Codesota · Models · TorchXRayVisionCohen Lab6 results · 6 benchmarks
Model card

TorchXRayVision.

Cohen Labopen-sourceDenseNet-121 / ResNetApache 2.02 current SOTA

Pre-trained on 8 datasets (MIMIC, CheXpert, NIH, etc.). Unified 18-pathology output.

§ 01 · Benchmarks

Every benchmark TorchXRayVision has a recorded score for.

#BenchmarkArea · TaskMetricValueRankDateSource
01NIH ChestX-ray14Medical · Disease Classificationauroc85.8%#1/4source ↗
02PadChestMedical · Disease Classificationauroc84.6%#1/1source ↗
03COVID-19 Image Data CollectionMedical · Disease Classificationauroc93.2%#2/2source ↗
04MIMIC-CXRMedical · Disease Classificationauroc86.3%#2/3source ↗
05VinDr-CXRMedical · Disease Classificationauroc87.9%#2/2source ↗
06CheXpertMedical · Disease Classificationauroc87.4%#6/7source ↗
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 02 · Strengths by area

Where TorchXRayVision actually performs.

Medical
6
benchmarks
avg rank #2.3 · 2 SOTA
§ 05 · Sources & freshness

Where these numbers come from.

github-readme
5
results
research-paper
1
result
0 of 6 rows marked verified.