Electromagnetic Portable Brain Imaging For Stroke

Background: Electromagnetic imaging (EMI) is an emerging technology that transmits low energy electromagnetic waves from a ring of transceivers around the head, modified as they pass through abnormal tissue, providing unique signatures for brain pathology. It promises to provide portable, non-ionizing, rapid neuroimaging for prehospital and bedside evaluation of stroke, based on the dielectric properties of the tissue. We aimed to assess the clinical utility of EMI in stroke diagnosis in a pilot study. Methods: In a prospective, observational, open, non-interventional pilot study, patients with imaging-proven ischemic (IS) or haemorrhagic stroke (ICH) within the preceding 48 hours were recruited. Using the EMVision scanner, EMI was performed within 1-24 hours of diagnostic CT or MRI. Images were obtained by processing signals from encircling transceiver antennae contained in an instrumented 18 kg helmet which emit and detect low energy non-ionising signals in the microwave frequency spectrum (0.5-2.0 GHz). Localisation was assessed by determining whether fusion images resulted in target detection in the same quadrant as comparable CT or MRI. Electromagnetic (EM) images were reconstructed by creating maps of the EM wave scattering arising from contrast in electrical parameters between IS or ICH lesions and normal brain. A blinded clinician assessed agreement between regional abnormalities on EMI and CT or MRI scans. Algorithms for distinction between IS and ICH were based on differences in EM transmission, reflection and scattering through brain tissue. Results: Thirty patients were studied, 21 IS and 9 ICH. Mean age was 66.7 years (range 37-87), 57% were female. Mean NIHSS at presentation was 5. Mean time to routine imaging was 5.5 hrs (range 1-48) and to EMI 24 hrs (range 6-60). Nineteen patients (63%) had only CT performed; 11 (37%) had both CT and MRI. EMI differentiated ICH from IS with 93% accuracy and localised the stroke to the correct brain quadrant with 87% accuracy. Conclusion: In this early validation pilot study we show the ability to distinguish between IS and ICH and stroke location within a given brain quadrant. Further developments may produce a valuable imaging tool to assist in prehospital and bedside stroke diagnosis and management.