# Knee Osteoarthritis Grading on OAI (official held-out test) Source: [Knee Osteoarthritis Grading on OAI (official held-out test)]() Collection index: [llms.txt]() Data updated: 2026-04-28T07:53:40.076Z Evaluation of model performance for grading Knee Osteoarthritis using the official held-out test split of the OAI dataset, measured using macro-averaged metrics like Accuracy, Precision, Recall, F1, Specificity, and AUROC. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Knee Osteoarthritis Grading | | Dataset (including any stated split) | OAI (official held-out test) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 15 of 15 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy](): Higher is better; unit: PERCENT. - [Precision (macro)](): Higher is better; unit: PERCENT. - [Recall (macro)](): Higher is better; unit: PERCENT. - [F1 Score (macro)](): Higher is better; unit: PERCENT. - [Specificity (macro)](): Higher is better; unit: PERCENT. - [AUROC (macro)](): Higher is better; unit: SCALAR. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | H-SemiS | 0.849 | Not recorded | 69f03eed55a3c32af89f9f15 | | Mean-Teacher | 0.803 | Not recorded | 69f03eed55a3c32af89f9f15 | | π-Model | 0.795 | Not recorded | 69f03eed55a3c32af89f9f15 | | MedViT | 0.769 | Not recorded | 69f03eed55a3c32af89f9f15 | | DenseNet201 | 0.748 | Not recorded | 69f03eed55a3c32af89f9f15 | | CMT | 0.732 | Not recorded | 69f03eed55a3c32af89f9f15 | | ResNet50 | 0.728 | Not recorded | 69f03eed55a3c32af89f9f15 | | ViT-B/32 | 0.704 | Not recorded | 69f03eed55a3c32af89f9f15 | | ViT–CORAL | 0.7029 | Not recorded | 695c96f453fdd5c36174355d | | VGG-19 | 0.696 | Not recorded | 695c96f453fdd5c36174355d | | ResNet101 | 0.69 | Not recorded | 695c96f453fdd5c36174355d | | InceptionV3 | 0.684 | Not recorded | 695c96f453fdd5c36174355d | | DenseNet121 | 0.682 | Not recorded | 695c96f453fdd5c36174355d | | MobileNetV2 | 0.67 | Not recorded | 695c96f453fdd5c36174355d | | Deep Siamese CNN | 0.6671 | Not recorded | 695c96f453fdd5c36174355d | ### H-SemiS Accuracy: 0.849. Result ID: 69f0677b47f34ce924d61132. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Mean-Teacher Accuracy: 0.803. Result ID: 69f0677b47f34ce924d61131. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### π-Model Accuracy: 0.795. Result ID: 69f0677b47f34ce924d61130. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MedViT Accuracy: 0.769. Result ID: 69f0677b47f34ce924d6112e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DenseNet201 Accuracy: 0.748. Result ID: 69f0677b47f34ce924d6112c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### CMT Accuracy: 0.732. Result ID: 69f0677b47f34ce924d6112f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ResNet50 Accuracy: 0.728. Result ID: 69f0677b47f34ce924d6112b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ViT-B/32 Accuracy: 0.704. Result ID: 69f0677b47f34ce924d6112d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ViT–CORAL Accuracy: 0.7029. Result ID: 6997f99c5698665ec4531036. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Ensemble inference":"true","TTA":"true","Ordinal threshold tuning":"true"}. This record does not assert independent reproduction. ### VGG-19 Accuracy: 0.696. Result ID: 6997f99c5698665ec4531033. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### ResNet101 Accuracy: 0.69. Result ID: 6997f99c5698665ec4531031. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### InceptionV3 Accuracy: 0.684. Result ID: 6997f99c5698665ec4531035. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### DenseNet121 Accuracy: 0.682. Result ID: 6997f99c5698665ec4531034. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### MobileNetV2 Accuracy: 0.67. Result ID: 6997f99c5698665ec4531032. [Reporting paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### Deep Siamese CNN Accuracy: 0.6671. Result ID: 6997f99c5698665ec4531030. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.