Skip to main navigation Skip to search Skip to main content

Contrastive Learning Under Noisy Temporal Self-supervision for Colonoscopy Videos

  • Luca Parolari
  • , Pietro Gori
  • , Lamberto Ballan
  • , Carlo Biffi
  • , Loic Le Folgoc
  • University of Padova
  • Institut Polytechnique de Paris
  • Cosmo Intelligent Medical Devices

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Learning robust representations of polyp tracklets is key to enabling multiple AI-assisted colonoscopy applications, from polyp characterization to automated reporting and retrieval. Supervised contrastive learning is an effective approach for learning such representations, but it typically relies on correct positive and negative definitions. Collecting these labels requires linking tracklets that depict the same underlying polyp entity throughout the video, which is costly and demands specialized clinical expertise. In this work, we leverage the sequential workflow of colonoscopy procedures to derive self-supervised associations from temporal structure. Since temporally derived associations are not guaranteed to be correct, we introduce a noise-aware contrastive loss to account for noisy associations. We demonstrate the effectiveness of the learned representations across multiple downstream tasks, including polyp retrieval and re-identification, size estimation, and histology classification. Our method outperforms prior self-supervised and supervised baselines, and matches or exceeds recent foundation models across all tasks, using a lightweight encoder trained on only 27 videos. Code is available at github.com/lparolari/ntssl.

Original languageEnglish
Title of host publicationMedical Image Computing and Computer Assisted Intervention – MICCAI 2026 - 29th International Conference, 2026, Proceedings
EditorsGuang Yang, Ehsan Adeli, Marleen de Bruijne, Marleen de Bruijne, Bartlomiej W. Papiez, Stefanie Speidel, Pallavi Tiwari, Guoyan Zheng, Mohammad Yaqub, Qi Dou, Islem Rekik
PublisherSpringer Science and Business Media Deutschland GmbH
Pages67-77
Number of pages11
ISBN (Print)9783032382320
DOIs
Publication statusPublished - 1 Jan 2027
Event29th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2026 - Strasbourg, France
Duration: 27 Sept 2026 → 1 Oct 2026

Publication series

NameLecture Notes in Computer Science
Volume16892 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference29th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2026
Country/TerritoryFrance
CityStrasbourg
Period27/09/26 → 1/10/26

Keywords

  • Colonoscopy Videos
  • Contrastive Learning
  • Representation Learning
  • Temporal Self-Supervision

Fingerprint

Dive into the research topics of 'Contrastive Learning Under Noisy Temporal Self-supervision for Colonoscopy Videos'. Together they form a unique fingerprint.

Cite this