Part-aware Personalized Segment Anything Model for Patient-Specific Segmentation
Keywords: Precision Medicine; Patient-Specific Segmentation; Out-of-Distribution Patient Adaptation
Abstract: Precision medicine, such as patient-adaptive treatments utilizing medical images, poses new challenges for image segmentation algorithms due to (1) the large variability across different patients and (2) the limited availability of annotated data for each patient. In this work, we propose a data-efficient segmentation method to address these challenges, namely $\textit{\textbf{P}art-aware}$ $\textit{\textbf{P}ersonalized}$ $\textit{\textbf{S}egment}$ $\textit{\textbf{A}nything}$ $\textit{\textbf{M}odel}$ ($\mathbf{{P}^{2}SAM}$). Without any model fine-tuning, enables seamless adaptation to any new patients relying only on one-shot patient-specific data. We introduce a novel part-aware prompt mechanism to select multiple-point prompts based on part-level features of the one-shot data, which can be extensively integrated into different promptable segmentation models, such as SAM and SAM 2. To further promote the robustness of the selected part-aware prompt, we propose a distribution-similarity-based retrieval approach to determine the optimal number of part-level features for a specific case. $\text{P}^{\text{2}}\text{SAM}$ improves the performance by $\texttt{+} 8.0$% and $\texttt{+} 2.0$% mean Dice score within two patient-specific segmentation tasks, and exhibits impressive generality across different domains, $\textit{e.g.}$, $\texttt{+} 6.4$% mIoU on the PerSeg benchmark. Code will be released upon acceptance.
Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 967
Loading