A historical perspective of biomedical explainable AI research.
A historical perspective of biomedical explainable AI research.
复制标题
DOI：
10.1016/j.patter.2023.100830
复制
发表时间：
2023-09-08
期刊：
影响因子：
6.5
通讯作者：
Rosen-Zvi, Michal
中科院分区：
文献类型：
--
作者：
Malinverno, Luca;Barros, Vesna;Ghisoni, Francesco;Visona, Giovanni;Kern, Roman;Nickel, Philip J.;Ventura, Barbara Elvira;Simic, Ilija;Stryeck, Sarah;Manni, Francesca;Ferri, Cesar;Jean-Quartier, Claire;Genga, Laura;Schweikert, Gabriele;Lovri, Mario;Rosen-Zvi, Michal
大多数人工智能（AI）模型的黑箱性质鼓励开发可解释性方法，以在AI决策过程中产生信任。这种方法可以大致分为两种主要类型：事后解释和固有的可解释算法。我们旨在分析COVID-19与可解释人工智能（XAI）推动生物医学研究前沿之间的可能关联。我们从PubMed数据库中自动提取了与因果关系或可解释性概念相关的生物医学XAI研究，并手动标记了1，603篇关于XAI类别的论文。为了比较COVID-19前后的趋势，我们拟合了一个变点检测模型，并评估了出版率的显著变化。我们发现，2020年初COVID-19的出现可能是XAI受到更多关注的驱动因素，在加速已经发展的趋势方面发挥着至关重要的作用。最后，我们讨论了XAI技术的未来社会使用和影响，以及那些寻求通过可解释的机器学习模型培养临床信任的人的潜在未来方向。理解机器学习模型的内部工作已经成为人工智能（AI）公平性和可靠性讨论的关键点。从这个角度来看，我们揭示了最近发表的关于生物医学科学中可解释AI（XAI）的科学著作的见解。具体而言，我们推测COVID-19大流行与该领域的出版率有关。目前的研究工作似乎更多地指向解释黑盒机器学习模型，而不是设计新颖的可解释架构。值得注意的是，2020年10月出现了发表率的拐点，当时XAI在生物医学科学方面的研究数量大幅上升。虽然普遍接受的可解释性定义不太可能，但正在进行的研究工作正在推动生物医学领域提高应用机器学习的鲁棒性和可靠性，我们认为这是一个积极的趋势。可解释人工智能（XAI）领域在过去几年中迅速加速，这有助于讨论可解释性的需求和定义。该观点首次系统分析了生物医学领域内XAI研究的趋势，并将COVID-19大流行视为可能的转折点。我们探讨了最近的研究动机，并强调了可解释模型与事后解释之间的对比，以及关于性能-可解释性权衡的争论。
The black-box nature of most artificial intelligence (AI) models encourages the development of explainability methods to engender trust into the AI decision-making process. Such methods can be broadly categorized into two main types: post hoc explanations and inherently interpretable algorithms. We aimed at analyzing the possible associations between COVID-19 and the push of explainable AI (XAI) to the forefront of biomedical research. We automatically extracted from the PubMed database biomedical XAI studies related to concepts of causality or explainability and manually labeled 1,603 papers with respect to XAI categories. To compare the trends pre- and post-COVID-19, we fit a change point detection model and evaluated significant changes in publication rates. We show that the advent of COVID-19 in the beginning of 2020 could be the driving factor behind an increased focus concerning XAI, playing a crucial role in accelerating an already evolving trend. Finally, we present a discussion with future societal use and impact of XAI technologies and potential future directions for those who pursue fostering clinical trust with interpretable machine learning models. Understanding the inner working of machine-learning models has become a crucial point of discussion in fairness and reliability of artificial intelligence (AI). In this perspective, we reveal insights from recently published scientific works on explainable AI (XAI) within the biomedical sciences. Specifically, we speculate that the COVID-19 pandemic is associated with the rate of publications in the field. Current research efforts seem to be directed more toward explaining black-box machine-learning models than designing novel interpretable architecture. Notably, an inflection period in the publication rate was observed in October 2020, when the quantity of XAI research in biomedical sciences surged upward significantly. While a universally accepted definition of explainability is unlikely, ongoing research efforts are pushing the biomedical field toward improving the robustness and reliability of applied machine learning, which we consider a positive trend. The field of explainable artificial intelligence (XAI) saw a rapid acceleration in the last few years that is contributing to discussions regarding requirements and definitions of interpretability. This perspective is the first to systematically analyze the trends in XAI research within the biomedical field considering the COVID-19 pandemic as a possible turning point. We explore the motivation of recent studies and highlight the contrast between interpretable models versus post hoc explanation and the debates regarding performance-interpretability tradeoff.
登录
查看更多内容
影响因子：
4.4
作者：
Carter H;Douville C;Stenson PD;Cooper DN;Karchin R
通讯作者：
Karchin R
影响因子：
16.6
作者：
通讯作者：
--
影响因子：
3.9
作者：
Adadi, Amina;Berrada, Mohammed
通讯作者：
Berrada, Mohammed
影响因子：
11
作者：
Kumari, Monika;Lu, Ruei-Min;Li, Mu-Chun;Huang, Jhih-Liang;Hsu, Fu-Fei;Ko, Shih-Han;Ke, Feng-Yi;Su, Shih-Chieh;Liang, Kang-Hao;Yuan, Joyce Pei-Yi;Chiang, Hsiao-Ling;Sun, Cheng-Pu;Lee, I-Jung;Li, Wen-Shan;Hsieh, Hsing-Pang;Tao, Mi-Hua;Wu, Han-Chung
通讯作者：
Wu, Han-Chung
影响因子：
0.9
作者：
Gunning, David;Aha, David W.
通讯作者：
Aha, David W.