Stroke Localization Using Multiple Ridge Regression Predictors Based on Electromagnetic Signals

Localizing stroke may be critical for elucidating underlying pathophysiology. This study proposes a ridge regression–meanshift (RRMS) framework using electromagnetic signals obtained from 16 antennas placed around the anthropomorphic head phantom. A total of 608 intracranial haemorrhage (ICH) and ischemic (IS) are collected evaluated RRMS, where each type signal contains two different diameters phantoms. Subsequently, multiple regression predictors then give target distances mean shift is used to cluster predicted location based on these distances. The test results show that training time economic cost significantly reduced as average prediction only takes 0.61 s achieve an accurate result (average position error = 0.74 cm) conventional laptop. It has great potential auxiliary standard medical method, or rapid diagnosis patients in underdeveloped areas, due its rapidity, good deployability, low hardware cost.