CPC G16H 50/30 (2018.01) | 52 Claims |
1. A method for quantitatively estimating a likelihood of a stroke condition of a subject, the method comprising:
acquiring non-invasive clinical measurement data pertaining to said subject, said clinical measurement data including at least one of image data, sound data, movement data, and tactile data;
constructing, via machine learning in an initial training phase, a positive stroke model from at least part of a positive stroke dataset acquired from a plurality of subjects positively diagnosed with at least one stroke condition and, in a steady-state operation phase continuously updating by training through machine learning said positive stroke model via its defining parameters through parameter estimation and optimization;
extracting from said clinical measurement data, potential stroke features according to at least one predetermined stroke assessment criterion;
comparing said potential stroke features with classified sampled data of said positive stroke dataset; and
determining, according to said comparing and said positive stroke model, without neuroimaging of said subject, a probability of a type of said stroke condition, and a probability of a corresponding stroke location of said stroke condition with respect to a particular brain location of said subject.
|