![]() Qualitative Evaluation Retinal Specialist QualificationsĮTable 3. Grader Experience and Set AssignmentĮTable 2. Distribution of Dice Similarity Coefficients (DSC) for All Scans and Stratified by Scan SubgroupĮTable 1. Distribution of Dice Similarity Coefficients (DSC) Stratified by 4 Ascending Mean Grader Volume BucketsĮFigure 18. Relationship Between Intergrader DSC and the Model-Grader DSC per FeatureĮFigure 17. Matrix Showing the Number of Scans Segmented by the Model and Segmented by Neither, One or Both Human Expert Graders, per FeatureĮFigure 16. Bland-Altman Plots Comparing Volumes of Individual Features Segmented Between GradersĮFigure 15. Stacked Bar Chart Showing Distribution of the Stack Rank Positions for the Expert Gradings and the Model SegmentationsĮFigure 14. Example of Disagreement Between Specialists for Qualitative EvaluationĮFigure 13. Example 2 of a Disagreement Between Specialists for Qualitative EvaluationĮFigure 12. Example 1 of a Model Failure and Large Disagreement in IRF/SRF VolumeĮFigure 11. Example of Model Failure and Largest Disagreement in PED Volume Between the Model and Expert GradersĮFigure 10. Example of Large Disagreement in SHRM and PED Volume Between Model and Manual GradingsĮFigure 9. Example of a Challenging Case Resulting in Large Disagreement in SRF Volume Between Model and Manual GradingsĮFigure 8. ![]() Example of a Scan Where All 3 Gradings Were Rated HighlyĮFigure 7. Example 2 of a Scan Where the Model Was Ranked as Most RepresentativeĮFigure 6. Example 1 of a Scan Where the Model Was Qualitatively Rated Higher Than the 2 Manual SegmentationsĮFigure 5. Example of a Scan Where the Model Was Qualitatively Rated Higher Than the 2 Manual SegmentationsĮFigure 4. Custom Viewer Used to Assess the 3 Segmentations for the Qualitative EvaluationĮFigure 3. Grading Protocol and Disease Severity Grading DefinitionĮFigure 1. Set 2: 164 OCT scans acquired using the Topcon device or a Spectralis OCT device from Heidelberg Engineering GmbH, resulting in 4 subsets: (1) Topcon-AMD, (2) Heidelberg-AMD, (3) Topcon-DME, and (4) Heidelberg-DME.).ĮMethods. (Note: Set 1: 15 OCT scans from patients with new severe AMD imaged using the 3D OCT-2000 device from Topcon Corporation. This video illustrates example model segmentations, model success cases, model failures, and examples of model-specialist disagreements when retinal experts graded automated vs manual OCT segmentations for intraretinal fluid (IRF) and subretinal fluid (SRF) (for DME scans) or for IRF, SRF, subreintal hyperreflective material, and pigment epithelial detachment (PED) (for AMD scans).Ĭlick the Related Article link for complete study details, and the Related Article Supplement PDF for full video image findings (eFigures 2 to 12 in the Supplement). ![]() Researchers developed a deep learning model to quantify volumes of clinically relevant pathology in OCT scans in individuals with AMD and DME. OCT segmentation offers the potential to objectively quantify disease burden and standardize treatment decisions. ![]() Treatment decisions for retinal conditions like wet age-related macular degeneration (AMD) and diabetic macular edema (DME) rely on subjective assessment of retinal fluid as a marker of disease severity. ![]()
0 Comments
Leave a Reply. |