Kai-Chun Liu

[t22302856]

Date: July 1, 2022

Frozen Shoulder Assessment (Clinical co-worker: 三總 復健科 張智雅醫師)

• Progress
  • Labeling tools
• Plan
  • Physical feature extraction tool (github)

Vestibular Disorder Assessment (Clinical co-worker: 北榮 丁冠中醫師)

• Progress
  • Discrimination between Healthy and patients with vestibular disorders (Co-worker: 陽明 林于傑) XGboost -> 73% accuracy
  • Data ready (78 patients, 48 healthy)
• Plan
  • try CNN amd MLP

Miscellaneous

• Support 貫蓁's EMG denoising (prepare for paper submission IEEE OJEMB)
• Support mmWave ECG實驗

---

Date: May 20, 2022

Frozen Shoulder Assessment (Clinical co-worker: 三總 復健科 張智雅醫師)

• Progress
  • Draft finished, title: Discrimination between Healthy and Frozen Shoulder from Daily Shoulder Task- A Machine Learning Approach, CNN->88.26% accuracy, SVM->83.91% accuracy
• Plan
  • update manuscript

Vestibular Disorder Assessment (Clinical co-worker: 北榮 丁冠中醫師)

• Progress
  • Discrimination between Healthy and patients with vestibular disorders (Co-worker: 陽明 林于傑) XGboost -> 73% accuracy
• Plan
  • try CNN amd MLP

Miscellaneous

• Support 在民‘s manuscript
• Support mmWave ECG實驗
• Support 貫蓁's EMG denoising

---

Date: Apr. 22, 2022

Frozen Shoulder Assessment (Clinical co-worker: 三總 復健科 張智雅醫師)

• Progress
  • Draft finished, title: Discrimination between Healthy and Frozen Shoulder from Daily Shoulder Task- A Machine Learning Approach
  • Focus on disability assessment (Co-worker: 陽明 盧廷揚) and pain assessment (Co-worker: 中研院 陳摯心)
• Issue
  • Analysis results are hard to explain
  • The number of subjects is small
• Plan
  • Keep discussion with Dr. 張智雅
  • Include more subjects

Vestibular Disorder Assessment (Clinical co-worker: 北榮 丁冠中醫師)

• Progress
  • Discrimination between Healthy and patients with vestibular disorders (Co-worker: 陽明 林于傑)
  • Automatic step detection for the clinical test (Co-worker: 陽明 黃宜如)
  • Analyze the correlation between DHI scores and IMU features (Co-worker: 陽明 宋育庭)
• Issue
  • Hard to recruit 30-50 years old healthy female subjects
• Plan

Fall Detection

• Progress
  • Draft preparation, title: Wearable-based Fall Detection Using Ensemble Learning of Coarse-fine CNN and GRU Networks (Co-worker: 陽明 劉建彬)

  • Pre-Impact Fall Detection (Co-worker: 陽明 紀廷翰)		Accuracy	Precision	Recall	F1-score
    XGB	90.05	33.06	87.27	47.96
    CNN	93.82	54.73	72.88	62.52
    CNN-LSTM	96.98	64.98	81.25	72.21
    self-attention CNN	98.12	81.10	97.47	88.54
    self-attention CNN-GRU	98.24	82.47	96.81	89.06

• Plan
  • Paper prepation, 2022 IEEE BIOCASE

Miscellaneous

• Support 在民‘s manuscript
• Support mmWave ECG實驗

---

example Date. e.g. Aug. 25th

Progress

• ProgressItem 1.
  • ProgressItem 1.1.

  • Table		Col. 1	Col. 2
    Row 1	w1_1	w1_2
    Row 2	w2_1	w2_2

  • ProgressItem 1.2.
• ProgressItem 2.

Current issues

• Describe what kind of helps you want (e.g., a survey of some topics, computational resources, code reviews, etc.)

Action items

• [ ] ActionItem1.
• ActionItem2.
  • [ ] ActionItem 2.1
  • [ ] ActionItem 2.2

[t22302856]

Date: May 6, 2022

Frozen Shoulder Assessment (Clinical co-worker: 三總 復健科 張智雅醫師)

• Progress
  • test the results using NN, title: Discrimination between Healthy and Frozen Shoulder from Daily Shoulder Task- A Machine Learning Approach
  • Focus on disability assessment (Co-worker: 陽明 盧廷揚) and pain assessment (Co-worker: 中研院 陳摯心)

Vestibular Disorder Assessment (Clinical co-worker: 北榮 丁冠中醫師)

• Progress
  • the number of 30-50 years old healthy female subjects is sufficient to support Discrimination between Healthy and patients with vestibular disorders (Co-worker: 陽明 林于傑)
• Plan
  • Analysis results between healthy subjects and pa

Miscellaneous

• Support 在民‘s manuscript
• Support mmWave ECG實驗
• Support 貫蓁‘'s work (prepare for IEEE BHI)
• Support Prof. Tzu-ming Liu's work