识别用于脑胶质瘤分割的临床适用机器学习算法：最新进展与发现
Identifying clinically applicable machine learning algorithms for glioma segmentation: recent advances and discoveries.
作者信息
Tillmanns Niklas, Lum Avery E, Cassinelli Gabriel, Merkaj Sara, Verma Tej, Zeevi Tal, Staib Lawrence, Subramanian Harry, Bahar Ryan C, Brim Waverly, Lost Jan, Jekel Leon, Brackett Alexandria, Payabvash Sam, Ikuta Ichiro, Lin MingDe, Bousabarah Khaled, Johnson Michele H, Cui Jin, Malhotra Ajay, Omuro Antonio, Turowski Bernd, Aboian Mariam S
机构信息
Brain Tumor Research Group, Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, Connecticut, USA.
University Dusseldorf, Medical Faculty, Department of Diagnostic and Interventional Radiology, Dusseldorf, Germany.
出版信息
Neurooncol Adv. 2022 Jun 14;4(1):vdac093. doi: 10.1093/noajnl/vdac093. eCollection 2022 Jan-Dec.
BACKGROUND
While there are innumerable machine learning (ML) research algorithms used for segmentation of gliomas, there is yet to be a US FDA cleared product. The aim of this study is to explore the systemic limitations of research algorithms that have prevented translation from concept to product by a review of the current research literature.
METHODS
We performed a systematic literature review on 4 databases. Of 11 727 articles, 58 articles met the inclusion criteria and were used for data extraction and screening using TRIPOD.
RESULTS
We found that while many articles were published on ML-based glioma segmentation and report high accuracy results, there were substantial limitations in the methods and results portions of the papers that result in difficulty reproducing the methods and translation into clinical practice.
CONCLUSIONS
In addition, we identified that more than a third of the articles used the same publicly available BRaTS and TCIA datasets and are responsible for the majority of patient data on which ML algorithms were trained, which leads to limited generalizability and potential for overfitting and bias.
背景
虽然有无数用于胶质瘤分割的机器学习（ML）研究算法，但尚未有美国食品药品监督管理局（FDA）批准的产品。本研究的目的是通过回顾当前研究文献，探讨阻碍从概念转化为产品的研究算法的系统性局限。
方法
我们在4个数据库上进行了系统的文献综述。在11727篇文章中，58篇符合纳入标准，并使用TRIPOD进行数据提取和筛选。
结果
我们发现，虽然有许多关于基于ML的胶质瘤分割的文章发表，并报告了高精度的结果，但论文的方法和结果部分存在重大局限，导致难以重现这些方法并转化为临床实践。
结论
此外，我们发现超过三分之一的文章使用了相同的公开可用的BRaTS和TCIA数据集，并且这些数据集是ML算法训练所依据的大部分患者数据的来源，这导致了有限的通用性以及过拟合和偏差的可能性。
Zotero 插件