Parkinson's Disease Hoehn&Yahr Dataset

The ALAMEDA_PD_HoehnYahr_dataset.csv contains 76 preprocessed features extracted from raw force data. In total, it includes 79 columns: The first two columns (ID, Datetime) correspond to the ID of PD patient and Date of the recording. The next 76 columns correspond to features extracted from raw force data collected with loadsol insoles throughout a short walk assessment during in-clinic visits. The features were extracted from both time and frequency domains of the recorded timeseries. The naming of the features contains information on the channel used to extract each feature. More specifically l1 corresponds to left heel, r1 to right heel, l8 to left forefoot and r8 to right forefoot. Features extracted from the frequency domain are denoted by the inclusion of the term freq in their name. The full list of the extracted features used to predict Hoehn & Yahr stage is demonstrated in the table below. The final column, HoehnYahr, indicates the Hoehn & Yahr stage of the patients.ALAMEDA_PD_HoehnYahr_dataset.csv数据集包含从原始力数据中提取的76项预处理特征，共计79列。前两列分别为ID与Datetime，对应帕金森病（Parkinson's Disease, PD）患者的唯一标识与记录采集日期。后续76列均为从Loadsol鞋垫在门诊就诊期间的短步行评估中采集的原始力数据所提取的特征，这些特征从所记录时序数据的时域与频域中提取得到。各特征的命名中包含了其提取所用传感通道的相关信息：具体而言，l1代表左足跟，r1代表右足跟，l8代表左前足，r8代表右前足。从频域提取的特征，其名称中会包含"freq"字段以作标识。用于预测霍恩-亚尔（Hoehn & Yahr）分期的全部提取特征列表详见下表。数据集的最后一列为HoehnYahr，用于标注对应患者的霍恩-亚尔分期。

搜集汇总

背景与挑战

背景概述

该数据集包含76个从帕金森病患者短距离行走评估中收集的原始力数据中提取的预处理特征，用于预测患者的Hoehn & Yahr分期。特征涵盖时间和频域，并区分左右脚的脚跟和前脚掌区域，总共79列数据，包括患者ID、记录日期和分期标签。数据集由ALAMEDA项目发布，适用于帕金森病相关研究。