文献类型：期刊文献

英文题名：A new low-rank adaptation method for brain structure and metastasis segmentation via decoupled principal weight direction and magnitude

作者：Zhu, Hancan[1,2];Yang, Hongxia[1];Wang, Yaqing[2];Hu, Keli[2];He, Guanghua[1];Zhou, Jia[3];Li, Zhong[4]

机构：[1]Shaoxing Univ, Sch Math Phys & Informat, 900 ChengNan Rd, Shaoxing 312000, Zhejiang, Peoples R China;[2]Shaoxing Univ, Affiliated Hosp, Shaoxing 312000, Zhejiang, Peoples R China;[3]Hangzhou Med Coll, Canc Ctr, Gamma Knife Treatment Ctr, Zhejiang Prov Peoples Hosp,Affiliated Peoples Hosp, Hangzhou 310014, Zhejiang, Peoples R China;[4]Huzhou Univ, Sch Informat Engn, Huzhou 313000, Zhejiang, Peoples R China

年份：2025

卷号：15

期号：1

起止页码：27388

外文期刊名：SCIENTIFIC REPORTS

收录：SCI-EXPANDED(收录号：WOS:001539314400020)、、Scopus(收录号：2-s2.0-105012784865)、WOS

基金：Data used in preparation of this article were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database (adni.loni.usc.edu). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report. A complete listing of ADNI investigators can be found at: http://adni.loni.usc.edu/wp-content/uploads/how_to_apply/ADNI_Acknowledgement_List.pdf.

语种：英文

外文关键词：Parameter-efficient fine-tuning; Deep learning; Low-rank adaptation; Hippocampus segmentation; Brain metastasis segmentation

外文摘要：Deep learning techniques have become pivotal in medical image segmentation, but their success often relies on large, manually annotated datasets, which are expensive and labor-intensive to obtain. Additionally, different segmentation tasks frequently require retraining models from scratch, resulting in substantial computational costs. To address these limitations, we propose PDoRA, an innovative parameter-efficient fine-tuning method that leverages knowledge transfer from a pre-trained SwinUNETR model for a wide range of brain image segmentation tasks. PDoRA minimizes the reliance on extensive data annotation and computational resources by decomposing model weights into principal and residual weights. The principal weights are further divided into magnitude and direction, enabling independent fine-tuning to enhance the model's ability to capture task-specific features. The residual weights remain fixed and are later fused with the updated principal weights, ensuring model stability while enhancing performance. We evaluated PDoRA on three diverse medical image datasets for brain structure and metastasis segmentation. The results demonstrate that PDoRA consistently outperforms existing parameter-efficient fine-tuning methods, achieving superior segmentation accuracy and efficiency. Our code is available at https://github.com/Perfect199001/PDoRA/tree/main.