Machine Learning for Bioelectromagnetics and Biomedical Engineering: Some Sample Applications

In a recent paper we have introduced the use of Machine-Learning to improve the effectiveness of electroporation treatments. On the basis of a wide literature analysis, and after building up a solid knowledge repository, we were able to build up an Artificial Neural Network (ANN) so to predict the impact of the treatment in terms of ablation area. In this paper, we demonstrate that the same approach can be extended so to allow the optimum choice and tuning of some parameters with an important impact on the quality of the treatment, such as the position of the electrodes, their size, geometry, etc. This allows the customization of the treatment to a wider variety of diseases, and its tailoring on specific cases or patients. We finally propose the extension of the ANN approach to a novel application area, extremely important in many biomedical applications: gesture recognition. We demonstrate that the approach, combined with the use of a special glove using chipless RF tags, can be effective in the detection of the movements of fingers in a human hand. For this application, we also investigate some open problems and future developments.

Automated summary

This paper introduces the use of machine learning and artificial neural networks for predicting and optimizing electroporation treatments, and extends the approach to gesture recognition.