临床医生对重症监护中可解释AI的偏好:对可解释模型和可视化用于输管决策支持的可比研究
Tiantian Xian1, Nikolay Mehandjiev2, Panos Constantinides1
1Alliance Manchester Business School, University of Manchester, Manchester, UK.
International journal of medical informatics
|January 22, 2026
概括
以临床医生为中心的视觉解释提高了对人工智能的信任,用于输管预测. 一个双编码的SHAP热图最好与临床工作流程保持一致,增强在重症监护病房的AI采用.
科学领域:
- 人工智能在医学中的应用
- 临床决策支持系统 临床决策支持系统
- 可解释的人工智能 (XAI)
背景情况:
- 复杂的人工智能模型面临采用障碍,原因是临床医生缺乏透明度.
- 这项研究解决了临床医生为中心的可解释AI (XAI) 接口的需求.
- 专注于用于输管预测的直观视觉解释,以提高接受度.
研究的目的:
- 将基于XAI的输管预测的三个时间意识的视觉解释进行比较.
- 评估临床医生对这些解释的接受,理解和感知效用.
- 测试假设,工作流兼容的设计可以提高人工智能接受度.
主要方法:
- 开发了机器学习模型,使用ICU时间序列数据来预测输管风险.
- 使用SHAP (夏普利添加式解释) 生成全球和本地解释.
- 设计了三种视觉格式:时间力图,时间条形图和双编码的SHAP热图.
- 临床医生 (n=206) 通过调查评估理解和可用性.
主要成果:
- 一个随机森林模型实现了0.94.9的曲线下的高面积 (AUC).
- 当地解释由206名临床医生评估.
- 双编码的SHAP热图被选为具有最高感知临床实用性和工作流程对齐的可视化.
结论:
- 解释格式必须根据临床推理和认知适应的任务背景量身定制.
- 热图与临床医生的心理模型对齐,提高了解释性和信任性.
- 语境相关性和呈现格式显著影响临床医生对XAI系统的信任.
相关概念视频
Interpreting R Charts
344
R chart, or range chart, is a fundamental tool in statistical process control used to monitor the variability within a process. It complements the X-bar (x̄) chart by focusing on the range of the data, rather than individual values, providing a clear picture of the process dispersion over time.
An R chart plots the range of subsets of measurements collected from a process. Each point on the chart represents the range—defined as the difference between the maximum and minimum...
An R chart plots the range of subsets of measurements collected from a process. Each point on the chart represents the range—defined as the difference between the maximum and minimum...
344
Critical Region, Critical Values and Significance Level
13.2K
The critical region, critical value, and significance level are interdependent concepts crucial in hypothesis testing.
In hypothesis testing, a sample statistic is converted to a test statistic using z, t, or chi-square distribution. A critical region is an area under the curve in probability distributions demarcated by the critical value. When the test statistic falls in this region, it suggests that the null hypothesis must be rejected. As this region contains all those values of the...
In hypothesis testing, a sample statistic is converted to a test statistic using z, t, or chi-square distribution. A critical region is an area under the curve in probability distributions demarcated by the critical value. When the test statistic falls in this region, it suggests that the null hypothesis must be rejected. As this region contains all those values of the...
13.2K
Self-Help Support Groups
339
Self-help support groups are voluntary, community-based organizations that provide a platform for individuals with shared concerns to exchange support, insights, and practical strategies for coping with life challenges. Typically led by group members or paraprofessionals, these groups form a cornerstone of mental health care, especially in reaching populations that are underserved by traditional healthcare systems.
Accessibility and Cost-Effectiveness
One of the primary strengths of self-help...
Accessibility and Cost-Effectiveness
One of the primary strengths of self-help...
339
Critical Values
10.2K
A critical value is a definite value obtained from a particular probability distribution at a predecided confidence level (or a predecided significance level) for a given population parameter. The critical value provides demarcation that separates the sample statistics that are likely to occur from the ones that are unlikely to occur based on the given probability distribution and the population parameter to be estimated. The critical value for normal distribution is obtained from the z...
10.2K
Decision Making
926
Decision-making is a fundamental cognitive process that involves evaluating alternatives and selecting among them. This process can range from simple choices, such as deciding what to wear, to complex decisions, like choosing a major in college or a career path. The complexity of the decision often dictates the approach we use, which can be broadly categorized into two types: automatic and controlled decision-making.
Automatic decision-making is fast, intuitive, and relies on gut feelings...
Automatic decision-making is fast, intuitive, and relies on gut feelings...
926
Interpreting Run Charts
3.1K
Run charts, essentially line graphs plotted over time, serve as fundamental yet effective tools for process analysis. They chronicle data sequentially, facilitating the identification of trends, shifts, or cyclical movements. This graphical representation is instrumental in determining whether a process is stable or exhibits signs of potential instability indicative of special cause variation. In the healthcare domain, run charts depict infection rates over time, enabling hospitals to monitor...
3.1K


