Marathon data.zip

Data accompanying manuscript Data of 7 runners during a Marathon is provided to accompany the manuscript “Quantifying and correcting for speed and stride frequency effects on running mechanics in fatiguing outdoor running”. For each runner the following files are provided: - Global Navigation Satellite System (GNSS) running speed - Inertial Measurement unit (IMU) running speed - Acceleration of foot sensor - Acceleration of lower leg/tibia sensor - Knee angles GNSS running speed GNSS running speed during the full marathon was based on different sports watches. Sampling frequencies between sports watches differed but was on average 0.7 (0.4) Hz. In line with the manuscript, missing latitude-longitude data was linearly interpolated before speed was computed as the distance between two latitude-longitude coordinates based on the Haversine formula. GNSS speeds above 20 km/h were deemed extremely unlikely and replaced with spline interpolation. GNSS speed was then resampled to 240 Hz to match the sampling frequency of the IMUs. Note that GNSS and IMU data are not yet time synchronized! - Filename: SubXX_gnss_speed.csv - Size of matrix: [1xN] IMU running speed IMU running speed was solely used for time synchronization of the GNSS running speed with IMU data. The scaled biomechanical model (as described in the manuscript) provided the velocity of the pelvis segment at 240 Hz. Pelvis IMU speed was then computed as the resultant pelvis IMU velocity. - Filename: SubXX_imu_speed.csv - Size of matrix: [Nx1] Acceleration of foot sensor Accelerations of the right foot were used for initial contact detection in the manuscript. 3D accelerations of a sensor on the right foot are provided in a sensor-fixed coordinate system. The sensor was placed on the midfoot within the shoes, the sensor was aligned with the long axis of the foot. The positive axis of the first dimension points towards the center of the ankle joint. The positive axis of the second dimension points to the right. The positive axis of the third dimensions is directed approximately upwards. - Filename: SubXX_rfoot_acc.csv - Size of matrix: [Nx3] Acceleration of lower leg/tibia sensor Accelerations of the lower leg were one of the quantities of interest in the manuscript. 1D acceleration of a sensor on the right lower leg at 240 Hz is provided in a sensor-fixed coordinate system. The sensor was aligned with the axial direction of the tibia. - Filename: SubXX_rtibia_acc.csv - Size of matrix: [Nx1] Knee angles Knee flexion/extension angles were one of the quantities of interest in the manuscript. Knee flexion/extension angles of the right lower leg at 240 Hz are provided. Knee flexion angles were defined 0° when the leg was fully extended during neutral standing. Flexion resulted in positive knee flexion angles. - Filename: SubXX_rknee_angle.csv - Size of matrix: [Nx1]

本数据集配套于学术论文《量化并校正疲劳户外跑步中速度与步频对跑步生物力学的影响》（"Quantifying and correcting for speed and stride frequency effects on running mechanics in fatiguing outdoor running"），包含7名马拉松跑者的跑步相关数据。每名跑者均附带以下文件：
1. 全球导航卫星系统（Global Navigation Satellite System, GNSS）跑步速度：全程马拉松的GNSS跑步速度数据源自多款运动腕表，不同腕表的采样频率存在差异，平均采样频率为0.7（标准差0.4）Hz。按照配套论文的处理流程，在基于哈弗辛公式（Haversine formula）计算两点经纬度坐标间的距离以推导速度前，先对缺失的经纬度数据进行线性插值。对于高于20 km/h的GNSS速度数据，因其不符合正常跑步速度范围，采用样条插值进行修正替换。随后将GNSS速度重采样至240 Hz，以匹配惯性测量单元（Inertial Measurement Unit, IMU）的采样频率。注意：GNSS与IMU数据尚未完成时间同步！
文件名：SubXX_gnss_speed.csv，数据矩阵维度：[1xN]
2. 惯性测量单元（Inertial Measurement Unit, IMU）跑步速度：本数据仅用于实现GNSS跑步速度与IMU数据的时间同步。配套论文中提及的缩放生物力学模型可输出240 Hz的骨盆节段速度，进而计算得到骨盆IMU合速度。
文件名：SubXX_imu_speed.csv，数据矩阵维度：[Nx1]
3. 足部传感器加速度数据：本研究使用右足传感器的加速度数据实现初始触地时刻检测。数据包含安装于右足的传感器在自身固定坐标系下的三维加速度分量。传感器被安置于鞋内的足弓位置，并与足部长轴对齐：第一维度的正方向指向踝关节中心，第二维度的正方向指向身体右侧，第三维度的正方向大致朝上。
文件名：SubXX_rfoot_acc.csv，数据矩阵维度：[Nx3]
4. 小腿/胫骨传感器加速度数据：小腿加速度是本论文关注的核心观测指标之一。本数据集提供了安装于右小腿的传感器在240 Hz采样频率下、自身固定坐标系下的一维加速度数据，传感器与胫骨轴向对齐。
文件名：SubXX_rtibia_acc.csv，数据矩阵维度：[Nx1]
5. 膝关节角度数据：膝关节屈伸角度是本论文关注的核心观测指标之一。本数据集提供了右小腿在240 Hz采样频率下的膝关节屈伸角度数据。定义中立站立状态下腿部完全伸直时的膝关节角度为0°，屈膝动作将产生正的膝关节屈曲角度。
文件名：SubXX_rknee_angle.csv，数据矩阵维度：[Nx1]