the selected confidence zone; a particle zone candidate selection step 130 of selecting a plurality of particle candidates based on the integral image method in the confidence zone from which the background is removed; a particle zone correction step (S140) of correcting overlapping particle candidates among a plurality of selected particle candidates into particle zones based on color properties (color, saturation, and brightness); and a cell zone determining step (S150) of determining a cell zone in an area in which the cell population of the extracted cell image is higher than a preset standard or in an area spaced apart by a preset distance in the opposite direction to the smear based on the corrected particle zone This can be achieved by providing a method for determining the cell zone of a slide sample image on which the bone marrow is plated.
컨피던스 존 선정단계(S110)에서, 도말된 슬라이드 샘플 이미지는 스케일이 다운되어 조정된 것일 수 있다. 그리고 컨피던스 존 선정단계(S110)에서, 선정된 컨피던스 존은 유효 영역에 대응하는 유효 픽셀이 추출되는 하이라이팅이 수행된 것일 수 있다.In the confidence zone selection step (S110), the smeared slide sample image may be scaled down and adjusted. In addition, in the confidence zone selection step ( S110 ), the selected confidence zone may be highlighted in which an effective pixel corresponding to an effective area is extracted.
골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법은, 컨피던스 존 선정단계(S110)와 백그라운드 서브스트랙션 단계(S120) 사이에, 유효 픽셀에 대하여 컬러 균등화를 수행하는 화이트 밸런싱 단계(S112); 및 유효 픽셀에 대하여 컬러 온도를 보정하는 컬러 온도 보정단계(S114)를 더 포함할 수 있다.The method for determining the cell zone of the slide sample image on which the bone marrow is smeared includes a white balancing step of performing color equalization on effective pixels between the confidence zone selection step (S110) and the background subtraction step (S120) (S112); and a color temperature correction step (S114) of correcting the color temperature of the effective pixel.
또한 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법은, 컬러 온도 보정단계(S114)와 백그라운드 서브스트랙션 단계(S120) 사이에, 컬러 온도가 보정된 유효 픽셀에 대하여 컬러와 브라이트니스에 대한 정규화를 수행하는 정규화 단계(S116)를 더 포함할 수 있다.In addition, in the cell zone determination method of the bone marrow-smeared slide sample image, between the color temperature correction step ( S114 ) and the background subtraction step ( S120 ), color and brightness are normalized for effective pixels whose color temperature has been corrected. It may further include a normalization step (S116) to perform.
아울러 백그라운드 서브스트랙션 단계(S120)는 1차적으로 블러된 영역과, 밝기가 일정 이하인 영역, 색감이 일정 이하인 영역을 제거하는 방식으로 진행될 수 있다.In addition, the background subtraction step ( S120 ) may be performed in a manner that primarily removes a blurred area, an area having a brightness equal to or less than a certain level, and an area having a color equal to or less than a certain level.
셀 존 결정단계(S150)는, 보정된 파티클 존 및 결정된 셀 존이 오버랩되는 경우, 결정된 셀 존에서 보정된 파티클 존의 오버랩 영역을 제거한 것을 셀 존으로 결정하는 단계일 수 있다.The cell zone determining step S150 may be a step of determining a cell zone by removing the overlapping area of the corrected particle zone from the determined cell zone when the corrected particle zone and the determined cell zone overlap.
한편 본 발명의 목적은, 전술한 셀 존 결정 방법으로 결정된 셀 존에 대하여 N×N 마스크 내에서 N×N 횟수로 고배율 촬상하되, N×N 마스크의 센터 포인트에서 수행된 하나의 오토 포커싱 결과를 N×N 마스크 내의 다른 포인트에서 이용하는 것을 특징으로 하는 셀 존의 고배율 촬상 방법을 제공함으로써 달성될 수 있다. [여기서 N은 3 이상의 홀수이다]On the other hand, an object of the present invention is to perform high-magnification imaging at N×N times in an N×N mask with respect to the cell zone determined by the cell zone determination method described above, and one autofocusing result performed at the center point of the N×N mask This can be achieved by providing a high-magnification imaging method of a cell zone characterized by use at different points in an NxN mask. [where N is an odd number greater than or equal to 3]
상기와 같은 본 발명의 일 실시예에 의하면, 골수 판독을 위해 준비된 슬라이드 샘플 이미지에서 신뢰성이 높은 셀 존의 자동 검출과 신뢰성 높은 셀 존을 디텍팅할 수 있는 효과가 있다.According to an embodiment of the present invention as described above, there is an effect of automatic detection of a highly reliable cell zone and detection of a highly reliable cell zone in a slide sample image prepared for bone marrow reading.
또한 골수 판독을 위해 준비된 슬라이드 샘플 이미지에서 셀 존을 결정함에 있어서 신속성 및 정확성을 높일 수 있는 효과가 있다.In addition, there is an effect of increasing the speed and accuracy in determining the cell zone in the slide sample image prepared for bone marrow reading.
그리고 결정된 셀 존에 대한 고배율 촬상 방법이 신속하고 효율적이다.And the high magnification imaging method for the determined cell zone is fast and efficient.
도 1은 본 발명의 일 실시예에 따른 셀 존 결정 방법을 순차적으로 나타낸 순서도이고,1 is a flowchart sequentially illustrating a method for determining a cell zone according to an embodiment of the present invention;
도 2는 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 리사징(resizing)의 예시를 나타낸 도면이고,2 is a diagram illustrating an example of resizing among a method for determining a cell zone according to an embodiment of the present invention;
도 3은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 컨피던스 존(또는 컨피던스 맵)의 선정 예시를 나타낸 도면이며,3 is a diagram illustrating an example of selecting a confidence zone (or a confidence map) in a method for determining a cell zone according to an embodiment of the present invention;
도 4는 도 3에서 선정된 컨피던스 존에 대하여 하이라이팅을 통해 유효 픽셀을 추출한 상태를 나타낸 도면이고,4 is a view showing a state in which effective pixels are extracted through highlighting for the confidence zone selected in FIG. 3;
도 5는 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 컨피던스 존에 대한 일련의 이미지 처리 과정을 나타낸 순서도이고,5 is a flowchart illustrating a series of image processing processes for a confidence zone in a method for determining a cell zone according to an embodiment of the present invention;
도 6는 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 오토 화이트 밸런싱(Auto White Balancing) 전(a)과 후(b)의 컨피던스 존의 예시를 나타낸 도면이고,6 is a diagram illustrating an example of a confidence zone before (a) and after (b) auto white balancing in a method for determining a cell zone according to an embodiment of the present invention;
도 7은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 컬러 온도 보정(Color Temperature Correction) 후의 컨피던스 존의 예시를 나타낸 도면이며,7 is a view showing an example of a confidence zone after color temperature correction in a method for determining a cell zone according to an embodiment of the present invention;
도 8은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 컬러-브라이트니스 정규화(Color-Brightness Normalization) 전(a)과 후(b)의 컨피던스 존의 예시를 나타낸 도면이고,8 is a view showing examples of confidence zones before (a) and after (b) color-brightness normalization in a method for determining a cell zone according to an embodiment of the present invention;
도 9는 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 백그라운드 서브트랙션(Background Subtraction) 후의 컨피던스 존의 예시(a)와 부분 확대 사진(b)을 나타낸 도면이며,9 is a view showing an example (a) and a partial enlarged photograph (b) of a confidence zone after background subtraction in a method for determining a cell zone according to an embodiment of the present invention;
도 10은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 적분 이미지 방식(Integral Image Method)을 통해 파티클 후보군(중복된 사각 박스)이 표시된 컨피던스 존의 예시를 나타낸 도면이고,10 is a view showing an example of a confidence zone in which a particle candidate group (overlapping square box) is displayed through an integral image method among the cell zone determination methods according to an embodiment of the present invention;
도 11은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 논 맥시마 서프레션 방식(Non-Maxima Suppression Method)을 통해 보정된 파티클 후보(중복되지 않는 사각 박스)가 표시된 컨피던스 존의 예시를 나타낸 도면이고,11 is a view showing an example of a confidence zone in which particle candidates (non-overlapping rectangular boxes) corrected through a Non-Maxima Suppression Method among the cell zone determination method according to an embodiment of the present invention are displayed; FIG. ego,
도 12는 본 발명의 일 실시예에 따른 셀 존 결정 방법에 따라 결정된 셀 존이 표시된 슬라이드 샘플 이미지를 나타낸 도면이며,12 is a view showing a slide sample image in which a cell zone determined according to a method for determining a cell zone according to an embodiment of the present invention is displayed;
도 13은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 파티클과 셀 존이 겹치는 상태(a)와, 고배율 촬상 영역이 되는 유효한 영역의 셀 존(b)을 나타낸 도면이고,13 is a view showing a state in which particles and cell zones overlap (a) and a cell zone (b) of an effective area serving as a high magnification imaging area in a method for determining a cell zone according to an embodiment of the present invention;
도 14는 본 발명의 일 실시예에 따른 셀 존 결정 방법에 따라 결정된 셀 존에 대하여 고배율 촬상을 위한 오토 포커싱을 설명하는 도면이다.14 is a view for explaining auto-focusing for high magnification imaging with respect to a cell zone determined according to a method for determining a cell zone according to an embodiment of the present invention.
이하 첨부 도면들 및 첨부 도면들에 기재된 내용들을 참조하여 본 발명의 실시예를 상세하게 설명하지만, 본 발명이 실시예에 의해 제한되거나 한정되는 것은 아니다.Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings and contents described in the accompanying drawings, but the present invention is not limited or limited by the embodiments.
아래 설명하는 실시예들에는 다양한 변경이 가해질 수 있다. 아래 설명하는 실시예들은 실시 형태에 대해 한정하려는 것이 아니며, 이들에 대한 모든 변경, 균등물 내지 대체물을 포함하는 것으로 이해되어야 한다.Various modifications may be made to the embodiments described below. It should be understood that the embodiments described below are not intended to limit the embodiments, and include all modifications, equivalents, and substitutes thereto.
한편, 본 발명을 설명함에 있어서, 관련된 공지 기능 또는 구성에 대한 구체적인 설명이 본 발명의 요지를 불필요하게 흐릴 수 있다고 판단되는 경우에는, 그 상세한 설명을 생략할 것이다. 그리고, 본 명세서에서 사용되는 용어(terminology)들은 본 발명의 실시예를 적절히 표현하기 위해 사용된 용어들로서, 이는 사용자, 운용자의 의도 또는 본 발명이 속하는 분야의 관례 등에 따라 달라질 수 있다. 따라서, 본 용어들에 대한 정의는 본 명세서 전반에 걸친 내용을 토대로 내려져야 할 것이다.Meanwhile, in describing the present invention, if it is determined that a detailed description of a related well-known function or configuration may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted. In addition, the terms used in this specification are terms used to properly express the embodiment of the present invention, which may vary according to the intention of a user or operator or customs in the field to which the present invention belongs. Accordingly, definitions of these terms should be made based on the content throughout this specification.
골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법Cell Zone Determination Method for Bone Marrow Smeared Slide Sample Images
도 1은 본 발명의 일 실시예에 따른 셀 존 결정 방법을 순차적으로 나타낸 순서도이다. 도 1을 참조하면, 본 발명의 일 실시예는 슬라이드 샘플 이미지에서 컨피던스 존 선정(S110), 백그라운드 서브트랙션(S120), 파티클 존 후보 선정(S130), 파티클 존 보정(S140) 및 셀 존 결정(S150) 단계로 이루어진다. 이러한 본 실시예에 의하면 골수가 도말된 슬라이드 샘플 이미지 상에서 셀 존(Cell Zone)을 결정하여 고배율 촬상을 수행할 수 있으므로, 진단검사의가 별도로 셀 존을 찾는 수고를 덜수 있다.1 is a flowchart sequentially illustrating a method for determining a cell zone according to an embodiment of the present invention. 1, an embodiment of the present invention selects a confidence zone from a slide sample image (S110), background subtraction (S120), particle zone candidate selection (S130), particle zone correction (S140) and cell zone determination ( S150). According to this embodiment, high-magnification imaging can be performed by determining a cell zone on the slide sample image on which the bone marrow is smeared, so that the diagnostician can save the effort of separately finding the cell zone.
이하 도면들을 참조하여 본 실시예에 대하여 상술한다.Hereinafter, the present embodiment will be described in detail with reference to the drawings.
본 실시예는 우선, 도 3에 도시된 바와 같이, 일 방향으로 골수가 도말된 슬라이드 샘플 이미지에서 염색이 두꺼운 딥 존(Deep Zone)이 제거된 컨피던스 존(Confidence Zone or Confidence Map)을 선정하는 컨피던스 존 선정단계(S110)가 수행된다. 컨피던스 존은 슬라이드 표본을 고배율로 관찰하고 판독하기에 적합한 이상적인 다수의 싱글 세포들의 영역인 셀 존(Cell Zone)이 분포할 가능성이 높은 신뢰도 높은 영역이다. 슬라이드를 투과하는 빛의 포화 특성 또는 특정 파장 대역 빛의 투과율 등 빛 특성에 기반하여 마스크 형상으로 맵핑된 컨피던스 존이 선정될 수 있다.In this embodiment, first, as shown in FIG. 3, the confidence zone (Confidence Zone or Confidence Map) in which the deep zone with thick staining is removed from the slide sample image on which the bone marrow is smeared in one direction. A zone selection step (S110) is performed. The confidence zone is a highly reliable area with a high probability of distribution of a cell zone, which is an ideal area of a large number of single cells suitable for observing and reading a slide specimen at high magnification. A confidence zone mapped to a mask shape may be selected based on light characteristics such as saturation characteristics of light passing through the slide or transmittance of light in a specific wavelength band.
선정된 컨피던스 존은 도 2에 도시된 바와 같이, 원본 이미지를 다운스케일링(Downscaling) 함으로써 작은 이미지에서의 연산량을 많이 줄일 수 있다. 또한 원본 이미지가 가지고 있는 노이즈 및 잡음 등을 제거할 수 있다. 따라서 다운스케일링 이후 축소된 이미지에서 컨피던스 존을 선정단계(S110)가 수행되는 것이 바람직하다.As shown in FIG. 2 , in the selected confidence zone, the amount of calculation in a small image can be greatly reduced by downscaling the original image. Also, noise and noise of the original image can be removed. Therefore, it is preferable that the step (S110) of selecting a confidence zone in the reduced image after downscaling is performed.
아울러 컨피던스 존 선정단계(S110)에서 선정된 컨피던스 존은, 도 4에 도시된 바와 같이, 유효 영역에 대응하는 유효 픽셀만을 추출하는 하이라이팅(Highlighting)이 수행됨으로써 유효 픽셀로 채워진 컨피던스 존인 것이 바람직하다.In addition, as shown in FIG. 4 , the confidence zone selected in the confidence zone selection step S110 is preferably a confidence zone filled with effective pixels by performing highlighting for extracting only effective pixels corresponding to the effective area.
여기서 유효 영역은 슬라이드 상에 염색이 되어 있는 부분, 즉 세포가 있는 부분에 해당하며, 빛의 슬라이드 투과 시 염색이 되어 있지 않아 포화(saturation)된 포화 영역을 제거한 영역이다. 이러한 유효 영역의 유효 픽셀을 추출함으로써 이후 이미지 처리의 연산량을 줄일 수 있다.Here, the effective area corresponds to the part on the slide that is stained, that is, the part with cells, and is an area in which the saturated area that is not stained when light is transmitted through the slide is removed. By extracting effective pixels in such an effective area, it is possible to reduce the amount of computation of subsequent image processing.
도 5는 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 컨피던스 존에 대한 일련의 이미지 처리 과정을 나타낸 순서도이다. 도 5에 도시된 바와 같이, 유효 픽셀에 대하여 컬러 균등화를 수행하는 화이트 밸런싱 단계(Auto White Balancing; S112)와 유효 픽셀에 대하여 컬러 온도를 보정하는 컬러 온도 보정단계(Color Temperature Correctness; S114), 그리고 컬러 온도가 보정된 유효 픽셀에 대하여 컬러와 브라이트니스에 대한 정규화를 수행하는 정규화 단계(Color-Brightness Normalization; S116)가 순차적으로 수행될 수 있다. S112, S114, S116과 같은 일련의 단계는, 컴퓨터 비젼 인식을 위한 이미지 전처리 과정이다. 슬라이드 샘플의 속성상 슬라이드 샘플 이미지는 항상 균일한 컬러와 균일한 밝기를 가질 수 없는 한계가 있으므로 이러한 전처리 과정을 통해 균일한 이미지를 획득하여 이용하는 것이 바람직하다.5 is a flowchart illustrating a series of image processing processes for a confidence zone in a method for determining a cell zone according to an embodiment of the present invention. As shown in FIG. 5 , a white balancing step (Auto White Balancing; S112) of performing color equalization on effective pixels and a color temperature correction step of correcting color temperature of effective pixels (Color Temperature Correctness; S114), and A normalization step (Color-Brightness Normalization; S116 ) of performing normalization of color and brightness with respect to an effective pixel whose color temperature has been corrected may be sequentially performed. A series of steps such as S112, S114, and S116 is an image preprocessing process for computer vision recognition. Due to the nature of the slide sample, the slide sample image cannot always have a uniform color and uniform brightness, so it is desirable to obtain and use a uniform image through this pre-processing process.
화이트 밸런싱 단계(S112)는, 도 6에 도시된 바와 같이, 이미지 전체적으로 컬러 균형을 잡는 것이다. 예를 들어 AWB(Auto White Balancing) 적용 전에는 도 6 (a)에서처럼 컬러 속성이 (254, 254, 245)이고 표준편차(SD)가 4.242인 포인트가 AWB(Auto White Balancing) 적용 이후에는 도 6 (b)에서처럼 컬러 속성이 (246, 245, 244)이며 표준편차(SD)가 0.816이 되었다.In the white balancing step (S112), as shown in FIG. 6 , the color balance of the image as a whole is performed. For example, as in Fig. 6 (a) before AWB (Auto White Balancing) is applied, the point having color properties of (254, 254, 245) and standard deviation (SD) of 4.242 is shown in Fig. 6 ( As in b), the color properties were (246, 245, 244) and the standard deviation (SD) was 0.816.
컬러 온도 보정단계(S114)는, 기초 LED 참고예(Basic LED Reference Example)와 이에 대응하는 켈빈 컬러 온도 스케일 차트(Kelvin Color Temperature Scale Chart)를 이용하여 특정 컬러 온도로 보정한다. 예를 들어, 본 실시예에서는 도 7에 도시된 바와 같이, 3800 K 보정을 수행하였다.In the color temperature correction step (S114), a specific color temperature is corrected using a Basic LED Reference Example and a corresponding Kelvin Color Temperature Scale Chart. For example, in this embodiment, as shown in FIG. 7 , 3800 K correction was performed.
컬러-브라이트니스 정규화 단계(S116)는 항상 균등한 밝기가 유지될 수 있도록 특정 밝기로 낮추거나 높이는 단계이다. 예를 들어, 본 실시예에서는 도 8에 도시된 바이, (a)에 비하여 (b)가 3 % 인텐서티(intensity)가 향상되었다.The color-brightness normalization step S116 is a step of lowering or increasing the brightness to a specific brightness so that an even brightness can be always maintained. For example, in the present embodiment, intensity (intensity) is improved by 3% in (b) compared to (a) shown in FIG. 8 .
도 9는 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 백그라운드 서브트랙션(Background Subtraction) 후의 컨피던스 존의 예시(a)와 부분 확대 사진(b)을 나타낸 도면이다. 백그라운드 서브트랙션 단계(120)는 선정된 컨피던스 존에서 세포 이미지의 컬러와 질감 특징에 기반하여 백그라운드를 제거하고 세포 이미지를 추출하는 과정이다. 9 is a view showing an example (a) and a partial enlarged photograph (b) of a confidence zone after background subtraction in a method for determining a cell zone according to an embodiment of the present invention. The background subtraction step 120 is a process of removing the background and extracting the cell image based on the color and texture characteristics of the cell image in the selected confidence zone.
백그라운드 서브스트랙션 단계(S120)는 1차적으로 블러(blurr)된 영역과, 밝기가 일정 이하인 영역, 색감이 일정 이하인 영역을 제거하는 방식으로 진행될 수 있다. 이러한 백그라운드 서브스트랙션을 통해 세포에 대응하는 컬러 특징(세포 이미지의 컬러와 질감)이 추출될 수 있다. 이러한 백그라운드 서브스트랙션 단계(S120) 이후에 이웃 픽셀 정보를 활용하여 이미지 끊김 현상을 보정하는 모폴로지(Morphology)가 수행될 수도 있다.The background subtraction step ( S120 ) may be performed in a manner that primarily removes a blurred area, an area having a brightness below a certain level, and an area having a color below a certain level. Through this background subtraction, color features (color and texture of the cell image) corresponding to the cell can be extracted. After the background subtraction step ( S120 ), a morphology of correcting image breakage by using neighboring pixel information may be performed.
도 10은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 적분 이미지 방식(Integral Image Method)을 통해 파티클 후보군(중복된 사각 박스)이 표시된 컨피던스 존의 예시를 나타낸 도면이다. 파티클 존 후보 선정단계(130)는 백그라운드가 제거된 컨피던스 존에서 적분 이미지 방식에 기반하여 다수의 파티클 후보를 선정하는 과정이다. 여기서 파티클은, 골수 샘플에서 얻어진 물질로서, 대부분 정상 세포가 아니며, 육안으로 확인 가능할 만큼 거대한 형상으로 나타나는 것이 일반적이다.10 is a diagram illustrating an example of a confidence zone in which a particle candidate group (overlapping rectangular box) is displayed through an integral image method among a cell zone determination method according to an embodiment of the present invention. The particle zone candidate selection step 130 is a process of selecting a plurality of particle candidates based on the integral image method in the confidence zone from which the background is removed. Here, the particles, which are substances obtained from bone marrow samples, are not mostly normal cells, and generally appear in a shape large enough to be identified with the naked eye.
도 11은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 논 맥시마 서프레션 방식(Non-Maxima Suppression Method)을 통해 보정된 파티클 후보(중복되지 않는 사각 박스)가 표시된 컨피던스 존의 예시를 나타낸 도면이다. 파티클 존 보정단계(S140)는 선정된 다수의 파티클 후보 중 상호 오버랩되는 파티클 후보에 대하여 컬러 속성(컬러, 채도 및 밝기)을 기준으로 파티클 존으로 보정하는 과정이다. 논 맥시마 서프레션 방식(Non-Maxima Suppression Method)을 통해 중복된 파티클 후보에 대하여 큰 값을 갖는 단일한 파티클 존으로 흡수시킴으로써 보정이 수행될 수 있다. 따라서 보정된 파티클 존은 상호 중복되지 않는 형태로 추출된다.11 is a view showing an example of a confidence zone in which particle candidates (non-overlapping rectangular boxes) corrected through a Non-Maxima Suppression Method among the cell zone determination method according to an embodiment of the present invention are displayed; FIG. to be. The particle zone correction step ( S140 ) is a process of correcting the particle candidates that overlap each other among a plurality of selected particle candidates as a particle zone based on color properties (color, saturation, and brightness). Correction may be performed by absorbing duplicate particle candidates into a single particle zone having a large value through the Non-Maxima Suppression Method. Therefore, the corrected particle zones are extracted in a form that does not overlap with each other.
도 12는 본 발명의 일 실시예에 따른 셀 존 결정 방법에 따라 결정된 셀 존이 표시된 슬라이드 샘플 이미지를 나타낸 도면이다. 도 12에 도시된 바와 같이, 셀 존 결정단계(S150)는 보정된 파티클 존을 기준으로 도말 반대 방향으로 기 설정된 거리만큼 이격된 영역에서 셀 존을 결정하는 과정이다. 12 is a diagram illustrating a slide sample image in which a cell zone determined according to a method for determining a cell zone according to an embodiment of the present invention is displayed. As shown in FIG. 12 , the cell zone determination step S150 is a process of determining a cell zone in an area spaced apart by a preset distance in the opposite direction to the smearing based on the corrected particle zone.
다만, 골수 샘플에 따라 파티클이 존재하지 않는 경우도 있을 수 있다. 이런 경우에는 추출된 세포 이미지의 세포 군집도가 기 설정된 기준보다 높은 영역 또는 독립된 단일 세포가 많이 분포하는 영역을 셀 존으로 결정할 수 있다. 이때에도 셀 존 결정방법의 통일화를 위해 파티클 존을 상정하고 상정된 파티클 존을 기준으로 도말 반대 방향으로 기 설정된 거리만큼 이격된 영역을 셀 존으로 결정할 수 있다.However, there may be cases in which particles do not exist depending on the bone marrow sample. In this case, a region in which the cell population of the extracted cell image is higher than a preset standard or a region in which a large number of independent single cells are distributed may be determined as the cell zone. Even at this time, a particle zone is assumed for the unification of the cell zone determination method, and an area spaced apart by a preset distance in the direction opposite to the smearing based on the assumed particle zone may be determined as the cell zone.
도 13은 본 발명의 일 실시예에 따른 셀 존 결정 방법 중 파티클과 셀 존이 겹치는 상태(a)와, 고배율 촬상 영역이 되는 유효한 영역의 셀 존(b)을 나타낸 도면이다. 셀 존 결정단계(S150)는, 보정된 파티클 존 및 결정된 셀 존이 오버랩되는 경우, 결정된 셀 존에서 보정된 파티클 존의 오버랩 영역을 제거한 것을 셀 존으로 결정하는 단계일 수 있다. 파티클 존 오버랩 영역을 제거함으로써 셀 존을 더욱 정확하게 결정할 수 있으며 연산량을 줄일 수 있어 신속하고 정확한 진단에 도움이 된다.13 is a diagram illustrating a state in which particles and cell zones overlap (a) and a cell zone (b) of an effective area serving as a high magnification imaging area in a method for determining a cell zone according to an embodiment of the present invention. The cell zone determining step S150 may be a step of determining a cell zone by removing the overlapping area of the corrected particle zone from the determined cell zone when the corrected particle zone and the determined cell zone overlap. By removing the overlapping area of the particle zone, the cell zone can be determined more accurately and the amount of computation can be reduced, which helps in quick and accurate diagnosis.
도 14는 본 발명의 일 실시예에 따른 셀 존 결정 방법에 따라 결정된 셀 존에 대하여 고배율 촬상을 위한 오토 포커싱을 설명하는 도면이다. 도 14에 도시된 바와 같이 전술한 셀 존 결정방법으로 결정된 셀 존에 대해서는 진단을 위한 고배율 촬상이 수행된다. 결정된 셀 존에 대하여 N×N 마스크 내에서 N×N 횟수로 고배율 촬상하되, N×N 마스크의 센터 포인트에서 수행된 하나의 오토 포커싱 결과를 N×N 마스크 내의 다른 포인트에서 이용하는 것이 바람직하다. 이렇게 함으로써 반복되는 다수의 오토포커싱에 따른 연산량을 감소시킬 수 있다[여기서 N은 3 이상의 홀수이다].14 is a view for explaining auto-focusing for high magnification imaging with respect to a cell zone determined according to a method for determining a cell zone according to an embodiment of the present invention. As shown in FIG. 14 , high magnification imaging for diagnosis is performed on the cell zone determined by the method for determining the cell zone. It is preferable to image the determined cell zone at high magnification at N×N times in an N×N mask, and use one autofocusing result performed at the center point of the N×N mask at another point in the N×N mask. By doing so, it is possible to reduce the amount of computation due to repeated autofocusing (where N is an odd number equal to or greater than 3).
이상 첨부된 도면을 참조하여 본 발명의 실시 예를 설명하였지만, 상술한 본 발명의 기술적 구성은 본 발명이 속하는 기술 분야의 당 업자가 본 발명의 그 기술적 사상이나 필수적 특징을 변경하지 않고서 다른 구체적인 형태로 실시될 수 있다는 것을 이해할 수 있을 것이다. 그러므로 이상에서 기술한 실시 예들은 모든 면에서 예시적인 것이며 한정적인 것이 아닌 것으로서 이해되어야 한다. 아울러, 본 발명의 범위는 상기의 상세한 설명보다는 후술하는 특허청구범위에 의하여 나타내어진다. 또한, 특허청구범위의 의미 및 범위 그리고 그 등가 개념으로부터 도출되는 모든 변경 또는 변형된 형태가 본 발명의 범위에 포함되는 것으로 해석되어야 한다.Although the embodiments of the present invention have been described above with reference to the accompanying drawings, the technical configuration of the present invention described above is another specific form for those skilled in the art to which the present invention pertains without changing the technical spirit or essential features of the present invention. It will be understood that it can be implemented as Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. In addition, the scope of the present invention is indicated by the following claims rather than the above detailed description. In addition, all changes or modifications derived from the meaning and scope of the claims and their equivalents should be construed as being included in the scope of the present invention.

Claims (7)

  1. 일 방향으로 골수가 도말된 슬라이드 샘플 이미지에서 염색이 두꺼운 딥 존이 제거된 컨피던스 존을 선정하는 컨피던스 존 선정단계(S110);Confidence zone selection step (S110) of selecting a confidence zone from which a deep zone with thick staining is removed from the slide sample image on which the bone marrow is smeared in one direction;
    상기 선정된 컨피던스 존에서 세포 이미지의 컬러와 질감 특징에 기반하여 백그라운드를 제거하고 상기 세포 이미지를 추출하는 백그라운드 서브트랙션 단계(120);a background subtraction step 120 of removing a background based on the color and texture characteristics of the cell image in the selected confidence zone and extracting the cell image;
    상기 백그라운드가 제거된 컨피던스 존에서 적분 이미지 방식에 기반하여 다수의 파티클 후보를 선정하는 파티클 존 후보 선정단계(130);a particle zone candidate selection step 130 of selecting a plurality of particle candidates based on an integral image method in the confidence zone from which the background is removed;
    상기 선정된 다수의 파티클 후보 중 상호 오버랩되는 파티클 후보에 대하여 컬러 속성을 기준으로 파티클 존으로 보정하는 파티클 존 보정단계(S140); 및a particle zone correction step (S140) of correcting overlapping particle candidates from among the selected plurality of particle candidates into particle zones based on color properties; and
    상기 추출된 세포 이미지의 세포 군집도가 기 설정된 기준보다 높은 영역 또는 상기 보정된 파티클 존을 기준으로 도말 반대 방향으로 기 설정된 거리만큼 이격된 영역에서 셀 존을 결정하는 셀 존 결정단계(S150)를 포함하는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법.A cell zone determination step (S150) of determining a cell zone in an area where the cell population of the extracted cell image is higher than a preset standard or in an area spaced apart by a preset distance in the opposite direction to the smear based on the corrected particle zone A method for determining the cell zone of an image of a slide sample containing bone marrow smeared.
  2. 제1 항에 있어서,According to claim 1,
    상기 컨피던스 존 선정단계(S110)에서,In the confidence zone selection step (S110),
    상기 도말된 슬라이드 샘플 이미지는 스케일이 다운되어 조정된 것을 특징으로 하는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법.The method for determining the cell zone of the bone marrow plated slide sample image, characterized in that the plated slide sample image is scaled down.
  3. 제1 항에 있어서,According to claim 1,
    상기 컨피던스 존 선정단계(S110)에서,In the confidence zone selection step (S110),
    상기 선정된 컨피던스 존은 유효 영역에 대응하는 유효 픽셀이 추출되는 하이라이팅이 수행된 것을 특징으로 하는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법.The selected confidence zone is a cell zone determination method of a slide sample image on which bone marrow is smeared, characterized in that highlighting is performed in which effective pixels corresponding to the effective area are extracted.
  4. 제3 항에 있어서,4. The method of claim 3,
    상기 컨피던스 존 선정단계(S110)와 상기 백그라운드 서브스트랙션 단계(S120) 사이에,Between the confidence zone selection step (S110) and the background subtraction step (S120),
    상기 유효 픽셀에 대하여 컬러 균등화를 수행하는 화이트 밸런싱 단계(S112); 및a white balancing step of performing color equalization on the effective pixels (S112); and
    상기 유효 픽셀에 대하여 컬러 온도를 보정하는 컬러 온도 보정단계(S114)를 더 포함하는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법.The cell zone determination method of the bone marrow-smeared slide sample image further comprising a color temperature correction step (S114) of correcting the color temperature for the effective pixels.
  5. 제4 항에 있어서,5. The method of claim 4,
    상기 컬러 온도 보정단계(S114)와 상기 백그라운드 서브스트랙션 단계(S120) 사이에,Between the color temperature correction step (S114) and the background subtraction step (S120),
    상기 컬러 온도가 보정된 유효 픽셀에 대하여 컬러와 브라이트니스에 대한 정규화를 수행하는 정규화 단계(S116)를 더 포함하는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법.The method for determining a cell zone of a slide sample image on which bone marrow is smeared further comprises a normalization step (S116) of performing normalization for color and brightness with respect to the effective pixel whose color temperature has been corrected.
  6. 제1 항에 있어서, According to claim 1,
    상기 셀 존 결정단계(S150)는,The cell zone determination step (S150) is,
    상기 보정된 파티클 존 및 상기 결정된 셀 존이 오버랩되는 경우, 상기 결정된 셀 존에서 상기 보정된 파티클 존의 오버랩 영역을 제거한 것을 셀 존으로 결정하는 단계인 것을 특징으로 하는 골수가 도말된 슬라이드 샘플 이미지의 셀 존 결정 방법.When the corrected particle zone and the determined cell zone overlap, the step of removing the overlapping area of the corrected particle zone from the determined cell zone is a step of determining as a cell zone of the bone marrow smeared slide sample image Cell Zone Determination Method.
  7. 제1 내지 제6 항 중 어느 한 항에 따른 셀 존 결정 방법으로 결정된 셀 존에 대하여 N×N 마스크 내에서 N×N 횟수로 고배율 촬상하되, 상기 N×N 마스크의 센터 포인트에서 수행된 하나의 오토 포커싱 결과를 상기 N×N 마스크 내의 다른 포인트에서 이용하는 것을 특징으로 하는 셀 존의 고배율 촬상 방법. The cell zone determined by the cell zone determination method according to any one of claims 1 to 6, but with high magnification imaging at N×N times in an N×N mask, one image performed at the center point of the N×N mask A high magnification imaging method of a cell zone, characterized in that the auto-focusing result is used at different points in the N×N mask.
    [여기서 상기 N은 3 이상의 홀수이다.][Wherein N is an odd number of 3 or more.]
PCT/KR2021/018680 2020-12-10 2021-12-09 Method for determining cell zone of sample image of slide smeared with bone marrow, and method for imaging cell zone at high magnification WO2022124832A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR1020200172022A KR102613961B1 (en) 2020-12-10 2020-12-10 Determination method for cell zone of slide sample image smeared with bone-marrow and high magnification imaging method of the same cell zone
KR10-2020-0172022 2020-12-10

Publications (1)

Publication Number Publication Date
WO2022124832A1 true WO2022124832A1 (en) 2022-06-16

Family

ID=81974519

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2021/018680 WO2022124832A1 (en) 2020-12-10 2021-12-09 Method for determining cell zone of sample image of slide smeared with bone marrow, and method for imaging cell zone at high magnification

Country Status (2)

Country Link
KR (1) KR102613961B1 (en)
WO (1) WO2022124832A1 (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040136581A1 (en) * 1996-11-27 2004-07-15 Chroma Vision Medical Systems, Inc., A California Corporation Method and apparatus for automated image analysis of biological specimens
JP2015052581A (en) * 2013-01-08 2015-03-19 キヤノン株式会社 Biological tissue image reconfiguration method and device, and image display device using the biological tissue image
KR102105489B1 (en) * 2018-12-31 2020-04-28 (주)유아이엠디 Microscopy based slide scanning system for bone marrow interpretation
KR20200046271A (en) * 2018-10-24 2020-05-07 (주)유아이엠디 Detection device and control method thereof
KR20200082746A (en) * 2018-12-31 2020-07-08 (주)유아이엠디 Cell-zone labeling apparatus and cell-zone detecting system including the same apparatus

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109121437A (en) * 2015-12-18 2019-01-01 雅培实验室 For assessing the method and system of histological stain
KR101962869B1 (en) * 2018-08-02 2019-03-27 주식회사 우리메디칼 Bone marrow interpretation supporting apparatus based on image analysis

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040136581A1 (en) * 1996-11-27 2004-07-15 Chroma Vision Medical Systems, Inc., A California Corporation Method and apparatus for automated image analysis of biological specimens
JP2015052581A (en) * 2013-01-08 2015-03-19 キヤノン株式会社 Biological tissue image reconfiguration method and device, and image display device using the biological tissue image
KR20200046271A (en) * 2018-10-24 2020-05-07 (주)유아이엠디 Detection device and control method thereof
KR102105489B1 (en) * 2018-12-31 2020-04-28 (주)유아이엠디 Microscopy based slide scanning system for bone marrow interpretation
KR20200082746A (en) * 2018-12-31 2020-07-08 (주)유아이엠디 Cell-zone labeling apparatus and cell-zone detecting system including the same apparatus

Also Published As

Publication number Publication date
KR20220082277A (en) 2022-06-17
KR102613961B1 (en) 2023-12-14

Similar Documents

Publication Publication Date Title
JP4071186B2 (en) Method and system for identifying an object of interest in a biological specimen
JP4550415B2 (en) Quantitative video microscopy and related system and computer software program products
AU2003236675B2 (en) Method for quantitative video-microscopy and associated system and computer software program product
CN100561186C (en) The method of testing and the device that are used for the textile color stability and color aberration grading
US20060067887A1 (en) Scoring estrogen and progesterone receptors expression based on image analysis
BRPI0610115A2 (en) methods for image analysis based on chromogen separation
AU2002334590A1 (en) Method quantitative video-microscopy and associated system and computer software program product
AU2003236675A1 (en) Method for quantitative video-microscopy and associated system and computer software program product
WO2020141812A1 (en) Region of interest labeling device for bone marrow interpretation, and region of interest detection system comprising same
EP3719739B1 (en) Image coloring device, image coloring method, image learning device, image learning method, program, and image coloring system
CN112990015A (en) Automatic lesion cell identification method and device and electronic equipment
CN111863118A (en) Method for carrying out TCT and DNA ploidy analysis based on TCT film-making
CN111656393A (en) Histological image analysis
RU2548399C2 (en) Method for producing processed virtual analysis image
JPWO2019171909A1 (en) Image processing method, image processing device and program
WO2022124832A1 (en) Method for determining cell zone of sample image of slide smeared with bone marrow, and method for imaging cell zone at high magnification
WO2021075785A1 (en) Method for photographing bone marrow cell image using bone marrow image analyzer
KR20060022266A (en) System for classifying slides using scatter plot distributions
JPH10115612A (en) Assay of cancer cell
US20210174147A1 (en) Operating method of image processing apparatus, image processing apparatus, and computer-readable recording medium
WO2023282462A1 (en) Blood cell detection and classification method based on examination area designation
Murakami et al. Color processing in pathology image analysis system for liver biopsy
WO2022250460A1 (en) Image-based method for counting platelets, and output method for counting information
WO2022173176A1 (en) Method for obtaining high power image of slide sample
WO2022215989A1 (en) Method of detecting optimal area to examine on smear slide

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: 21903878

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21903878

Country of ref document: EP

Kind code of ref document: A1