直肠癌基于磁共振成像的深度学习研究进展
Progress in deep learning based on magnetic resonance imaging for rectal cancer research
石晟铭 1肖玲清 2马佳琪 1刘晗 1武玉鹏 1李晓夫1
作者信息
- 1. 哈尔滨医科大学附属第二医院磁共振成像诊断科,哈尔滨 150086
- 2. 新疆生产建设兵团第十师北屯医院总医院影像中心,北屯 836099
- 折叠
摘要
直肠癌(rectal cancer,RC)是一种在全球范围内发病率较高的恶性肿瘤,对其的管理和治疗均面临着显著的挑战.MRI是评估RC的常规手段,然而常规MRI及功能MRI由于各自固有的局限性常常无法为RC患者提供足够的信息,以支持个性化的治疗计划.随着近年来人工智能在医疗界的迅速进展,深度学习技术在评估RC分期、治疗反应、RC分割和基因分型等方面已经展现出极大的潜力和广阔的应用前景.本文就基于MRI的深度学习技术在RC中的应用作一综述,以期为RC患者选择最佳治疗方案提供参考,从而改善患者的预后,并希望能对未来的科研工作提供新思路与新方向.
Abstract
Rectal cancer(RC)is a malignancy with a high incidence rate worldwide,posing significant challenges in its management and treatment.MRI is the conventional modality used to assess RC.However,both traditional MRI and functional MRI frequently fall short in providing sufficient information for the development of personalized treatment plans for RC patients due to their inherent limitations.With the rapid advancements in artificial intelligence within the medical field in recent years,deep learning technologies have demonstrated tremendous potential and broad prospects for applications in areas such as RC staging,treatment response evaluation,RC segmentation,and genetic typing.These advancements suggest that deep learning could pave new ways for enhancing the precision and personalization of treatment decisions in RC in the future.This article presents a comprehensive review on the application of MRI-based deep learning techniques in RC,aiming to assist in selecting the optimal treatment strategy for RC patients,thereby improving patient outcomes,and providing new insights and directions for future research endeavors.
关键词
直肠癌/磁共振成像/深度学习/预测性能Key words
rectal cancer/magnetic resonance imaging/deep learning/predicted performance引用本文复制引用
基金项目
新疆生产建设兵团财政科技计划(2021AB029)
出版年
2024
国家科技期刊平台
NSTL
万方数据