English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Purpose: Depth estimation in robotic surgery is vital in 3D reconstruction,surgical navigation and augmented reality visualization. Although thefoundation model exhibits outstanding performance in many vision tasks,including depth estimation (e.g., DINOv2), recent works observed itslimitations in medical and surgical domain-specific applications. This workpresents a low-ranked adaptation (LoRA) of the foundation model for surgicaldepth estimation. Methods: We design a foundation model-based depth estimationmethod, referred to as Surgical-DINO, a low-rank adaptation of the DINOv2 fordepth estimation in endoscopic surgery. We build LoRA layers and integrate theminto DINO to adapt with surgery-specific domain knowledge instead ofconventional fine-tuning. During training, we freeze the DINO image encoder,which shows excellent visual representation capacity, and only optimize theLoRA layers and depth decoder to integrate features from the surgical scene.Results: Our model is extensively validated on a MICCAI challenge dataset ofSCARED, which is collected from da Vinci Xi endoscope surgery. We empiricallyshow that Surgical-DINO significantly outperforms all the state-of-the-artmodels in endoscopic depth estimation tasks. The analysis with ablation studieshas shown evidence of the remarkable effect of our LoRA layers and adaptation.Conclusion: Surgical-DINO shed some light on the successful adaptation of thefoundation models into the surgical domain for depth estimation. There is clearevidence in the results that zero-shot prediction on pre-trained weights incomputer vision datasets or naive fine-tuning is not sufficient to use thefoundation model in the surgical domain directly. Code is available athttps://github.com/BeileiCui/SurgicalDINO.

Surgical-DINO: Adapter Learning of Foundation Models for Depth Estimation in Endoscopic Surgery