In New Zealand alone, one in twenty people currently live with heart disease, a condition which includes disruptions to heart rhythm, known as arrhythmia. Despite being a prolific and life-threatening issue, it is often difficult for at-risk individuals to be monitored in a clinical context constantly. This is due to costs, healthcare system constraints and the fact that many at-risk people continue to function normally in their day to-day lives.
With advancements in real-time heart monitoring and machine learning techniques, our research aims to design a relatively low-cost system for providing automatic emergency action during a heart anomaly.
A commercially available smart wearable will be used to capture heart condition data from the patient and send it for processing on a smartphone app. Machine learning will be used to initially train an algorithm to identify various heart states, and thus be able to help classify heart condition data as it is received on the smartphone. If issue is detected, the app will provide warnings and trigger an emergency callout to the appropriate medical services.
- Design and implementation of neural network for classification of heart arrhythmia states
- Prototype (Raspberry Pi/Arduino) emitting heart condition data, with real-time classification via neural network
- Smartphone application interfacing with wearable heart tracker to produce emergency callouts
- Computer Systems Engineering
- Software Engineering
Allocated (Not available for preferences)
Lab allocations have not been finalised