Microwave Antenna-Assisted Machine Learning: A Paradigm Shift in Non-Invasive Brain Hemorrhage Detection
Adarsh Singh,Bappaditya Mandal,Bishakha Biswas,Sankhadeep Chatterjee,Soumen Banerjee,Debasis Mitra,Robin Augustine
DOI: https://doi.org/10.1109/access.2024.3371886
IF: 3.9
2024-03-15
IEEE Access
Abstract:Brain hemorrhages have become increasingly common and can be fatal if left untreated. Current methods for monitoring the progression of the disorder that rely on MRI and PET scans are inconvenient and costly for patients. This has spurred research toward portable and cost-effective techniques for predicting the current stage and malignancy of the hemorrhages. In this study, simulated S-parameter data obtained from a two-antenna system placed over the head is used in conjunction with machine learning to detect the dielectric changes in the brain caused by hemorrhage non-invasively. Several machine learning classifiers are used to analyze the data, and their performance metrics are compared to determine the optimal classifier for this case. The study revealed that Decision Tree, KNN, and Random Forest classifiers are better than SVM and MLP classifiers in terms of accuracy, precision, and recall in predicting Brain hemorrhage at the most probable locations. Contrary to conventional microwave imaging systems requiring several antennas for brain hemorrhage detection, this study demonstrates that integrating machine learning with microwave sensors enables accurate solutions with a reduced antenna count. The results present a transformative strategy for monitoring systems in clinics, where a simple, safe, and low-cost microwave antenna-based system can be intelligently integrated with machine learning to diagnose the presence of Brain hemorrhage.
computer science, information systems,telecommunications,engineering, electrical & electronic