Classification applied to brain haemorrhage detection: Initial phantom studies using electrical impedance measurements

Machine learning and classification algorithms are applied to data collected from an EIT system. The system is used with an anatomically accurate head phantom setup in a variety of situations modelling normal and haemorrhagic brain. This initial study demonstrates the promise of classification, but also indicates challenges.The research leading to these results has received funding from the European Research Council under the European Union’s Horizon 2020 Programme/ ERC Grant Agreement BioElecPro n.637780, and from the Hardiman Research Scholarship, NUIG