SAMI3D-DW improves interactive 3D medical image segmentation accuracy
SAMI3D-DW: Interactive Segmentation of Any 3D Medical Images
Computer Vision and Pattern Recognition
Summary
Segmenting 3D medical images accurately is important but challenging, especially for complex body parts and diseases. The authors present SAMI3D-DW, a model trained on a large set of medical scans, that helps users segment these images with just a few clicks. It performs better than previous methods on many types of anatomy and pathologies, making it faster and easier to analyze medical scans. This could help doctors quickly measure tumors and other structures for diagnosis and treatment monitoring.
What this means in practice
- •For radiologists and clinicians: Enable fast and accurate segmentation of complex anatomical structures in CT and MR scans with minimal user input.
- •For medical imaging software developers: Incorporate SAMI3D-DW to provide interactive 3D segmentation tools that reduce annotation time and improve accuracy for diverse clinical targets.$Commercial implications: Supports development of commercial medical imaging products that speed up diagnosis and treatment assessment.
Authors
Ping Gong, Shiyuan Su, Fandong Zhang, Xinchen Han, Haowei Sun, Yiming Li, Yizhou Yu
Abstract
Interactive segmentation of 3D medical images supports quantitative analysis of anatomical structures and disease while allowing users to specify and refine their targets. Despite substantial progress by nnInteractive and VISTA3D, reliable segmentation across diverse clinical targets remains challenging, particularly for complex anatomical structures and the heterogeneous, long-tailed spectrum of pathology. We present SAMI3D-DW V1 (hereafter SAMI3D-DW), an interactive 3D segmentation model trained on Deepwise's large-scale proprietary medical image datasets. We evaluate the model under simulated user interactions on a CT/MR benchmark comprising 4,326 cases from 219 source datasets, spanning 107 anatomical and pathological categories, organized by a medical taxonomy and evaluated with a category-balanced DSC score. SAMI3D-DW achieves the highest category-macro Dice among evaluated methods in both interaction modes. With one point, it scores 0.5764 versus 0.5315 for nnInteractive, the strongest baseline, rising to 0.7771 versus 0.7494 with five points. With bounding-box initialization, the scores are 0.7130 versus 0.6530. After five corrective clicks, SAMI3D-DW reaches 0.8002 versus 0.7868, making it the only evaluated box-compatible model to exceed 0.80. For radiologists and clinicians, SAMI3D-DW enables segmentation of complex anatomical structures, including intracranial vessel trees on CT and MR angiography, with a few clicks. In a preliminary in-house comparison involving neurofibromatosis type 1 (NF1), SAMI3D-DW-assisted tumor annotation took minutes per case and approximately one-fifteenth of the time required for manual annotation, highlighting its potential to support volumetric treatment-response assessment.