Defining The Pattern of the Thai Osteoarthritis Movement Data Model

This study proposed an algorithm to identify the patterns of Thai osteoarthritis, a condition that has seen an increase in prevalence in recent years, significantly affecting patients' quality of life and mobility. We aimed to define the data patterns for analysis. For data collection, we employed a Razor-IMU sensor and a WIFI transmitter, allowing testers to independently perform full-range knee motions. Data collection was categorized into two groups: healthy and unhealthy. Our goal was to categorize the data and establish criteria for pattern classification. Once the patterns and criteria were established, our objective was to validate movement models for both normal individuals and osteoarthritis patients. To achieve this, we conducted statistical hypothesis testing to verify the accuracy of the data. This testing comprised three main steps: first, evaluating the accuracy of data collection and data cleaning. Second, assessing the precision of converting data from linear to angular format, including degree coordinates selection. The last, Evaluating the accuracy of data sorting and grouping using Louvain clustering. The researcher thoroughly scrutinized each step to confirm the results. Each step demonstrated an accuracy test As Thailand transitions into an aging society, the prevalence of osteoarthritis is increasing due to the natural deterioration of the body. This deterioration can be decelerated by avoiding risky behaviors. Osteoarthritis significantly impacts patients' physical and mental well-being, making it a critical health concern. The objective of this research is to develop movement prototypes for both normal individuals and osteoarthritis patients by leveraging computer knowledge. This includes data collection with motion sensor devices, enhancing data quality using data mining techniques such as data cleaning and data transformation into suitable formats, and grouping of data. Additionally, the study seeks to validate the accuracy of the algorithm using statistical hypothesis testing methods and graph pattern detection. Based on the experimental results, an accuracy rate of 97% was achieved, demonstrating a high level of reliability. This prototype can be applied in treatment analysis, monitoring treatment outcomes, and even injury prevention. Furthermore, the dataset can serve as a model for the Thai population and can be expanded to accommodate larger datasets.