Tissue tracking under long-horizon occlusions with contrastive learning - Takeaways - MDSpire

Tissue tracking under long-horizon occlusions with contrastive learning

  • By

  • Inglezou, Myrto

  • Kegkeroglou, Nikolaos

  • Delimpasis, Leonidas

  • Chatzakos, Panagiotis

  • Porichis, Antonios

  • March 6, 2026

  • 0 min

Share

  • 1

    The proposed methodology enhances tissue tracking during surgery by integrating dense optical flow estimation with camera localization.

  • 2

    A self-supervised template matching method using contrastive learning is developed to improve tissue re-identification under occlusions.

  • 3

    The approach effectively mitigates localization drift without scene mapping, ensuring continuous tracking over long durations.

  • 4

    Validation of the method is conducted on the public SurgT benchmark and a synthetic dataset designed for long-horizon occlusions.

  • 5

    Existing tracking methods struggle with long-horizon occlusions, highlighting the significance of the proposed solution in surgical environments.

Original Source(s)

Tissue tracking under long-horizon occlusions with contrastive learning

  • by Inglezou, Myrto, Kegkeroglou, Nikolaos, Delimpasis, Leonidas, Chatzakos, Panagiotis, Porichis, Antonios

  • March 6, 2026

Related Content

Anatomical, functional, and patient-reported outcomes following anterior urethroplasty. Can we predict when and why are patients with anatomical recurrences requiring reinterventions?

Layered precision suturing vs. traditional double-layer closure at cesarean: a randomized trial of uterine scar healing and maternal outcomes

Outcomes, Indications, and Complications of Revisional Surgery Following Biliopancreatic Diversion in a High-Volume Surgical Center