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

The Segment Anything Model (SAM) serves as a fundamental model for semanticsegmentation and demonstrates remarkable generalization capabilities across awide range of downstream scenarios. In this empirical study, we examine SAM'srobustness and zero-shot generalizability in the field of robotic surgery. Wecomprehensively explore different scenarios, including prompted and unpromptedsituations, bounding box and points-based prompt approaches, as well as theability to generalize under corruptions and perturbations at five severitylevels. Additionally, we compare the performance of SAM with state-of-the-artsupervised models. We conduct all the experiments with two well-known roboticinstrument segmentation datasets from MICCAI EndoVis 2017 and 2018 challenges.Our extensive evaluation results reveal that although SAM shows remarkablezero-shot generalization ability with bounding box prompts, it struggles tosegment the whole instrument with point-based prompts and unprompted settings.Furthermore, our qualitative figures demonstrate that the model either failedto predict certain parts of the instrument mask (e.g., jaws, wrist) orpredicted parts of the instrument as wrong classes in the scenario ofoverlapping instruments within the same bounding box or with the point-basedprompt. In fact, SAM struggles to identify instruments in complex surgicalscenarios characterized by the presence of blood, reflection, blur, and shade.Additionally, SAM is insufficiently robust to maintain high performance whensubjected to various forms of data corruption. We also attempt to fine-tune SAMusing Low-rank Adaptation (LoRA) and propose SurgicalSAM, which shows thecapability in class-wise mask prediction without prompt. Therefore, we canargue that, without further domain-specific fine-tuning, SAM is not ready fordownstream surgical tasks.

SAM Meets Robotic Surgery: An Empirical Study on Generalization, Robustness and Adaptation