Unpaired multi-modal multi-label learning for detecting endometriosis signs

Zhang Y, Yuan Zhang, Hu Wang, Wang H, Yutong Xie, Xie Y, Minh-Son To, To MS, Minh‐Son To, Knox S, Steven Knox, Leonardi M, Mathew Leonardi, Condous G, George Condous, G. Condous, Jodie Avery, Jodie C Avery, Avery JC, Hull ML, M. Louise Hull, Carneiro G, Gustavo Carneiro
Artificial intelligence in medicine · 2026 · vol. 181 , pp. 103503 · doi:10.1016/j.artmed.2026.103503 · PMID:42594747 · W7202069939
article OA: hybrid CC0
🔓 Open OA copy View on OpenAlex View on PubMed View at publisher
⚙ AI-generated summary by gemini-2.5-flash-lite, 2026-08-12 ⓘ

This paper introduces an unpaired multi-modal learning framework designed to detect endometriosis by integrating diverse data sources without direct sample matching.

One-sentence paraphrase of the abstract; not a substitute for reading it. No clinical advice. How this works

Abstract

Endometriosis is a widespread gynecological disorder causing severe pain and infertility, with diagnosis currently relying on slow, costly, and risky laparoscopy. This highlights the critical need for non-invasive imaging diagnostics using transvaginal ultrasound (TVUS) and magnetic resonance imaging (MRI). A key challenge is that patients typically receive only one scan modality in practice, despite TVUS and MRI offering differing diagnostic strengths for endometriosis signs like Pouch of Douglas (POD) obliteration and bowel nodules (BN). Previous work partially addressed this challenge by leveraging unpaired multi-modal data for detecting a single marker: Pouch of Douglas (POD) obliteration. However, this is restrictive because endometriosis signs, such as POD obliteration and bowel nodules (BN), often provide correlated diagnostic cues. Capturing these correlations is essential for accurate detection of endometriosis imaging signs, particularly when combined with multi-modal learning, as each modality offers complementary strengths for different signs. To overcome these limitations, we propose EndoFusion, a novel unpaired multi-modal, multi-label learning framework that enables the detection of POD and BN from TVUS and MRI. Our approach introduces three key innovations: (1) label-based pairing, mixup, and cross-modal feature exchange for robust single-modality inference; (2) Dynamic Mutual Knowledge Distillation (DMKD), which adaptively selects teachers using a worst-student-oriented strategy for effective cross-modal transfer; and (3) label correlations modeling with multi-head attention and a specialized loss to handle imbalance and boost accuracy. This design ensures that knowledge from the superior modality and from co-occurring signs is effectively transferred, mitigating modality-specific weaknesses and improving robustness in imaging sign detection. Experiments on our endometriosis dataset show that our method significantly outperforms all comparison methods, achieving an average AUC of 0.827 (95% CI: 0.790-0.861) when evaluated using single-modality inference. These results represent an initial proof-of-concept toward multi-modal, non-invasive assessment of selected endometriosis imaging signs from MRI and TVUS.

My notes (saved in your browser only)

MeSH descriptors

Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis Endometriosis

Citation neighborhood

Papers in the corpus that this work cites (lower rings, blue) and that cite this one (upper rings, green). Dot size scales with the paper's in-corpus citation count — bigger dot = more influential within the endo/adeno field. Click a dot to open that paper. [ expand to 2 hops ] — adds papers reached through this work's immediate citers/citees. Heavier; up to 60 extra dots.

References (33)

Source provenance

europepmc
last seen: 2026-09-26T06:15:21.694264+00:00
openalex
last seen: 2026-09-26T06:07:23.847345+00:00
pubmed
last seen: 2026-09-26T06:09:14.188871+00:00
unpaywall
last seen: 2026-09-09T06:31:01.691765+00:00
License: CC0 · commercial use OK