Papers › CASS: Cross Architectural Self-Supervision for Medical Image Analysis

CASS: Cross Architectural Self-Supervision for Medical Image Analysis

8 Jun 2022arXiv:2206.04170archive 2025-07-28

Pranav Singh, Elena Sizikova, Jacopo Cirrone

Recent advances in deep learning and computer vision have reduced many barriers to automated medical image analysis, allowing algorithms to process label-free images and improve performance. However, existing techniques have extreme computational requirements and drop a lot of performance with a reduction in batch size or training epochs. This paper presents Cross Architectural - Self Supervision (CASS), a novel self-supervised learning approach that leverages Transformer and CNN simultaneously. Compared to the existing state of the art self-supervised learning approaches, we empirically show that CASS-trained CNNs and Transformers across four diverse datasets gained an average of 3.8% with 1% labeled data, 5.9% with 10% labeled data, and 10.13% with 100% labeled data while taking 69% less time. We also show that CASS is much more robust to changes in batch size and training epochs. Notably, one of the test datasets comprised histopathology slides of an autoimmune disease, a condition with minimal data that has been underrepresented in medical imaging. The code is open source and is available on GitHub.

Paper PDF Code

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

pranavsinghps1/CASS officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Classification Medical Image Analysis Medical Image Classification Partial Label Learning Self-Supervised Learning

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Classification	Autoimmune Dataset	CASS	F1 score	0.8894	#2 of 4	Archive leaderboard	report
Classification	Autoimmune Dataset	DINO	F1 score	0.8639	#3 of 4	Archive leaderboard	report
Classification	Brain Tumor MRI Dataset	DINO	F1 score	0.9909	#1 of 1	Archive leaderboard	report
Classification	ISIC 2019	CASS	Balanced Multi-Class Accuracy	0.6519	#1 of 1	Archive leaderboard	report
Partial Label Learning	Autoimmune Dataset	CASS	F1 score	0.8717	#1 of 2	Archive leaderboard	report
Partial Label Learning	Autoimmune Dataset	DINO	F1 score	0.8445	#2 of 2	Archive leaderboard	report
Partial Label Learning	ISIC 2019	CASS	Balanced Multi-Class Accuracy	0.7258	#1 of 1	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position Encodings Adam Attention BPE Dense Connections Dropout Label Smoothing Layer Normalization Linear Layer Multi-Head Attention Position-Wise Feed-Forward Layer Residual Connection Softmax Test Transformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections