Dataset for: Effects of Handle Diameter and Center of Gravity on Tremor Suppression

Description

1. Research Hypothesis. This study hypothesizes that modifying the physical morphology of a spoon—specifically, increasing the handle diameter and shifting the center of gravity (CG) proximally—will biomechanically suppress hand tremors during feeding tasks in older adults. Furthermore, we hypothesize that these objective biomechanical benefits may conflict with users' subjective preferences due to their long-term cognitive reliance on traditional slender utensils. 2. Data Gathering Methodology. The data was collected from 20 older adults (aged 75 and above) who performed a simulated 8-second static feeding holding task. A 2 × 2 repeated-measures design was employed, with handle diameter (12 mm vs. 30 mm) and CG position (proximal vs. distal) as independent variables. (a) Objective Kinematic Data: Captured using a wireless 6-axis Inertial Measurement Unit (IMU) mounted on the spoon handle, recording tri-axial acceleration and angular velocity at 200 Hz. (b) Subjective Data: Collected via structured interviews immediately following the tasks to assess operational preference, perceived effort, and willingness to use, using a 3-point forced-choice scale. 3. Notable Findings & What the Data Shows. (a) Objective Biomechanics: The kinematic data indicates that the 30 mm thick handle significantly improved linear stability (reduced acceleration CV, p < 0.05) and demonstrated greater rotational robustness during CG shifts compared to the 12 mm thin handle. (b) Subjective Preference-Performance Paradox: Despite the objective physical advantages of the thick handle, the subjective data reveals that 65% of participants significantly preferred the thin handle (p < 0.01), driven by established mental models (e.g., chopstick usage habits). For CG, 80% preferred the proximal weighting (p < 0.001), aligning with mechanical intuition. 4. How to Interpret and Use the Data. This dataset is structured to guarantee full reproducibility. It includes raw IMU time-series signals, processed/filtered data segments, and the final statistical feature dataset (containing Root Mean Square and Coefficient of Variation values, along with their Log10 transformations). A Jupyter Notebook (.ipynb) containing all Python processing scripts is also provided.Other researchers can use this dataset to: (a) Replicate the repeated-measures ANOVA to verify the kinematic and statistical findings. (b) Explore alternative signal processing and filtering algorithms for tremor quantification. (c) Investigate the correlation and discrepancies between objective biomechanical metrics and subjective ergonomic preferences in age-friendly design. For a detailed explanation of folder structures, condition codes (e.g., A1, B2), and variable definitions, please refer to the included README.txt file.

Steps to reproduce

1. Experimental Setup & Material Preparation - Hardware Modification: Prepare 4 adaptive spoon prototypes based on a standard stainless-steel spoon. Use ultra-light clay to manipulate the handle diameter (12 mm for Thin, 30 mm for Thick). Use neodymium magnetic weights (30 g total) to manipulate the Center of Gravity (Proximal vs. Distal). - Instrumentation: Attach a WitMotion WT9011DCL-BT50 wireless 6-axis Inertial Measurement Unit (IMU) to the proximal end of the grip area. Configure the sampling rate and Bluetooth transmission to 200 Hz. 2. Data Acquisition (Task Execution) - Recruit older adults (aged 75+) and seat them in a comfortable, natural feeding posture. - Instruct participants to grasp the spoon with their dominant hand and perform an 8-second static holding task at a designated mid-air target. - Use a 4 × 4 Latin Square Design to randomize the presentation order of the four experimental conditions for each participant to mitigate learning effects and fatigue. - Immediately following the tasks, conduct structured interviews using a 3-point scale to collect subjective evaluation metrics (operational preference, perceived effort, and willingness to use). 3. Kinematic Signal Processing (Python Workflow) Execute the provided Jupyter Notebook (`.ipynb`) to reproduce the signal processing steps: - Preprocessing: Import the raw time-series CSV files. Detrend the tri-axial acceleration and angular velocity data to remove DC components (e.g., gravity). - Filtering: Apply a 4th-order Butterworth high-pass filter (cutoff frequency = 2 Hz, zero-phase shift) independently to each axis to eliminate low-frequency drift induced by postural adjustments. - Resultant Calculation: Calculate the 3D resultant magnitude of the purified tri-axial signals frame-by-frame. - Secondary Truncation: Apply a sliding-window algorithm (fixed window of 5 seconds / 1000 samples). The algorithm traverses the sequence and extracts the most stable contiguous 5-second segment by identifying the window with the lowest standard deviation. 4. Feature Extraction & Statistical Analysis - From the extracted 5-second core data, compute the Root Mean Square (RMS) to quantify absolute tremor intensity, and the Coefficient of Variation (CV) to assess relative fluctuation instability for both acceleration and angular velocity. - Apply a Log_10 transformation to the raw RMS and CV values to satisfy normality assumptions (Shapiro-Wilk test). - Finally, use the transformed dataset (`03_Final_Statistical_Data.csv`) to conduct a 2 × 2 repeated-measures ANOVA (Handle × CG) and perform nonparametric tests (Chi-square, Wilcoxon signed-rank) on the subjective data.

Institutions

• Nanjing University of Science and Technology

  Jiangsu, Nanjing

Categories

Funders

• High School Philosophy and Social Science Foundation of Department of Education of Jiangsu Province of China

  Grant ID: No. 2021SJZDA016
• National Nature Science Foundation of China

  Grant ID: No. 72401136

Licence