Microwave Radar Imaging as a Tool for Medical Diagnostics

In brain stroke detection “timeliness” is mandatory. Indeed, medical actions should be provided quickly and within the 60 minutes, which is the “door-to-needle Golden Hour”. It was shown that every minute an ischemic stroke lasts and remains untreated, a patient on average loses 1.9 million neurons, 13.8 billion synapses, and 12 km of axonal fibers. For each hour delay in treatment the brain loses as many neurons as it does in nearly 3.6 years of normal aging.

In this framework, a quick diagnostic is obviously highly required. However, well assessed technologies, such as magnetic resonance imaging (MRI) and X-ray computed tomography (CT), do not meet this requirement since they can be used only in hospital. In addition, CT uses ionizing radiation and MRI is expensive. Also, continuous monitoring is hardly feasible.

For these reasons, research has been exploring different imaging modalities. Among these, microwave imaging appears a good candidate for complementing MRI and CT. Indeed, the related technology is relatively low cost and compact. Hence, portable imaging systems which are transportable in ambulance are feasible.

In this paper we introduce the microwave imaging prototype for stroke detection that we have developed and compare it with few similar systems spread in literature.

The system consists of a helmet within which a number of antennas, properly designed to work in contact to the human head, are arranged. Therefore, no matching liquid is necessary. The imaging algorithm is based on an incoherent (with respect to the frequencies) approach. As a results, dispersive behaviour of antennas and head tissues are not critical issues. Also, the required frequency band is narrower than for other similar systems.

The system is light and take few minutes to provide the result. Hence, it meets the need for quickness and transportability. The achievable performance is checked against physical phantom of the head. The obtained results show that on site quick stroke diagnostics is feasible.