Real‐time brain stroke detection through a learning‐by‐examples technique—An experimental assessment

Abstract

The real‐time detection of brain strokes is addressed within the Learning‐by‐Examples (LBE) framework. Starting from scattering measurements at microwave regime, a support vector machine (SVM) is exploited to build a robust decision function able to infer in real‐time whether a stroke is present or not in the patient head. The proposed approach is validated in a laboratory‐controlled environment by considering experimental measurements for both training and testing SVM phases. The obtained results prove that a very high detection accuracy can be yielded even though using a limited amount of training data.

Instantaneous brain stroke classification and localization from real scattering data

“…For each t ‐th illumination, the complex scattering coefficients S it , i = 1, …, T , are measured, S it being the i ‐th transmission ( i ≠ t ) or reflection ( i = t ) coefficient, respectively. Due to reciprocity, the set of independent collected data is composed by the following ( T + 1) × T real‐valued measured features where the symbol ℜ (.…”

“…In such a framework, a LBE ‐based MI inversion scheme has been proposed in Ref. for the robust and quick detection of a brain stroke in a monitored head phantom. Still following that line‐of‐reasoning, but addressing the more challenging problem of identifying the stroke typology (ie, ischemic or hemorrhagic) along with its position within the brain, this letter proposes a 2‐step LBE ‐based classification approach.…”

Robust real‐time inversion of electrical impedance tomography data for human lung ventilation monitoring

“…The inversion of EIT data for recovering the conductivity distribution within the human chest (directly related to the air content of the lungs) is a nonlinear and ill‐posed problem needing proper regularization schemes and suitable inversion strategies . Recently, learning‐by‐examples (LBE) approaches have emerged as promising alternatives to standard inverse scattering techniques since they proved to yield accurate estimations of the DUT with quasi real‐time performance starting from the knowledge of an input‐output ( I/O ) database (ie, the training set) . In such a framework, this work presents an innovative LBE method for monitoring the human lungs.…”

Wrapped ultra‐wide band radar sensor based on resistively loaded bow‐tie antenna for near zone sensing applications

“…In this regard, the impulse radar characterized by UWB pulse has widely utilized for near zone sensing applications. The applications in the vicinity of the targets can be included for not only medical applications, such as imaging of the breast tumor, imaging of the bleeding area for stroke, and assessment of bone mineral density, but also industrial applications such as ground‐penetrating radar (GPR), through‐wall penetrating radar, and foliage‐penetrating radar (FPR) …”

“…In this regard, the impulse radar characterized by UWB pulse has widely utilized for near zone sensing applications. The applications in the vicinity of the targets can be included for not only medical applications, such as imaging of the breast tumor, [2][3][4][5] imaging of the bleeding area for stroke, [6][7][8] and assessment of bone mineral density, [9][10][11] but also industrial applications such as ground-penetrating radar (GPR), [12][13][14][15] through-wall penetrating radar, 16,17 and foliage-penetrating radar (FPR). 18,19 For near zone sensing applications, the proposed antenna should radiate a short pulse with less distorted waveform to receive a clear echo pulse reflected by the target.…”