รายละเอียดวารสารวิชาการ

  1. หน้าแรก
  2. รายชื่อนักวิจัย
  3. มนรดา ศิริมงคล
  4. ข้อมูลวารสารวิชาการ

ชื่องานวิจัย ( Article Name)

Abnormal Gait Pattern Recognition of Stroke Patient in Initial Stage Using Smartphone and Hybrid Classification Methods

บทคัดย่อ ( Abstract )

This paper proposes abnormal gait pattern recognition of Stroke patient in the initial stage using smartphone and hybrid classification methods. Our proposed research is divided into 2 parts: model building for abnormal gait pattern detection and system development. For a model of abnormal gait pattern detection, sensors (accelerometer and gyroscope sensor) on a smartphone are used to collect accelerometer and gyroscope data of gait pattern. Then, data of gait pattern are transformed and selected appropriated attributes to build a model using hybrid classification methods (Multilayer Perceptron, Decision Tree, and Support Vector Machine). From our experiments, the result shows that our proposed method achieved high accuracy of 99.40% for abnormal gait pattern recognition of Stroke patient in the initial stage and compared with previous research using accelerometer and gyroscope. Finally, our research developed a gait stroke detection system for abnormal gait pattern detection. When the system detects abnormal gait pattern, the system will send notification to caregiver and physician to reduce the risk of fall and accidence, especially, in the case of patient living alone at home. In addition, the system can help the physician to diagnose and train the gait pattern of post-stroke patients for rehabilitation.

เอกสารวิจัย ( Paper )

VIEW

ผู้เขียน ( Authors )

  • มนรดา ศิริมงคล

ข้อมูลการเผยแพร่ ( Journal Infomation )