Healthcare simulation
Presentations
An emergency-room patient simulation for practising symptom recognition.
Presentations is a Unity simulation of patient presentations in an emergency-room setting. I built it to model diseases with software, and to practise recognizing disease patterns from a patient’s demographics, symptoms and vital signs.
Generating a patient
The game covers 18 disease presentations. It generates demographic and risk-factor information such as age, weight, activity, smoking and family history, uses those factors to weight disease likelihood, and then generates a symptom profile. Players choose a diagnosis and receive feedback. Hover text explains physiological ranges and terminology, giving the case more context as the player works through it.
Hand-written rules made it difficult to represent the interplay between many risk factors and symptoms. Working on those relationships became part of my motivation to explore machine learning.