Cross-shaped windows transformer with self-supervised pretraining for clinically significant prostate cancer detection in bi-parametric MRI

IF 3.2 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Medical physics Pub Date : 2024-11-26 DOI:10.1002/mp.17546

Yuheng Li, Jacob Wynne, Jing Wang, Richard L. J. Qiu, Justin Roper, Shaoyan Pan, Ashesh B. Jani, Tian Liu, Pretesh R. Patel, Hui Mao, Xiaofeng Yang

{"title":"Cross-shaped windows transformer with self-supervised pretraining for clinically significant prostate cancer detection in bi-parametric MRI","authors":"Yuheng Li, Jacob Wynne, Jing Wang, Richard L. J. Qiu, Justin Roper, Shaoyan Pan, Ashesh B. Jani, Tian Liu, Pretesh R. Patel, Hui Mao, Xiaofeng Yang","doi":"10.1002/mp.17546","DOIUrl":null,"url":null,"abstract":"<div>\n \n \n <section>\n \n <h3> Background</h3>\n \n <p>Bi-parametric magnetic resonance imaging (bpMRI) has demonstrated promising results in prostate cancer (PCa) detection. Vision transformers have achieved competitive performance compared to convolutional neural network (CNN) in deep learning, but they need abundant annotated data for training. Self-supervised learning can effectively leverage unlabeled data to extract useful semantic representations without annotation and its associated costs.</p>\n </section>\n \n <section>\n \n <h3> Purpose</h3>\n \n <p>This study proposes a novel self-supervised learning framework and a transformer model to enhance PCa detection using prostate bpMRI.</p>\n </section>\n \n <section>\n \n <h3> Methods and materials</h3>\n \n <p>We introduce a novel end-to-end Cross-Shaped windows (CSwin) transformer UNet model, CSwin UNet, to detect clinically significant prostate cancer (csPCa) in prostate bpMRI. We also propose a multitask self-supervised learning framework to leverage unlabeled data and improve network generalizability. Using a large prostate bpMRI dataset (PI-CAI) with 1476 patients, we first pretrain CSwin transformer using multitask self-supervised learning to improve data-efficiency and network generalizability. We then finetune using lesion annotations to perform csPCa detection. We also test the network generalization using a separate bpMRI dataset with 158 patients (Prostate158).</p>\n </section>\n \n <section>\n \n <h3> Results</h3>\n \n <p>Five-fold cross validation shows that self-supervised CSwin UNet achieves 0.888 ± 0.010 aread under receiver operating characterstics curve (AUC) and 0.545 ± 0.060 Average Precision (AP) on PI-CAI dataset, significantly outperforming four comparable models (nnFormer, Swin UNETR, DynUNet, Attention UNet, UNet). On model generalizability, self-supervised CSwin UNet achieves 0.79 AUC and 0.45 AP, still outperforming all other comparable methods and demonstrating good generalization to external data.</p>\n </section>\n \n <section>\n \n <h3> Conclusions</h3>\n \n <p>This study proposes CSwin UNet, a new transformer-based model for end-to-end detection of csPCa, enhanced by self-supervised pretraining to enhance network generalizability. We employ an automatic weighted loss (AWL) to unify pretext tasks, improving representation learning. Evaluated on two multi-institutional public datasets, our method surpasses existing methods in detection metrics and demonstrates good generalization to external data.</p>\n </section>\n </div>","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 2","pages":"993-1004"},"PeriodicalIF":3.2000,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Medical physics","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/mp.17546","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING","Score":null,"Total":0}

引用次数: 0

Abstract

Background

Bi-parametric magnetic resonance imaging (bpMRI) has demonstrated promising results in prostate cancer (PCa) detection. Vision transformers have achieved competitive performance compared to convolutional neural network (CNN) in deep learning, but they need abundant annotated data for training. Self-supervised learning can effectively leverage unlabeled data to extract useful semantic representations without annotation and its associated costs.

Purpose

This study proposes a novel self-supervised learning framework and a transformer model to enhance PCa detection using prostate bpMRI.

Methods and materials

We introduce a novel end-to-end Cross-Shaped windows (CSwin) transformer UNet model, CSwin UNet, to detect clinically significant prostate cancer (csPCa) in prostate bpMRI. We also propose a multitask self-supervised learning framework to leverage unlabeled data and improve network generalizability. Using a large prostate bpMRI dataset (PI-CAI) with 1476 patients, we first pretrain CSwin transformer using multitask self-supervised learning to improve data-efficiency and network generalizability. We then finetune using lesion annotations to perform csPCa detection. We also test the network generalization using a separate bpMRI dataset with 158 patients (Prostate158).

Results

Five-fold cross validation shows that self-supervised CSwin UNet achieves 0.888 ± 0.010 aread under receiver operating characterstics curve (AUC) and 0.545 ± 0.060 Average Precision (AP) on PI-CAI dataset, significantly outperforming four comparable models (nnFormer, Swin UNETR, DynUNet, Attention UNet, UNet). On model generalizability, self-supervised CSwin UNet achieves 0.79 AUC and 0.45 AP, still outperforming all other comparable methods and demonstrating good generalization to external data.

Conclusions

This study proposes CSwin UNet, a new transformer-based model for end-to-end detection of csPCa, enhanced by self-supervised pretraining to enhance network generalizability. We employ an automatic weighted loss (AWL) to unify pretext tasks, improving representation learning. Evaluated on two multi-institutional public datasets, our method surpasses existing methods in detection metrics and demonstrates good generalization to external data.

查看原文本刊更多论文

在双参数磁共振成像中利用十字形窗口变换器和自监督预训练技术检测具有临床意义的前列腺癌。

背景：双参数磁共振成像（bpMRI）在前列腺癌（PCa）检测方面取得了可喜的成果。与卷积神经网络（CNN）相比，视觉变换器在深度学习中取得了具有竞争力的性能，但它们需要大量标注数据进行训练。自监督学习可以有效利用未标注数据提取有用的语义表征，而无需标注及其相关成本。目的：本研究提出了一种新型自监督学习框架和转换器模型，以利用前列腺 bpMRI 增强 PCa 检测：我们引入了一种新颖的端到端异形窗（CSwin）转换器 UNet 模型 CSwin UNet，用于检测前列腺 bpMRI 中具有临床意义的前列腺癌（csPCa）。我们还提出了一种多任务自我监督学习框架，以利用未标记数据并提高网络的泛化能力。通过使用包含 1476 名患者的大型前列腺 bpMRI 数据集（PI-CAI），我们首先使用多任务自我监督学习对 CSwin 变换器进行预训练，以提高数据效率和网络泛化能力。然后，我们使用病变注释进行微调，以进行 csPCa 检测。我们还使用一个包含 158 名患者的独立 bpMRI 数据集（Prostate158）对网络泛化进行了测试：五倍交叉验证结果表明，自监督 CSwin UNet 在 PI-CAI 数据集上实现了 0.888 ± 0.010 的接收器操作特性曲线下误差（AUC）和 0.545 ± 0.060 的平均精度（AP），明显优于四个同类模型（nnFormer、Swin UNETR、DynUNet、Attention UNet、UNet）。在模型泛化能力方面，自监督 CSwin UNet 的 AUC 为 0.79，AP 为 0.45，仍然优于所有其他同类方法，并对外部数据表现出良好的泛化能力：本研究提出了 CSwin UNet，这是一种基于变压器的新型 csPCa 端到端检测模型，通过自我监督预训练增强了网络泛化能力。我们采用自动加权损失（AWL）来统一前置任务，从而改进表征学习。在两个多机构公共数据集上进行评估后，我们的方法在检测指标上超越了现有方法，并对外部数据表现出良好的泛化能力。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Medical physics 医学-核医学

CiteScore

6.80

自引率

15.80%

发文量

660

审稿时长

1.7 months

期刊介绍： Medical Physics publishes original, high impact physics, imaging science, and engineering research that advances patient diagnosis and therapy through contributions in 1) Basic science developments with high potential for clinical translation 2) Clinical applications of cutting edge engineering and physics innovations 3) Broadly applicable and innovative clinical physics developments Medical Physics is a journal of global scope and reach. By publishing in Medical Physics your research will reach an international, multidisciplinary audience including practicing medical physicists as well as physics- and engineering based translational scientists. We work closely with authors of promising articles to improve their quality.