Implementation and optimization of an ultra-low power vagus nerve sensing system for epileptic seizures detection

Epilepsy is a neurological disorder that affects more than 50 million of people around the world. Among those people, two-thirds cannot be cured with medications. The need for finding a suitable non-invasive solution for those people, medications resistant, is a medical and vital stake to help them lead a more bearable daily-life by reducing the frequency and the intensity of the epilepsy crisis. To this aim, it is necessary to detect the epileptic seizures. The vagus nerve is of interest to this as it contains bio-markers associated with epileptic seizures. This work proposes a design and implementation of a system combining analog front-end and micro-controller in order to isolate the vagus nerve electroneurogram and detect those bio-markers. The design is made according to two figures of merit, the added noise and the total power consumption, in order to make the system suitable for a long-term medical implantation. This work succeeded in detecting those bio-markers and getting a power consumption of 1.623 mW, the same range of power consumption as already existing implantable medical devices like cochlear or muscle stimulator. The results obtained allow to consider the developed system as a suitable solution to epilepsy detection. Experiments on real subjects still need to be conducted as the efficiency of this design was tested only on signals generated in laboratory.