Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Reverse Classification Accuracy: Predicting Segmentation Performance in the Absence of Ground Truth.

Vanya V Valindria, Ioannis Lavdas, Wenjia Bai

    IEEE Transactions on Medical Imaging
    |April 25, 2017
    PubMed
    Summary

    Related Concept Videos

    Receiver Operating Characteristic Plot01:15

    Receiver Operating Characteristic Plot

    434
    A ROC (Receiver Operating Characteristic) plot is a graphical tool used to assess the performance of a binary classification model by illustrating the trade-off between sensitivity (true positive rate) and specificity (false positive rate). By plotting sensitivity against 1 - specificity across various threshold settings, the ROC curve shows how well the model distinguishes between classes, with a curve closer to the top-left corner indicating a more accurate model. The area under the ROC curve...
    434

    You might also read

    Related Articles

    Articles linked to this work by shared authors, journal, and citation graph.

    Sort by
    Same author

    Development of a target product profile for artificial intelligence in diabetic eye screening in England: a modified Delphi consensus study.

    The Lancet. Digital health·2026
    Same author

    Disparate privacy risks from medical AI.

    Nature·2026
    Same author

    A computational model to describe multi-regional brain architecture during neurodegeneration in Alzheimer's disease.

    Scientific reports·2026
    Same author

    Latent Causal Modeling for 3D Brain MRI Counterfactuals.

    Deep generative models : 5th MICCAI workshop, DGM4MICCAI 2025, held in conjunction with MICCAI 2025, Daejeon, South Korea, September 23, 2025, Proceedings. DGM4MICCAI (Workshop) (5th : 2025 : Taejon-si, Korea)·2026
    Same author

    Average Calibration Losses for Reliable Uncertainty in Medical Image Segmentation.

    IEEE transactions on medical imaging·2026
    Same author

    A radio-genomics biomarker for precision epidermal growth factor receptor mutation targeting therapy in non-small cell lung cancer.

    Scientific reports·2026

    Assessing automatic segmentation accuracy without ground truth is challenging. Reverse Classification Accuracy (RCA) predicts segmentation performance on new clinical data, enabling reliable deployment of AI tools in healthcare.

    Area of Science:

    • Medical image analysis
    • Computational pathology
    • Artificial intelligence in medicine

    Background:

    • Clinical integration of automatic segmentation tools requires performance assessment on new data.
    • Lack of ground truth hinders accuracy evaluation after deployment.
    • Cross-validation may overestimate performance due to overfitting during development.

    Purpose of the Study:

    • Introduce Reverse Classification Accuracy (RCA) as a framework to predict segmentation performance on new data without ground truth.
    • Enable reliable detection of automatic segmentation method failures in clinical practice.

    Main Methods:

    • Utilize predicted segmentation from a new image to train a reverse classifier.
    • Evaluate the reverse classifier on reference images with available ground truth.

    Related Experiment Videos

  • Validate the RCA approach on multi-organ segmentation using diverse classifiers and methods.

    • Main Results:

    • Demonstrated the feasibility of predicting individual segmentation quality without ground truth.
    • Confirmed that good quality segmentations lead to strong reverse classifier performance.
    • RCA successfully predicted segmentation performance across different methods and classifiers.

    Conclusions:

    • Reverse Classification Accuracy (RCA) provides a robust method for evaluating segmentation performance in the absence of ground truth.
    • RCA is suitable for clinical routine integration and large-scale image analysis.
    • Enables reliable deployment and monitoring of automatic segmentation tools in healthcare settings.