Mobile Health

Aims

The course aims to explore how wearable and mobile systems sensors can be used to gather data relevant to understand health, how the data can be analysed with advanced signal processing and machine learning and the performance of these systems in terms of diagnostics and disease progression detection.

Syllabus

• Course Overview. Introduction to Mobile Health. Evaluation metrics and methodology. Basics of Signal Processing.
• Inertial Measurement Units, Human Activity Recognition (HAR) and Gait Analysis and Machine Learning for IMU data.
• Radios, Bluetooth, GPS and Cellular. Epidemiology and contact tracing, Social interaction sensing and applications. Location tracking in health monitoring and public health.
• Audio Signal Processing. Voice and Speech Analysis: concepts and data analysis. Body Sounds analysis.
• Photoplethysmogram and Light sensing for health (heart and sleep)
• Contactless and wireless behaviour and physiological monitoring
• Mobile Devices and Behaviour intervention
• Topical Guest Lectures

Objectives

The course aims to explore how wearable and mobile systems sensors can be used to gather data relevant to understand health, how the data can be analysed with advanced signal processing and machine learning and the performance of these systems in terms of diagnostics and disease progression detection.

Roughly, each lecture contains a theory part about the working of “sensor signals” or “data analysis methods” and an application part which contextualises the concepts.

At the end of the course students should: Understand how mobile/wearable sensors capture data and their working. Understand different approaches to acquiring and analysing sensor data from different types of sensors. Understand the concept of signal processing applied to time series data and their practical application in health. Be able to extract sensor data and analyse it with basic signal processing and machine learning techniques. Be aware of the different health applications of the various sensor techniques. The course will also touch on privacy and ethics implications of the approaches developed in an orthogonal fashion.

Recommended Reading

Please see Course Materials for recommended reading for each session.

Assessment - Part II students

Two assignments will be based on two datasets which will be provided to the students:

Assignment 1 (shared for Part II and Part III/MPhil): this will be based on a dataset and will be worth 40% of the final mark. The task of the assessment will be to perform pre-processing and basic data analysis in a "colab" and an answer sheet of no more than 1000 words.

Assignment 2 (Part II): This assignment (worth 60% of the final mark) will be a fuller analysis of a dataset focusing on machine learning algorithms and metrics. Discussion and interpretation of the findings will be reported in a colab and a report of no more than 1200 words.

Assessment  - Part III and MPhil students

Two assignments will be based on datasets which will be provided to the students:

Assignment 1: this will be based on a dataset and will be worth 40% of the final mark. The task of the assessment will be to perform pre-processing and basic data analysis in a "colab" and an answer sheet of no more than 1000 words.

Assignment 2: The second assignment (worth 60% of the final mark) will be based on multiple datasets. The students will be asked to compare and contract ML algorithms/solutions (trained and tested on the different data) and discuss the findings and interpretation in terms of health context. This will be in the form of a colab and a report of 1800 words.

Further Information

Current Cambridge undergraduate students who are continuing onto Part III or the MPhil in Advanced Computer Science may only take this module if they did NOT take it as a Unit of Assessment in Part II.