On-Board Platform
Within the METABO project, an integrated vehicle/driver subsystem will be developed for the detection and prediction of hypoglycaemic events (or other metabolic diseases related to diabetes mellitus) while driving exploiting indirect measures. In fact the driving activity can be impaired by several factors, among which hypoglycaemia is included. Such pathological state, produced by a lower than normal level of blood glucose, prevents the central nervous system to maintain its normal level of activity. One of the most important issues related to hypoglycaemia is that diabetics tend to lose awareness of the early autonomic symptoms, as their diseases advances. This phenomenon, known as “hypoglycaemia unawareness”, may be very dangerous in context (such as driving or performing risky activity as an air traffic controller) where the loss of concentration may entail serious consequences. In the specific case of driving, detecting in advance these signs and providing suitable audiovisual alerting to the driver can prevent fatal traffic accidents.
This vehicle/driver subsystem, called On-Board METABO Platform (OBMP) will realize the detection on the basis of a combined analysis of four concurrent strategies, related to:
- The BEHAVIORAL state
- The PHYSIOLOGICAL state
- The EMOTIONAL state
- The HEALTH state
These strategies will provide independent detection and diagnosis extracting the information from different type of sources which composes the man-car-environment system.
The behavioural module
This module will investigate the hypoglycaemic symptoms from the vehicle side in signals such as the steering wheel angle, the vehicle speed and accelerations and the driver’s usage of the gas and brake pedals. The assumption is that hypoglycaemic symptoms arising on the human side thanks to the human-vehicle-mechanical coupling are detectable also on the vehicle side.
The physiological module
This module will use biomedical sensors connected to the on-board ultra-mobile PC. Focusing in hypoglycaemia detection the most important bio-signal will be the electrocardiogram (ECG), the waveforms of which are known to change in hypoglycaemia. More specifically, the repolarisation time of the heart muscle is known to increase and the amplitude of the T wave is known to decrease in hypoglycaemia. Also other bio-signals such as blood pressure, Galvanic Skin Response (GSR), electromyogram (EMG), respiratory rate and skin temperature have expected correlations with hypoglycaemia and these correlations will be investigated along the prosecution of the project.
The emotional module
This module will investigate the possibility to predict glucose level changes based on actual emotional state under a short-term observation condition. This module is based on multimodal emotion recognition combining together information extracted from physiological signals and visual signal.
Physiological signals will be extracted from biomedical sensors in the physiological module.
