WO2013175683A1 - 病理診断結果判定システム、病理診断結果判定方法および病理診断結果判定装置 - Google Patents
病理診断結果判定システム、病理診断結果判定方法および病理診断結果判定装置 Download PDFInfo
- Publication number
- WO2013175683A1 WO2013175683A1 PCT/JP2013/001451 JP2013001451W WO2013175683A1 WO 2013175683 A1 WO2013175683 A1 WO 2013175683A1 JP 2013001451 W JP2013001451 W JP 2013001451W WO 2013175683 A1 WO2013175683 A1 WO 2013175683A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- report
- unit
- diagnosis
- feature
- pathological diagnosis
- 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
Images
Classifications
-
- 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
- G01N33/4833—Physical analysis of biological material of solid biological material, e.g. tissue samples, cell cultures
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30024—Cell structures in vitro; Tissue sections in vitro
Definitions
- the present invention relates to a pathological diagnosis result determination system, a pathological diagnosis result determination method, and a pathological diagnosis result determination apparatus for finding a discrepancy between a result of diagnosis of a tissue specimen image and a report content.
- pathological diagnosis method there is a method in which a specimen (pathological specimen) such as a diseased tissue or cell collected from a human body is pasted on a glass slide and observed and diagnosed using a microscope.
- pathological diagnosis image diagnosis
- a scanner that captures a specimen on a slide as a digital image
- pathological diagnosis is performed with a microscope or digital image
- the pathologist summarizes the result in a pathological diagnosis report.
- Japanese Patent Application Laid-Open No. 2010-035756 uses gaze detection to detect which feature and how much time in a medical diagnostic image is viewed, and has been set in advance by feature extraction. Methods have been proposed to give warnings about oversight of areas. This method is effective when a gaze detection device can be introduced at the diagnosis site and a warning is given.
- the objective of this invention is providing the technique which solves the above-mentioned subject.
- the pathological diagnosis result determination system performs a pathological diagnosis of a tissue specimen image, generates a diagnostic record information, and accumulates a report describing the pathological diagnosis result of the tissue specimen image
- the report storage unit that performs analysis, the report analysis unit that analyzes the diagnosis results described in the reports stored in the report storage unit, the diagnosis results analyzed by the report analysis unit, and the diagnosis record information are compared, and diagnosis is performed.
- a report verification unit for determining the degree of coincidence of the results.
- the pathological diagnosis result determination method of the present invention performs a pathological diagnosis of a tissue specimen image, generates a diagnostic record information, and accumulates a report describing the pathological diagnosis result of the tissue specimen image in a report storage unit. Compare the diagnosis record information with the report accumulation step, the report analysis step that analyzes the diagnosis results described in the reports accumulated in the report accumulation unit, and the diagnosis record information. And a report verification step for determining the degree of coincidence.
- the pathological diagnosis result determination apparatus of the present invention includes a diagnostic unit that performs pathological diagnosis of a tissue specimen image and generates diagnostic record information, and a report storage unit that accumulates a report in which the pathological diagnosis result of the tissue specimen image is described.
- the report analysis unit that analyzes the diagnosis results described in the reports stored in the report storage unit, the diagnosis results analyzed by the report analysis unit, and the diagnosis record information are compared, and the degree of coincidence of the diagnosis results is determined.
- a report verification unit for judging.
- the present invention it is possible to prevent a difference between the result of diagnosing the tissue specimen image and the report contents and to reduce the risk of misdiagnosis.
- Embodiment 1 FIG.
- exemplary embodiments of the present invention will be described in detail with reference to the drawings.
- the components described in the following embodiments are merely examples, and are not intended to limit the technical scope of the present invention.
- FIG. 1 is a block diagram showing a configuration example of a first embodiment of a pathological diagnosis determination system according to the present invention.
- the image storage unit 101 reads a specimen slide of a diseased tissue with an imaging device such as a scanner, and stores the generated tissue specimen image.
- the diagnosis unit 102 requests the image storage unit 101 to transmit the tissue specimen image, and extracts the outline of the tissue in the obtained tissue image. Furthermore, the diagnosis unit 102 has a function of extracting various features existing in the tissue.
- features used for pathological diagnosis include nuclei, gland ducts, signet ring cells, mucus, and necrotic cells.
- the diagnosis unit 102 extracts a feature region from the target region using a general image processing unit such as an edge extraction filter or template matching after limiting the rough region including the target region with the color value of the pixel.
- the feature region is a region including a feature used for pathological diagnosis.
- the diagnosis unit 102 determines and analyzes a gaze area (ROI) based on various features extracted from the tissue specimen image, and performs pathological diagnosis. At this time, values (feature amounts) relating to the color and shape of the features are collectively used as a feature vector.
- the diagnosis unit 102 inputs a feature vector to a classifier previously learned by a machine learning method such as a neural network (NN) or a support vector machine (SVM), and performs pathological diagnosis.
- NN neural network
- SVM support vector machine
- Tissue specimen identification information is information for identifying a tissue specimen.
- a barcode including identification information is attached to the tissue specimen.
- a gaze area is an area to be a pathological diagnosis target in a tissue specimen image including an important feature in pathological diagnosis.
- the diagnosis result is a pathological diagnosis result of the tissue specimen.
- the diagnosis result indicates whether the tissue specimen is “malignant” or “benign”.
- the pathological report is accumulated in the report accumulation unit 104.
- the pathology report includes patient information (tissue specimen identification information (ID), age, gender), pathologist information, specimen information (organ name, biopsy sampling position in the organ, sampling method) ), Text information such as diagnostic information (diagnosis results, findings) is described.
- patient information tissue specimen identification information (ID)
- age, gender a registered trademark
- pathologist information a registered trademark
- specimen information an extracted material
- organ name organ name, biopsy sampling position in the organ, sampling method
- Text information such as diagnostic information (diagnosis results, findings) is described.
- other items necessary for pathological diagnosis may be added as necessary.
- the report analysis unit 105 extracts a diagnosis result by syntax analysis of natural language processing based on the description content of the pathology report.
- the diagnosis result indicates whether the tissue specimen is “malignant” or “benign”.
- the diagnosis result is accumulated as diagnosis record information in the report analysis unit 105 together with the tissue specimen identification information (ID) and the gaze area (ROI).
- the report verification unit 106 determines the degree of coincidence between the diagnosis result stored in the diagnosis unit 102 and the diagnosis record information stored in the report storage unit 104. In the first embodiment, the report verification unit 106 determines the degree of coincidence between the diagnosis record information accumulated in the diagnosis unit 102 and the diagnosis record information accumulated in the report accumulation unit 104.
- FIG. 2 is a flowchart showing an example of an operation procedure of the pathological diagnosis determination system according to the first embodiment.
- the diagnosis unit 102 requests the tissue specimen image stored in the image storage unit 101 (S100).
- the diagnosis unit 102 extracts the tissue contour in the requested tissue specimen image by the same method as the feature extraction described above. Further, the diagnosis unit 102 extracts various features existing in the organization, and calculates feature amounts that are attribute information (representing feature density, feature shape, size, etc.) of the various features (S101a, b).
- the diagnosis unit 102 determines a gaze area (ROI) based on the extracted various feature information, and performs pathological diagnosis (S102).
- the diagnosis unit 102 may use a method described in Japanese Patent Application Laid-Open No. 2009-175040 as a method of determining a gaze area (ROI), or may use another method.
- the diagnosis record information in the first embodiment is information on tissue specimen identification information (ID), a gaze area (ROI), and a diagnosis result.
- the diagnostic information storage unit 103 stores the diagnostic record information obtained by the diagnostic unit 102 as a table as shown in FIG. 3A (S103).
- the report analysis unit 105 requests a report stored in the report storage unit 104 (S200).
- the report analysis unit 105 extracts the diagnostic record information described in the report (S201), and records it in the report analysis unit 105 itself as a table as shown in FIG. 3b (S202).
- the report verification unit 106 compares the diagnosis record information (FIGS. 3a and 3b) recorded in the diagnosis unit 102 and the report analysis unit 105, and determines the degree of coincidence between them.
- diagnosis record information (FIGS. 3a and 3b) recorded in the diagnosis unit 102 and the report analysis unit 105, and determines the degree of coincidence between them.
- the report verification unit 106 first compares tissue specimen identification information (ID) (S300). If the tissue specimen identification information (ID) matches, the report verification unit 106 then compares the gaze area (ROI) (S301). If the tissue specimen identification information (ID) does not match, the flow ends there (S303).
- Step S300 If the tissue specimen identification information (ID) matches in step S300, the report verification unit 106 next compares the gaze area (ROI) (S301).
- ROI gaze area
- the report verification unit 106 sets a rectangular region (bounding box) in contact with the upper end, lower end, left end, and right end of the tissue region extracted by the diagnosis unit 102.
- the report verification unit 106 divides the inside of the rectangular area into a plurality of blocks. For example, the report verification unit 106 divides the vertical direction into four parts and the horizontal direction into four parts, generates a total of 16 blocks, and assigns each block a number from 1 to 16.
- Correlation of path information and location information of each block is performed as follows.
- Information regarding the gaze area (ROI) described in the pathology report is often indicated only by a rough position such as “upper right”.
- the correspondence between the gaze area (ROI) and the block number in advance can be specified.
- the report analysis unit 105 performs syntax analysis of the pathology report and acquires gaze area (ROI) information.
- ROI gaze area
- the word “upper right” can be extracted as the attention area (ROI) information.
- the position in the tissue specimen image corresponding to the “upper right” of the pathology report is defined as block numbers 3, 4 and 8 as shown in FIG. It can be determined (S301). If the gaze areas do not match, the flow ends there (S303).
- the report verification unit 106 compares the diagnosis results recorded in the diagnosis unit 102 and the report analysis unit 105 (S302). If both diagnosis results match with “benign”, the report verification unit 106 sets the final diagnosis result to “benign”, and the flow ends (S303).
- the report verification unit 106 sets the final diagnosis result to “malignant”, and the flow ends (S303).
- the report verification unit 106 may perform detailed evaluation of cancer malignancy (for example, grading). At this time, the report verification unit 106 may display the diagnostic record information on the display screen.
- the report verification unit 106 displays a warning on the display screen, and the flow ends (S303).
- the present invention it is possible to prevent the difference between the result of diagnosis of the tissue specimen image and the report content created by the pathologist.
- Embodiment 2 features and attribute information necessary for pathological diagnosis are extracted based on the description contents of the pathological report accumulated in the report accumulation unit 104. Then, the features included in the diagnostic record information stored in the diagnostic information storage unit 103 are compared with the attribute information, and the degree of coincidence between them is determined. Further, at this time, a score corresponding to the degree of coincidence between the two is calculated, and a warning is displayed when the calculated score is more than a predetermined value beyond a preset standard value.
- FIG. 7 is a block diagram showing a configuration example of the second embodiment of the pathological diagnosis determination system according to the present invention.
- the diagnosis unit 102 includes a diagnosis feature recognition unit 112 and a diagnosis feature analysis unit 122.
- the report verification unit 106 includes a feature-specific importance table recording unit 126 and a score calculation unit 136. Note that the description of the configuration common to the first embodiment is omitted.
- the diagnosis unit 102 includes a diagnosis feature recognition unit 112, a diagnosis feature analysis unit 122, and a diagnosis recording unit 132.
- the diagnostic feature recognition unit 112 requests the image storage unit 101 to transmit a tissue specimen image, and extracts the tissue contour in the obtained tissue image. Further, the diagnostic feature recognition unit 112 has a function of extracting various features existing in the tissue.
- features used for pathological diagnosis include nucleus, gland duct, signet ring cell, mucus, necrotic cell and the like. Since the feature extraction method is the same as that of the first embodiment, the description thereof is omitted here.
- the diagnostic feature analysis unit 122 analyzes the tissue and features in the tissue image extracted by the diagnostic feature recognition unit 112, and performs pathological diagnosis.
- the diagnostic feature analysis unit 122 extracts a feature region from the tissue image.
- the diagnostic feature analysis unit 122 determines the amount of the area shape (area, circumference, circularity, major axis / minor axis and ratio thereof, number of thinned pixels and geodesic lines, end points, intersections), and The average value and variance of the color elements (RGB, HSV, etc.) of the pixels in the region are calculated as features.
- the diagnostic feature analysis unit 122 collects these values into a feature vector, and uses a threshold learning or a machine learning method represented by a neural network (NN), support vector machine (SVM), or the like to perform benign or malignant. Determine the degree.
- NN neural network
- SVM support vector machine
- Diagnostic feature analysis unit 122 may determine the degree by binary values such as 0 and 1. Further, the diagnostic feature analysis unit 122 may determine the degree using a continuous value such as a distance from the threshold or a numerical value calculated by a machine learning method. The diagnostic feature analysis unit 122 may use generally known technology of feature analysis methods including the above method.
- the diagnosis recording unit 132 records tissue specimen identification information (ID), gaze area (ROI), feature name, feature attribute information, and the like of the tissue specimen image analyzed by the diagnostic feature analysis unit 122 as diagnostic record information.
- the report analysis unit 105 includes an extraction unit 115 and a recording unit 125.
- the extraction unit 115 extracts features and attribute information of the features by syntax analysis of natural language processing based on the description content of the pathology report.
- the feature attribute information represents feature density, feature shape, size, and the like.
- the extraction unit 115 performs pathological diagnosis based on the obtained features and feature attribute information. As shown in FIG. 10, the diagnostic result is accumulated as diagnostic record information in the diagnostic information accumulating unit 103 together with the tissue specimen identification information (ID) and the gaze area (ROI).
- the recording unit 125 records the feature extracted by the extracting unit 115 and its attribute information as shown in FIG.
- the report verification unit 106 includes a comparison unit 116, a feature-specific importance table recording unit 126, and a score calculation unit 136.
- the comparison unit 116 compares the diagnostic record information stored in the diagnostic information storage unit 103 with the features and attribute information recorded in the recording unit 125 of the report analysis unit 105.
- the feature-specific importance table recording unit 126 records the weighting values used for score calculation as shown in FIG. This weighting value is set in advance corresponding to each of the feature and its attribute information.
- the weighting value is extremely large as it strongly supports malignant judgment based on the presence or absence of such features as signet ring cells in gastric cancer diagnosis and comedoednecrosis in prostate cancer diagnosis.
- a weighting value is set according to the malignancy occurrence rate and the seriousness (according to cancer, the progress is fast, etc.) based on the attribute information.
- the attribute information includes words representing size, density, shape, and the like.
- a finding such as “the gland ducts having different sizes are densely packed” is described in the pathological report.
- a weighting value of +5 is set for the item “the gland ducts are dense” and +7 is set for the item “the size of the gland ducts are different”.
- the score calculation unit 136 calculates a warning score by a weighted sum or the like using the matching score given by the comparison unit 116 and the weighting value recorded in the feature-specific importance table recording unit 126.
- the coincidence score is a score given by the degree of coincidence when features are compared.
- the diagnostic record information of the same tissue specimen image recorded in the diagnostic information recording unit 103 and the recording unit 125 of the report analysis unit 105 when the features of the two are compared, if the features of the two match, +1 If they do not match, ⁇ 1 is assigned as the matching score.
- the warning score is a score derived by multiplying the feature or attribute information matching items by a feature matching score and a preset weighting value as shown in FIG. 11 and calculating the sum. is there.
- the warning score is an index indicating how far the pathological diagnosis results of the diagnostic information recording unit 103 and the report analysis unit 105 are far apart.
- the diagnostic feature recognition unit 112 extracts the tissue and various feature regions in the tissue specimen image. Further, the diagnostic feature analysis unit 122 determines a gaze area (ROI) based on the extracted various feature information and performs pathological diagnosis. Since the steps up to here are the same as those in the first embodiment, description thereof will be omitted.
- ROI gaze area
- FIG. 12 shows that a plurality of gaze regions (ROI) are set based on the tissue in the tissue specimen image recognized by the diagnostic feature recognition unit 112 and the feature information, and one gaze region (ROI) is determined. It is explanatory drawing shown.
- the diagnostic feature analysis unit 122 sets the feature existing in the gaze area to a magnification suitable for diagnosis. Since the pathological diagnosis method of the gaze area (ROI) performed by the diagnostic feature analysis unit 122 is the same as that in the first embodiment, the description thereof is omitted.
- the diagnostic feature analysis unit 122 sends the tissue specimen identification information (ID), gaze region (ROI) information, feature name, and attribute information obtained when analyzing the gland duct to the diagnostic information storage unit 103.
- the diagnostic information storage unit 103 records the sent information as shown in FIG.
- the report accumulated in the report accumulation unit 104 is transmitted to the report analysis unit 105.
- the extraction unit 115 of the report analysis unit 105 extracts pathological specimen identification information (ID), gaze area (ROI), feature name and attribute information described in the report, and records them in a table format as shown in FIG. Recorded in section 125.
- the report analysis unit 105 extracts the feature name and the attribute information thereof based on the description contents of the report accumulated in the report accumulation unit 104 . For example, suppose that there is a finding in the pathology report that gland ducts of different sizes are dense. In this case, the report analysis unit 105 extracts “gland duct” as a feature, “different in size”, and “dense” as attribute information. The information obtained here is recorded in the recording unit 125 of the report analysis unit 105 in a table format as shown in FIG.
- the comparison unit 116 of the report verification unit 106 compares the diagnostic record information including the characteristics recorded in the recording unit 125 and the attribute information thereof with the diagnostic record information stored in the diagnostic information storage unit 103, The degree of matching is determined.
- the flow of this determination process will be described with reference to FIG.
- the report verification unit 106 compares the tissue specimen identification information (ID) accumulated in the diagnostic information accumulation unit 103 with the tissue specimen identification information (ID) of the report analyzed by the report analysis unit 105 (S600). If they match, the report verification unit 106 compares the gaze area (ROI) (S601). If the gaze area (ROI) does not match, the flow ends here (S607).
- the report verification unit 106 compares the diagnosis record information accumulated in the diagnosis information accumulation unit 103 with the gaze area (ROI) in the diagnosis record information of the report analyzed by the report analysis unit 105 (S601). Since the comparison method here is the same as that of the first embodiment, a description thereof will be omitted.
- the comparison unit 116 of the report verification unit 106 starts with the information stored in the diagnosis information storage unit 103 (FIG. 9a) and the features and attribute information (FIG. 10) stored in the report analysis unit 105.
- the features are compared (S602).
- the feature is both “gland duct”. For this reason, the comparison unit 116 determines that the features match and adds +1 as the feature matching score.
- the comparison unit 116 of the report verification unit 106 compares the diagnostic information storage unit 103 and the attribute information of the features recorded in the report analysis unit 105.
- a threshold condition corresponding to each attribute information is set in advance. When the attribute information recorded in the diagnostic record information 103 and the report analysis unit 105 satisfies the threshold condition, the comparison unit 116 determines that the two match.
- the diagnostic information storage unit 103 stores information (FIG. 9a) indicating “density 70%”. This means that the gland duct occupies the entire gaze area (ROI).
- the report analysis unit 105 stores information (FIG. 10) indicating “crowded”.
- the diagnostic information storage unit 103 stores information (FIG. 9a) indicating “size variation: large”, and the report analysis unit 105 stores information indicating that “the sizes are different” ( FIG. 10) is stored. At this time, both satisfy a predetermined threshold condition. Therefore, the report verification unit 106 determines that the two match.
- the score calculation unit 136 of the report verification unit 106 multiplies the feature or attribute information matching item by the matching score of each feature and the weighting value set in advance in the feature-specific importance table recording unit 126.
- the warning score is derived by calculating the total sum (S604).
- the feature score is “+1” because it is determined that the feature is “gland duct”. Further, from the information recorded in the feature-specific importance table recording unit 126, the item “differing in size” in the gland duct has a weighting value +7, and the item “densely” has a weighting value +5.
- the score calculation unit 136 derives a warning score by multiplying two weighting values by the matching score +1 obtained in this way and calculating a sum.
- the warning score calculated here is +12. Note that the score calculation unit 136 may add and / or subtract the warning score calculated as necessary according to the feature in the report sentence and the arrangement order of the attribute information.
- the warning score calculated by the score calculation unit 136 of the report verification unit 106 is sent to the warning display unit 107.
- the warning display unit 107 compares the preset threshold value with the warning score transmitted from the report verification unit 106 (S605).
- the warning display unit 107 displays a warning for the user (S606) and ends the process (S607).
- the warning display unit 107 ends the process as it is (S607). In any case, the warning display unit 107 may display the warning score simultaneously with the warning display (S607).
- the comparison unit 116 of the report verification unit 106 starts with the diagnosis record information stored in the diagnosis information storage unit 103 (FIG. 9b) and the diagnosis record information stored in the report analysis unit 105 (FIG. 10). Compare features.
- the features are “gland duct” and “signature ring cell”, respectively. Therefore, the comparison unit 116 determines that the features do not match and sets the matching score to -1. The result is sent to the warning display unit 107.
- the warning display unit 107 displays a warning that the features used for the pathological diagnosis do not match (S603), and the flow ends as it is (S607). Further, the warning display unit 107 displays that the features do not match, and the features and their attribute information (S606).
- the present embodiment it is possible to prevent the difference between the result of diagnosis of the tissue specimen image and the report content created by the pathologist, and it is possible to evaluate not only the pathological diagnosis result but also the detailed reason.
- Embodiment 3 the diagnosis record information described in the report accumulated in the report accumulation unit 104 is compared with the diagnosis record information accumulated in the diagnosis information accumulation unit 103, and the degree of coincidence between them is determined. .
- the difference from the second embodiment is that there are a plurality of pieces of diagnostic record information stored in the diagnostic information storage unit 103.
- tissue specimen identification information (ID) and the gaze area (ROI) extraction method and comparison method recorded in the diagnostic information storage unit 103 and the report analysis unit 105 are the same as those in the first embodiment, here The description is omitted (S600, S601).
- FIG. 9 a flow for comparing features and their attribute information when there are a plurality of pieces of diagnostic result information recorded in the diagnostic information storage unit 103 will be described with reference to FIGS. 9 and 10.
- FIG. 9 a flow for comparing features and their attribute information when there are a plurality of pieces of diagnostic result information recorded in the diagnostic information storage unit 103 will be described with reference to FIGS. 9 and 10.
- the pathology report stored in the report storage unit 104 states that “the gland ducts of different sizes are dense”. Since the feature and its attribute information are extracted by the same method as in the first embodiment, the description thereof is omitted here.
- the extracted feature is “gland duct”, and the extracted attribute information is “different in size” and “dense”.
- the pathological report (FIG. 10) recorded in the diagnostic information storage unit 103 and stored in the report storage unit 104 matches the pathological specimen identification information (ID) and the gaze area (ROI).
- ID pathological specimen identification information
- ROI gaze area
- the comparison unit 116 of the report verification unit 106 compares the gaze area 1 recorded in the diagnostic information storage unit 103 with the feature described in the pathology report and its attribute information (S602). First, the comparison unit 116 compares the diagnostic record information (FIG. 9 a) stored in the diagnostic information storage unit 103 with the characteristics of the diagnostic record information (FIG. 10) stored in the report analysis unit 105. To do. In this example, the feature is both “gland duct”. Therefore, the comparison unit 116 determines that the features match and sets the match score to +1.
- the comparison unit 116 compares the attribute information of the features.
- the diagnostic information storage unit 103 stores information (FIG. 9a) indicating “density 70%”, and the report analysis unit 105 stores information indicating “crowded” (FIG. 10). .
- the comparison unit 116 determines that they match.
- the comparison unit 116 compares information on the feature size.
- the diagnostic information storage unit 103 stores information indicating that “there is a variation in size”, and the report analysis unit 105 stores information indicating that the size is different (FIG. 10). Has been. Also in this case, using the same method as in the second embodiment, the comparison unit 116 determines that they match.
- the item “different in size” in the gland duct has a weighting value +7
- the item “dense” has a weighting value +5.
- the score calculation unit 136 calculates a warning score using the matching score and the weighted value thus obtained (S604). Since the method for calculating the warning score is the same as that in the second embodiment, description thereof is omitted here.
- the comparison unit 116 compares the diagnostic record information (FIG. 9 b) accumulated in the diagnostic information accumulation unit 103 with the characteristics of the diagnostic record information (FIG. 10) stored in the report analysis unit 105. In this example, the features are different. Therefore, the comparison unit 116 determines that the features do not match and sets the matching score to -1.
- the warning display unit 107 displays that the features do not match, the calculated warning score, the features, and their attribute information. At this time, the warning display unit 107 may display these pieces of information in order from the lowest matching score.
- the comparison unit 116 compares the characteristics of the diagnostic record information (FIG. 9 c) stored in the diagnostic information storage unit 103 and the diagnostic record information (FIG. 10) stored in the report analysis unit 105. This method is also the same as the method for comparing the diagnostic record information stored in the diagnostic information storage unit 103 (FIG. 9b) and the diagnostic record information stored in the report analysis unit 105 (FIG. 10). . Therefore, the description is omitted here.
- the present embodiment it is possible to prevent the difference between the result of diagnosing the tissue specimen image and the report content created by the pathologist, and to evaluate not only the pathological diagnosis result but also the detailed reason.
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Chemical & Material Sciences (AREA)
- General Physics & Mathematics (AREA)
- Pathology (AREA)
- Public Health (AREA)
- Hematology (AREA)
- Molecular Biology (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Food Science & Technology (AREA)
- Urology & Nephrology (AREA)
- Immunology (AREA)
- Optics & Photonics (AREA)
- Biophysics (AREA)
- Medicinal Chemistry (AREA)
- Theoretical Computer Science (AREA)
- Radiology & Medical Imaging (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Databases & Information Systems (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Data Mining & Analysis (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Description
以下、図面を参照して、本発明の実施形態について例示的に詳しく説明する。ただし、以下の実施形態に記載されている構成要素はあくまでも例示であり、本発明の技術的範囲をそれらのみに限定する趣旨のものではない。
レポート検証部106は、まず、組織標本識別情報(ID)を比較する(S300)。組織標本識別情報(ID)が一致したら、次に、レポート検証部106は、注視領域(ROI)を比較する(S301)。組織標本識別情報(ID)が一致しなかった場合には、そこでフローは終了する(S303)。
ステップS300において、組織標本識別情報(ID)が一致したら、次にレポート検証部106は、注視領域(ROI)を比較する(S301)。以下、注視領域(ROI)を比較するフローを説明する。
さらに、レポート検証部106は、診断部102とレポート解析部105に記録されている診断結果を比較する(S302)。両者の診断結果が、「良性」で一致した場合、レポート検証部106は、最終診断結果を「良性」とし、フローが終了する(S303)。
第2の実施形態では、レポート蓄積部104に蓄積されている病理レポートの記載内容を基に病理診断に必要な特徴とその属性情報が抽出される。そして、診断情報蓄積部103に蓄積されている診断記録情報に含まれる特徴とその属性情報とが比較され、両者の一致度が判断される。また、このとき両者の一致度に応じたスコアが算出され、算出されたスコアが予め設定された標準値よりも一定値以上かけ離れている場合には、警告表示が行われる。
レポート検証部106は、診断情報蓄積部103に蓄積された組織標本識別情報(ID)と、レポート解析部105で解析されたレポートの組織標本識別情報(ID)とを比較する(S600)。一致していた場合、続いてレポート検証部106は、注視領域(ROI)を比較する(S601)。注視領域(ROI)が一致していない場合、ここでフローは終了する(S607)。
レポート検証部106は、診断情報蓄積部103に蓄積された診断記録情報と、レポート解析部105で解析されたレポートの診断記録情報のうち注視領域(ROI)とを比較する(S601)。ここでの比較方法は、第1の実施形態と同様のため、説明を省略する。
レポート検証部106の比較部116は、診断情報蓄積部103に蓄積された情報(図9a)と、レポート解析部105に保存されている特徴とその属性情報(図10)のうち、まず最初に特徴を比較する(S602)。本例の場合、特徴は両者ともに「腺管」である。そのため、比較部116は、特徴が一致すると判断し、特徴の一致度スコアとして+1を付与する。
第3実施形態では、レポート蓄積部104に蓄積されているレポートに記載された診断記録情報と、診断情報蓄積部103に蓄積された診断記録情報とが比較され、両者の一致度が判断される。第2実施形態との違いは、診断情報蓄積部103に蓄積されている診断記録情報が複数存在することである。
レポート検証部106の比較部116は、診断情報蓄積部103に記録された注視領域1と、病理レポートに記載された特徴とその属性情報の比較を行う(S602)。まず最初に、比較部116は、診断情報蓄積部103に蓄積された診断記録情報(図9a)と、レポート解析部105に保存されている診断記録情報(図10)のうちの特徴とを比較する。本例の場合、特徴は両者ともに「腺管」である。そのため比較部116は、特徴が一致すると判断し、一致度スコアを+1とする。
比較部116は、診断情報蓄積部103に蓄積された診断記録情報(図9b)と、レポート解析部105に保存されている診断記録情報(図10)のうちの特徴とを比較する。本例の場合、特徴が異なる。そのため比較部116は、特徴が一致しないと判断し、一致度スコアを-1とする。
102 診断部
103 診断情報蓄積部
104 レポート蓄積部
105 レポート解析部
106 レポート検証部
107 警告表示部
112 診断特徴認識部
122 診断特徴解析部
132 診断記録部
115 抽出部
125 記録部
116 比較部
126 特徴別重要度テーブル記録部
136 スコア算出部
Claims (15)
- 組織標本画像の病理診断を行い、診断記録情報を生成する診断部と、
前記組織標本画像の病理診断結果が記載されたレポートを蓄積するレポート蓄積部と、
前記レポート蓄積部に蓄積されているレポートに記載されている診断結果を解析するレポート解析部と、
前記レポート解析部で解析された診断結果と、前記診断記録情報とを比較し、診断結果の一致度を判断するレポート検証部とを備えた
ことを特徴とする病理診断結果判定システム。 - 組織標本画像の病理診断を行い、診断記録情報を蓄積する診断記録蓄積部を備え、
レポート解析部は、レポート蓄積部に蓄積されているレポートから注視領域、特徴名、及びその属性情報を抽出し、
レポート検証部は、前記レポート解析部で抽出された注視領域、特徴名、及び当該特徴の属性情報と、前記診断記録蓄積部の診断記録情報とを比較し、特徴名の一致により一致度スコアを設定し、当該一致度スコアと、特徴名及びその属性情報毎に予め設定された重み付け値を用いて警告スコアを算出する
請求項1に記載の病理診断結果判定システム。 - 重み付け値は、特徴と当該特徴の属性情報に応じて設定される請求項2に記載の病理診断結果判定システム。
- 診断部は、複数の注視領域を抽出し、
レポート検証部は、前記複数の注視領域ごとに診断記録情報を比較し、設定された一致度スコアを用いて警告スコアを算出する
請求項2または請求項3に記載の病理診断結果判定システム。 - 一致していない特徴に加え、一致した特徴を警告スコア順に表示する警告表示部を備えた
請求項4に記載の病理診断結果判定システム。 - 組織標本画像の病理診断を行い、診断記録情報を生成する診断ステップと、
前記組織標本画像の病理診断結果が記載されたレポートをレポート蓄積部に蓄積するレポート蓄積ステップと、
前記レポート蓄積部に蓄積されているレポートに記載されている診断結果を解析するレポート解析ステップと、
前記レポート解析ステップで解析された診断結果と、前記診断記録情報とを比較し、診断結果の一致度を判断するレポート検証ステップとを含む
ことを特徴とする病理診断結果判定方法。 - 組織標本画像の病理診断を行い、診断記録情報を診断記録蓄積部に蓄積する診断記録蓄積ステップを含み、
レポート解析ステップで、レポート蓄積部に蓄積されているレポートから注視領域、特徴名、及びその属性情報を抽出し、
レポート検証ステップで、前記レポート解析ステップで抽出された注視領域、特徴名、及び当該特徴の属性情報と、前記診断記録蓄積部の診断記録情報とを比較し、特徴名の一致により一致度スコアを設定し、当該一致度スコアと、特徴名及びその属性情報毎に予め設定された重み付け値を用いて警告スコアを算出する
請求項6に記載の病理診断結果判定方法。 - 重み付け値は、特徴と当該特徴の属性情報に応じて設定される請求項7に記載の病理診断結果判定方法。
- 診断ステップで、複数の注視領域を抽出し、
レポート検証ステップで、前記複数の注視領域ごとに診断記録情報を比較し、設定された一致度スコアを用いて警告スコアを算出する
請求項7または請求項8に記載の病理診断結果判定方法。 - 一致していない特徴に加え、一致した特徴を警告スコア順に表示する警告表示ステップを含む
請求項9に記載の病理診断結果判定方法。 - 組織標本画像の病理診断を行い、診断記録情報を生成する診断部と、
前記組織標本画像の病理診断結果が記載されたレポートを蓄積するレポート蓄積部と、
前記レポート蓄積部に蓄積されているレポートに記載されている診断結果を解析するレポート解析部と、
前記レポート解析部で解析された診断結果と、前記診断記録情報とを比較し、診断結果の一致度を判断するレポート検証部とを備えた
ことを特徴とする病理診断結果判定装置。 - 組織標本画像の病理診断を行い、診断記録情報を蓄積する診断記録蓄積部を備え、
レポート解析部は、レポート蓄積部に蓄積されているレポートから注視領域、特徴名、及びその属性情報を抽出し、
レポート検証部は、前記レポート解析部で抽出された注視領域、特徴名、及び当該特徴の属性情報と、前記診断記録蓄積部の診断記録情報とを比較し、特徴名の一致により一致度スコアを設定し、当該一致度スコアと、特徴名及びその属性情報毎に予め設定された重み付け値を用いて警告スコアを算出する
請求項11に記載の病理診断結果判定装置。 - 重み付け値は、特徴と当該特徴の属性情報に応じて設定される請求項12に記載の病理診断結果判定装置。
- 診断部は、複数の注視領域を抽出し、
レポート検証部は、前記複数の注視領域ごとに診断記録情報を比較し、設定された一致度スコアを用いて警告スコアを算出する
請求項12または請求項13に記載の病理診断結果判定装置。 - 一致していない特徴に加え、一致した特徴を警告スコア順に表示する警告表示部を備えた
請求項14に記載の病理診断結果判定装置。
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2014516635A JP5958534B2 (ja) | 2012-05-24 | 2013-03-07 | 病理診断結果判定システム、病理診断結果判定方法および病理診断結果判定装置 |
| US14/395,166 US9383347B2 (en) | 2012-05-24 | 2013-03-07 | Pathological diagnosis results assessment system, pathological diagnosis results assessment method, and pathological diagnosis results assessment device |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2012118726 | 2012-05-24 | ||
| JP2012-118726 | 2012-05-24 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2013175683A1 true WO2013175683A1 (ja) | 2013-11-28 |
Family
ID=49623399
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/JP2013/001451 Ceased WO2013175683A1 (ja) | 2012-05-24 | 2013-03-07 | 病理診断結果判定システム、病理診断結果判定方法および病理診断結果判定装置 |
Country Status (3)
| Country | Link |
|---|---|
| US (1) | US9383347B2 (ja) |
| JP (1) | JP5958534B2 (ja) |
| WO (1) | WO2013175683A1 (ja) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015134668A1 (en) * | 2014-03-04 | 2015-09-11 | The Regents Of The University Of California | Automated quality control of diagnostic radiology |
| WO2016130879A1 (en) * | 2015-02-13 | 2016-08-18 | Prairie Ventures, Llc | System and method to objectively measure quality assurance in anatomic pathology |
| CN109564622A (zh) * | 2016-08-12 | 2019-04-02 | 威里利生命科学有限责任公司 | 增强病理诊断 |
| CN112562816A (zh) * | 2020-11-13 | 2021-03-26 | 陈卫霞 | 肿瘤影像报告诊断结果与病理结果对应与评价系统及方法 |
| JP2022049586A (ja) * | 2020-09-16 | 2022-03-29 | BonBon株式会社 | プログラム、情報処理装置、方法 |
Families Citing this family (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5557091B2 (ja) * | 2010-03-30 | 2014-07-23 | 日本電気株式会社 | 画像処理装置、画像読取装置、画像処理方法及び画像処理プログラム |
| KR20160054992A (ko) * | 2014-11-07 | 2016-05-17 | 삼성전자주식회사 | 관심영역의 재검출 회피 장치 및 방법 |
| CN106296638A (zh) * | 2015-06-04 | 2017-01-04 | 欧姆龙株式会社 | 显著性信息取得装置以及显著性信息取得方法 |
| US10269114B2 (en) * | 2015-06-12 | 2019-04-23 | International Business Machines Corporation | Methods and systems for automatically scoring diagnoses associated with clinical images |
| US9858385B2 (en) | 2015-07-23 | 2018-01-02 | International Business Machines Corporation | Identifying errors in medical data |
| JP6897042B2 (ja) * | 2016-09-27 | 2021-06-30 | 日本電気株式会社 | 画像検査装置、画像検査方法および画像検査プログラム |
| JP7224757B2 (ja) * | 2017-10-13 | 2023-02-20 | キヤノン株式会社 | 診断支援装置、情報処理方法、診断支援システム及びプログラム |
| CN108172276A (zh) * | 2017-12-01 | 2018-06-15 | 杭州远舟医疗科技有限公司 | 一种红外影像自动评估方法、电子设备及存储介质 |
| US10832808B2 (en) | 2017-12-13 | 2020-11-10 | International Business Machines Corporation | Automated selection, arrangement, and processing of key images |
| EP3493213B1 (en) * | 2018-02-13 | 2020-12-30 | Siemens Healthcare GmbH | Method, system, computer program product and computer-readable medium for supporting a visual analysis of a medical image data set |
| JP2019170621A (ja) | 2018-03-28 | 2019-10-10 | 富士フイルム株式会社 | 読影支援装置、読影支援プログラム |
| WO2020033593A1 (en) * | 2018-08-07 | 2020-02-13 | Britescan, Llc | Portable scanning device for ascertaining attributes of sample materials |
| US10740901B2 (en) * | 2018-12-17 | 2020-08-11 | Nvidia Corporation | Encoder regularization of a segmentation model |
| KR20230021665A (ko) | 2020-05-08 | 2023-02-14 | 페이지.에이아이, 인크. | 디지털 병리학에서 현저한 정보를 결정하기 위해 전자 이미지들을 처리하는 시스템들 및 방법들 |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2918663B2 (ja) * | 1990-09-06 | 1999-07-12 | 株式会社日立製作所 | 顕微鏡視野内表示装置及びこれを含む顕微鏡画像処理装置 |
| JP2001059842A (ja) * | 1999-08-25 | 2001-03-06 | Nec Corp | 病理診断装置 |
| JP3451307B2 (ja) * | 1994-01-13 | 2003-09-29 | 独立行政法人放射線医学総合研究所 | 染色体画像表示装置 |
| JP3916395B2 (ja) * | 1997-07-17 | 2007-05-16 | アキュメッド インターナショナル インコーポレイテッド | 検体の予備処理機能を備えた検査体系 |
| JP2007192821A (ja) * | 2006-01-20 | 2007-08-02 | Sakura Finetex Usa Inc | 生物学的標本を処理するための自動システムおよび方法 |
| JP2011181015A (ja) * | 2010-03-03 | 2011-09-15 | Olympus Corp | 診断情報配信装置および病理診断システム |
| JP2012073179A (ja) * | 2010-09-29 | 2012-04-12 | Dainippon Screen Mfg Co Ltd | 病理診断支援装置、病理診断支援方法、病理診断支援のための制御プログラムおよび該制御プログラムを記録した記録媒体 |
| WO2012157201A1 (ja) * | 2011-05-18 | 2012-11-22 | 日本電気株式会社 | 情報処理システム、情報処理方法、情報処理装置およびその制御方法と制御プログラム |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6148096A (en) * | 1995-09-15 | 2000-11-14 | Accumed International, Inc. | Specimen preview and inspection system |
| US6091842A (en) * | 1996-10-25 | 2000-07-18 | Accumed International, Inc. | Cytological specimen analysis system with slide mapping and generation of viewing path information |
| US6430309B1 (en) * | 1995-09-15 | 2002-08-06 | Monogen, Inc. | Specimen preview and inspection system |
| US20030215936A1 (en) * | 2000-12-13 | 2003-11-20 | Olli Kallioniemi | High-throughput tissue microarray technology and applications |
| JP2010035756A (ja) | 2008-08-04 | 2010-02-18 | Fujifilm Corp | 診断支援装置及び診断支援方法 |
-
2013
- 2013-03-07 US US14/395,166 patent/US9383347B2/en not_active Expired - Fee Related
- 2013-03-07 WO PCT/JP2013/001451 patent/WO2013175683A1/ja not_active Ceased
- 2013-03-07 JP JP2014516635A patent/JP5958534B2/ja active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2918663B2 (ja) * | 1990-09-06 | 1999-07-12 | 株式会社日立製作所 | 顕微鏡視野内表示装置及びこれを含む顕微鏡画像処理装置 |
| JP3451307B2 (ja) * | 1994-01-13 | 2003-09-29 | 独立行政法人放射線医学総合研究所 | 染色体画像表示装置 |
| JP3916395B2 (ja) * | 1997-07-17 | 2007-05-16 | アキュメッド インターナショナル インコーポレイテッド | 検体の予備処理機能を備えた検査体系 |
| JP2001059842A (ja) * | 1999-08-25 | 2001-03-06 | Nec Corp | 病理診断装置 |
| JP2007192821A (ja) * | 2006-01-20 | 2007-08-02 | Sakura Finetex Usa Inc | 生物学的標本を処理するための自動システムおよび方法 |
| JP2011181015A (ja) * | 2010-03-03 | 2011-09-15 | Olympus Corp | 診断情報配信装置および病理診断システム |
| JP2012073179A (ja) * | 2010-09-29 | 2012-04-12 | Dainippon Screen Mfg Co Ltd | 病理診断支援装置、病理診断支援方法、病理診断支援のための制御プログラムおよび該制御プログラムを記録した記録媒体 |
| WO2012157201A1 (ja) * | 2011-05-18 | 2012-11-22 | 日本電気株式会社 | 情報処理システム、情報処理方法、情報処理装置およびその制御方法と制御プログラム |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015134668A1 (en) * | 2014-03-04 | 2015-09-11 | The Regents Of The University Of California | Automated quality control of diagnostic radiology |
| US10438347B2 (en) | 2014-03-04 | 2019-10-08 | The Regents Of The University Of California | Automated quality control of diagnostic radiology |
| WO2016130879A1 (en) * | 2015-02-13 | 2016-08-18 | Prairie Ventures, Llc | System and method to objectively measure quality assurance in anatomic pathology |
| CN109564622A (zh) * | 2016-08-12 | 2019-04-02 | 威里利生命科学有限责任公司 | 增强病理诊断 |
| CN109564622B (zh) * | 2016-08-12 | 2023-09-01 | 威里利生命科学有限责任公司 | 增强病理诊断 |
| JP2022049586A (ja) * | 2020-09-16 | 2022-03-29 | BonBon株式会社 | プログラム、情報処理装置、方法 |
| CN112562816A (zh) * | 2020-11-13 | 2021-03-26 | 陈卫霞 | 肿瘤影像报告诊断结果与病理结果对应与评价系统及方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| JPWO2013175683A1 (ja) | 2016-01-12 |
| JP5958534B2 (ja) | 2016-08-02 |
| US9383347B2 (en) | 2016-07-05 |
| US20150072371A1 (en) | 2015-03-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5958534B2 (ja) | 病理診断結果判定システム、病理診断結果判定方法および病理診断結果判定装置 | |
| Li et al. | A comprehensive review of computer-aided whole-slide image analysis: from datasets to feature extraction, segmentation, classification and detection approaches | |
| Koh et al. | Detection of meibomian glands and classification of meibography images | |
| US20220207730A1 (en) | Systems and Methods for Automated Image Analysis | |
| KR102531400B1 (ko) | 인공 지능 기반 대장 내시경 영상 진단 보조 시스템 및 방법 | |
| TW202105245A (zh) | 用於分析活檢樣本影像以決定其癌變機率之方法 | |
| CN107230198B (zh) | 胃镜图像智能处理方法及装置 | |
| Diao et al. | Weakly supervised framework for cancer region detection of hepatocellular carcinoma in whole-slide pathologic images based on multiscale attention convolutional neural network | |
| JP2016087370A (ja) | 大腸拡大内視鏡画像診断支援装置 | |
| Zhang et al. | Deep learning‐based detection and classification of lumbar disc herniation on magnetic resonance images | |
| Srivastava et al. | Computer-aided identification of ovarian cancer in confocal microendoscope images | |
| US10430905B2 (en) | Case search device and method | |
| Kalbhor et al. | DeepCerviCancer-deep learning-based cervical image classification using colposcopy and cytology images | |
| Geva et al. | Pneumothorax detection in chest radiographs using local and global texture signatures | |
| Tang et al. | Splat feature classification: Detection of the presence of large retinal hemorrhages | |
| Govind et al. | Automated erythrocyte detection and classification from whole slide images | |
| Corredor et al. | Training a cell-level classifier for detecting basal-cell carcinoma by combining human visual attention maps with low-level handcrafted features | |
| JP4925111B2 (ja) | 類似画像検索装置 | |
| KR20130082849A (ko) | 진단 근거를 제공하는 의료영상 진단 장치 및 방법 | |
| Yap et al. | Breast ultrasound lesions classification: a performance evaluation between manual delineation and computer segmentation | |
| Guo et al. | Anatomical landmark segmentation in uterine cervix images using deep learning | |
| Li et al. | Comparison of deep learned and texture features in mammographic mass classification | |
| Gandomkar et al. | Predicting radiologists' true and false positive decisions in reading mammograms by using gaze parameters and image-based features | |
| Fuhrman et al. | Detection and classification of coronary artery calcifications in low dose thoracic CT using deep learning | |
| US20220277451A1 (en) | Systems, methods and apparatuses for visualization of imaging data |
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: 13794460 Country of ref document: EP Kind code of ref document: A1 |
|
| ENP | Entry into the national phase |
Ref document number: 2014516635 Country of ref document: JP Kind code of ref document: A |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 14395166 Country of ref document: US |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 13794460 Country of ref document: EP Kind code of ref document: A1 |