Machine Learning for Smart Health Systems

January 22, 2024

For whom is this course?

This course is for anyone interested in applying machine learning for disease detection, monitoring, and other biomedical applications.

Learn how to process physiological signals e.g. ECG, EEG, PPG, respiration captured from the human body
Feature extraction from physiological signals
Feature selection for the detection of specific diseases of interest
Importance of confounding factor analysis in biomedical applications
Disease classification and severity estimation using machine learning and deep learning models

Technical skills are essential, but not enough, non-technical and domain fields of studies are still essential if you want to understand data science vs its application.
Current and future global challenges in the sector
How data science or artificial intelligence would be applied.
Data science and the necessities to keep learning for life.
Instructor-led online course with guided labs
Real-world, practical assignment(s) leading to project
Application in biomedical and health systems

Week	Instruction (1 hr)	Lab (guided + unguided) 1+ 3 hrs
Week 1 (5 hrs)	Intro to smart health, and physiological signals. ETL and visualization in smart health	Basic signal processing-loading, denoising ECG, acceleration, PPG signal and detecting biomarkers
Week 2 (5 hrs)	Feature extraction, Feature selection in machine learning, and co-variate analysis	Extract features from ECG and PPG, acceleration signals, Feature selection for sleep apnea & heart disease classification, and gait analysis
Week 3 (5 hrs)	Developing machine learning & deep learning models for disease classification	Sleep apnea and heart disease classification using ensemble and deep neural networks
Week 4 (5 hrs)	Deploying machine learning & deep learning models for disease classification	Export portable and deployable model using WEKA & TF-Lite, cloud computing
Week 5 (10 hrs)	Case study guidance & evaluation	Real-world case study (10 hrs)

Juber Rahman, Ph.D.