Explainable ML for sleep-related respiratory disorders
We want to investigate the use of state-of-the art interpretable ML and explainable AI solutions for detection and observation of sleep-related respiratory disorders through one or several Theses. Major questions to be investigated are:
- Can existing explainable ML solutions be used for sleep monitoring data, like polysomnography data, data from non-invasive ventilation devices, and data from journals from patients treated for laryngomalacia?
- Which kind of explanations can be given and are they useful for ML developers, medical staff, or patients?
There will be in the first phase some literature analysis necessary to select some candidates for explainable ML solutions that should be investigated empirically on sleep monitoring data. For the empirical part the explainable ML solutions have to be implemented (if not available as open source), models need to be trained and evaluated.
Knowledge in ML is a requirement for this work which will be embedded with other MSc. Theses in the Respire project: