解释基于加速时间序列的与年龄相关的步态分类的深度学习模型.
Xiaoping Zheng1, Egbert Otten1, Michiel F Reneman2
1University of Groningen, University Medical Center Groningen, Department of Human Movement Sciences, 9713 AV, Groningen, the Netherlands.
Computers in biology and medicine
|November 13, 2024
概括
可解释的人工智能增强了对老年人的步态分析的深度学习. SHAP强调了跟鞋接触数据是区分与年龄相关的步态模式的关键,提高了临床透明度.
科学领域:
- 生物医学工程 生物医学工程
- 数据科学数据科学数据科学
- 老年学是一门学科.
背景情况:
- 步态分析对于监测老年人的健康至关重要.
- 传感器技术的进步为步态分析产生了大数据.
- 深度学习 (DL) 是有前途的,但缺乏对临床使用的透明度.
研究的目的:
- 提高与年龄相关模式的基于DL的步态分类的透明度.
- 使用可解释的人工智能 (SHAP) 来解释DL模型.
- 提高AI在步态分析中的临床适用性.
主要方法:
- 与244名参与者 (成人和老年人) 进行的横截面研究.
- 在3分钟的步行中,在L3上使用了加速度计.
- 训练有素的卷积神经网络 (CNN) 在 1 层数据上,以及在 8 层数据上的门式循环单元 (GRU).
- 应用SHAP用于模型解释.
主要成果:
- 美国有线电视新闻 (CNN) 获得了81.4%的准确率 (AUC 0.89),大陆空军 (GRU) 获得了84.5%的准确率 (AUC 0.94).
- SHAP认为跟接触周围的垂直和行走方向数据是最重要的.
- 格鲁的分析考虑了间的变化,与CNN不同.
结论:
- 美国有线电视新闻网根据单步数据对步态进行分类;GRU使用跨步关系.
- 鞋跟接触数据对于区分成年人和老年人的步行模式至关重要.
- 可解释的人工智能 (SHAP) 为步态分析提供了DL模型的洞察力.
更多相关视频
相关概念视频
01:29Aggregates Classification
Aggregate classification is generally based on its size, petrographic characteristics, weight, and source. Size classification ranges from coarse to fine aggregates, defined by the size of the particles. Coarse aggregates are particles that do not pass through ASTM sieve No. 4, and aggregates that pass through the sieve are fine aggregates.
Petrographic classification groups aggregates based on common mineralogical characteristics. Some of the common mineral groups found in aggregates are...
Petrographic classification groups aggregates based on common mineralogical characteristics. Some of the common mineral groups found in aggregates are...
01:26Exponential Equations for Modeling Growth
Exponential models are essential for describing rapid, multiplicative changes in natural systems, such as population growth. When a population doubles at regular intervals, the process can be modeled using a suitable base. For instance, a bacterial culture that doubles every three hours follows the model n(t)=n0⋅2t/3, where n(t) is the population at the time t.A more general model uses the natural base e, especially for continuous growth. This takes the form n(t)=n0⋅ert, where r is the relative...