Publish time: 2022-05-09 00:00:00
26A-1110211
TW I 855635
Personalized modeling has long been anticipated to approach precise noninvasive blood glucose (NIBG) measurements, but challenged by limited data for training personal model and its unavoidable outlier predictions. We largely enhanced the training efficiency with the limited personal data by an innovative Deduction Learning (DL). We also developed an automatic screening algorithm to identify low confidence outlier predictions. Among 30 diabetic patients, with only a dozen rounds of training data, DL with automatic screening achieved a correlation coefficient of 0.81, an accuracy score of 93.5, a root mean squared error (RMSE) of 13.93 mg/dl, a mean absolute error (MAE) of 12.07 mg/dl, and 100% predictions in zone A of CEG.
- High accuracy of Non-Invasive Blood Glucose Prediction
- Require less data for model training
- Novel deep learning algorithm model structure
- Efficient use of data
- Non-Invasive Blood Glucose Prediction
- Personalized medicine
- Precision medicine
- Wearable products

Predictions of rounds 13~15 by models built with a dozen rounds of training data (rounds 1~12) for (a) IL, (b) DL, and (c) DL+S
Fu-Liang Yang, Wei-Ru Lu, Wen-Tse Yang, Justin Chu, Tung-Han Hsieh
If you are interested in this technology, please fill in the information on the webpage under the manufacturer option (Request for Technology Licensing, Material Transfer, or Sponsored Research) on the website of this office

Precise Non-Invasive Blood Glucose Prediction by Deduction Learning System