Microwave System for Head Imaging

PCA-based artifact removal algorithm for stroke detection using UWB radar imaging

“…Tables 2 and 3 show the same comparisons in terms of the metrics previously introduced. First of all, comparisons show that both for the monostatic and for the multistatic modes and with a moderate number of antennas (8 rather than 16-32 as considered in [13]), the proposed preprocessing algorithm shows better performance. In the monostatic mode, the entropy-based algorithm and the symmetry-based algorithm have metrics comparable with the ones achieved with the PCA-based algorithm.…”

“…So far, some feasibility studies and proof of concepts on the use of MWI for stroke detection have been reported [6,8,12,13,18,23]. A relevant work has been done at Chalmers University [6,12,18], where clinical trials are ongoing.…”

“…Several works have been done on the antennas design [13,28]. In this paper, we focus on the second challenge.…”

“…In this paper, we focus on the second challenge. Most of the works on UWB radar for stroke detection consider a monostatic system and the delay-and-sum (DAS) algorithm [8,13], which was originally proposed for breast cancer detection [11]. The DAS algorithm has the attracting feature of having low computational complexity but it offers limited robustness to artifacts.…”

“…To achieve good performance using the DAS algorithm, a very high number of antennas must be considered. For instance, in [13], 16 and 32 antennas are used. To improve the performance and reduce the number of antennas, much more complex beamforming algorithms have been proposed such as the MIST, microwave imaging space-time [1,16], also originally proposed for breast cancer detection.…”