Brain Injury Localization in Electromagnetic Imaging using Symmetric Crossing Lines Method

Abstract

To avoid death or disability, patients with brain injury should undertake a diagnosis at the earliest time and accept frequent monitoring after starting any medical intervention. This paper presents a novel approach to localize brain injury using the intersection of pairs of signals from symmetrical antennas based on the hypothesis that healthy brains are approximately symmetric that the bleeding targets will lead to significantly different amplitude and phase changes if a pair of symmetrical transmit signals cross targets with the same distances. The scattered signals (S-parameters) across the band 0.5-2 GHz are acquired using 100 realistic brain models and 150 experimental data measurements. Each signal is divided into two frequency bands and then they are converted into two graphs. The neighbour pair signals from symmetric hemisphere are evaluated to detect and localize the potential abnormalities in the brain. The higher weighted pairs across the axes of symmetry are detected and used to localize any abnormal targets. The results indicate that crossing pairs of antenna signals from the hemisphere with a blood mass exhibit significantly different signal amplitude in the graph features compared to those without the target (p<0.003). The experiments show that our novel localization algorithm can achieve an accuracy of 0.85±0.08 Dice similarity coefficient based on 150 experimental measurements using an elliptical container, which is suitable for brain injury localization.

Stroke Localization Using Multiple Ridge Regression Predictors Based on Electromagnetic Signals

“…This section applies the intersection of circles to obtain the stroke targets. Line intersection methods have been applied in [12] to localize the stroke targets. The intersection of at least two circular trajectories can obtain a potential existing location of the target.…”

“…Wu et al [11] convert it to classification by using the interconnection point of two lines that pass the stroke. Similarly, Zhu et al [12] use geometric methods to localize ICH stroke which causes an asymmetric response to electromagnetic signals, based on the assumption that the healthy brain is approximately symmetric. In contrast, the proposed method can also be used to localize IS stroke while achieving higher prediction accuracy.…”

“…The application of AI methods to the classification of stroke subtypes has been explored [7,9,10]. Over the years, AI methods have also been increasingly applied to stroke location prediction and have shown great potential [9][10][11][12]. Many studies focus on the localization of hemorrhagic stroke and made progress [9,11,12].…”

“…Over the years, AI methods have also been increasingly applied to stroke location prediction and have shown great potential [9][10][11][12]. Many studies focus on the localization of hemorrhagic stroke and made progress [9,11,12]. For example, a framework for stroke detection on a deep-learned stochastic model was proposed recently, achieving mean and median IoU (intersection over union) of 0.84 and 0.94, respectively [9].…”