INTER-SUBJECT VARIABILITY EVALUATION TOWARDS A ROBUST MICROWAVE SENSOR FOR PNEUMOTHORAX DIAGNOSIS

Pneumothorax is the medical condition caused by the air concentration inside the pleural cavity, the space between the lung and the chest wall. Apart from traditional diagnostic methods, it can be detected by using microwave sensors that capture variations in reflected electromagnetic field (EMF). Sex and obesity, related to the internal composition of the biological tissues, can influence the reflected EMF and therefore the sensor diagnostic ability. This paper investigates the effect on the performance of a proposed on-body dual-patch antenna sensor for pneumothorax diagnosis, due to inter-subject variability in underlying tissue structure. The sensor operates at frequency range of 1-4 GHz. The challenge of the paper is to propose frequency bands for robust and safe sensor operation. S12 parameter alternation versus frequency is assessed for healthy and pathological cases. Implemented thorax numerical models include modified (i) closed rectangular multilayered and (ii) MRI-based anatomical ones. In rectangular models, thickness and configuration of muscle, fat and bone tissues are varied, according to literature. Additionally, sex-related anatomical differences are taken into account in MRI-based models. All scenarios are solved using Finite Difference Time Domain method. Results revealed that the proposed frequency bands lie within 1-2.7 and 2.9-3.5 GHz, for muscle, 1.4-3.5 GHz for fat and 1-2.2 and 2.8-3.5 GHz, for bone variations. Numerical evaluations for accurate anatomical models verify the findings.