Lesion Segmentation from Imperfect Medical Data via Weakly Supervised Learning with SAM and Explainable Vision Transformer

人工智能 计算机科学 分割 不完美的 变压器 图像分割 计算机视觉 模式识别(心理学) 监督学习 医学影像学 训练集 机器学习 合成数据 半监督学习 尺度空间分割 标记数据 特征提取 深度学习 医学诊断 病变 语音识别 数据建模 特征学习 医学知识
作者
Ahmad Nawaz Zaheer,Suirong Cao,Muhammad Farhan,Awais Ahmad,Sohail Jabbar,Shaharyar Kamal,Alaa Mohammad Menshawi
出处
期刊:IEEE Journal of Biomedical and Health Informatics [Institute of Electrical and Electronics Engineers]
卷期号:PP: 1-8
标识
DOI:10.1109/jbhi.2026.3737415
摘要

Accurate lesion segmentation remains a critical yet challenging task in medical imaging due to scarce pixel-level annotations, inter-observer variability and modality-specific noise. Conventional fully supervised models require exhaustive manual labeling, which limits scalability and clinical applicability. To address these challenges, this paper proposes a hybrid weakly supervised framework that integrates the Segment Anything Model (SAM) with an Explainable Vision Transformer (ViT) for robust and interpretable lesion segmentation. The SAM module converts weak annotations such as bounding boxes and sparse points into high-quality pseudo-masks refined through ensemble consistency, shape priors, and confidence masking, reducing dependency on dense labels. A ViT-based multiple-instance learning (MIL) module then performs weakly supervised classification and mask refinement using uncertainty-aware optimization and self-training, ensuring resilience to noisy supervision. A SAM-ViT alignment loss enforces consistency between pseudo-masks and transformer attention maps, improving anatomical plausibility. Additionally, an explainable AI (XAI) mechanism aligns attention rollout and gradient-based saliency with lesion masks, embedding interpretability directly into the training process. This unified pipeline delivers noise-robust segmentation, clinically transparent predictions, and efficient label utilization under imperfect supervision. Experimental results on CAMELYON16/17 and BraTS datasets demonstrate state-of-the-art performance, achieving AUCs of 0.972, 0.949 and 0.941, with a Dice coefficient of 0.883 and mIoU of 0.826, outperforming existing weakly supervised and SAM-only baselines. This framework establishes a scalable and interpretable foundation for clinical-grade lesion segmentation, reducing annotation cost while improving reliability in real-world medical imaging workflows.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
飞远发布了新的文献求助10
刚刚
1秒前
lyh完成签到,获得积分10
2秒前
2秒前
倾听阳光完成签到,获得积分10
3秒前
3秒前
当0卷死坏女人完成签到,获得积分10
3秒前
Akim的应助被怕黑冥王星采纳,获得10
3秒前
Makoto1377完成签到,获得积分10
3秒前
Rida302完成签到,获得积分10
4秒前
念姬发布了新的文献求助10
4秒前
4秒前
酷波er的应助被WangXingQuan采纳,获得10
5秒前
隐形曼青的应助被Gibson采纳,获得10
6秒前
溧谦发布了新的文献求助10
6秒前
6秒前
停云发布了新的文献求助10
6秒前
充电宝的应助被霸气雯采纳,获得10
6秒前
6秒前
在水一方的应助被思川采纳,获得10
7秒前
顺心觅风发布了新的文献求助10
7秒前
8秒前
Lucas的应助被粗心的若采纳,获得10
9秒前
9秒前
情怀的应助被long-zhang采纳,获得10
9秒前
9秒前
Wiki发布了新的文献求助30
9秒前
hh完成签到,获得积分10
10秒前
11秒前
小博士328完成签到,获得积分10
11秒前
12秒前
高大真完成签到,获得积分10
12秒前
852的应助被Marshall采纳,获得10
13秒前
酷波er的应助被Marshall采纳,获得10
13秒前
13秒前
桐桐的应助被霸气雯采纳,获得10
16秒前
iwhisper完成签到,获得积分10
16秒前
16秒前
Xx完成签到,获得积分20
17秒前
桐桐的应助被热心宛菡采纳,获得10
17秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
八旗制度の研究 600
Improving the wear resistance of vehicle gear oil by oil soluble LaF3 nanoparticles and typical additives of gear oil 500
Boundaries and Bodhisattvas 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8183601
求助须知:如何正确求助?哪些是违规求助? 9572875
关于积分的说明 21827816
捐赠科研通 7780716
什么是DOI,文献DOI怎么找? 3434031
关于科研通互助平台的介绍 2558197
邀请新用户注册赠送积分活动 2480485