促进医疗保健中患者监测和评估系统的活动识别算法的公平性
Ciro Mennella1, Massimo Esposito1, Giuseppe De Pietro2
1Institute for High-Performance Computing and Networking (ICAR) Research National Council of Italy (CNR), Italy.
Computers in biology and medicine
|July 9, 2024
概括
定义受试者特征对于强大的人类活动识别算法至关重要. 训练数据异质性,特别是残疾属性,显著提高了对不同人群的AI模型准确性.
科学领域:
- 人工智能的人工智能
- 生物医学工程 生物医学工程
- 计算机科学 计算机科学
背景情况:
- 定义人类活动识别 (HAR) 算法的主题选择标准具有挑战性.
- 确保跨多种人群的算法稳定性需要理解数据异质性.
- 物理和软生物识别属性影响HAR性能,需要进行调查.
研究的目的:
- 分析HAR算法的训练数据异质性.
- 评估物理和软生物识别特征对HAR性能的影响.
- 识别影响运动分析算法的特征.
主要方法:
- 在IntelliRehab (IRDS) 数据集上评估了最先进的深度神经网络 (tCNN,混合LSTM,变压器).
- 在基于人类特征的培训数据中引入了故意的偏见.
- 多通道时间序列活动识别数据中的量化表示偏差.
主要成果:
- 在CNN-LSTM模型实现了最高准确度88%.
- 在异质残疾属性上训练的模型显示出更高的准确性 (51%) 与没有 (33%) 相比.
- 主体特征显著影响活动识别性能.
结论:
- 对象特征对于强大的HAR算法性能至关重要.
- 不同质的数据分布,特别是残疾属性,提高了AI模型的准确性.
- 这项研究通过量化偏见来提高医疗保健AI的公平性和可信度.
相关概念视频
Pulse rhythm
782
Pulse rhythm refers to the pattern of pulsations within specific intervals, offering valuable insights into the regularity or irregularity of the heart's beats as observed through the pattern of pulsation within specific intervals. A regular pulse exhibits a consistent heart rate with uniform waveforms and pulsation force, variations of which can be classified as normal, weak, or bounding.
Conversely, an irregular pulse pattern is termed dysrhythmia, stemming from disruptions in cardiac...
Conversely, an irregular pulse pattern is termed dysrhythmia, stemming from disruptions in cardiac...
782
Errors occurring during blood pressure monitoring
681
Blood pressure monitoring is a crucial clinical procedure in diagnosing and managing various cardiovascular conditions. Despite its significance, the accuracy of blood pressure measurements can be compromised by multiple factors, potentially leading to either falsely high or low readings. These inaccuracies are critical as they can significantly impact patient care. So, it is vital to understand these challenges deeply and adopt strategic approaches to minimize errors.
Several factors...
Several factors...
681


