Page 45 - GTM-2-3
P. 45

Global Translational Medicine                                           Deep learning by NMR-biochemical




                           A













































                           B

























            Figure 12. (A) Segmentation of multiple sclerosis lesions is shown as trained data set with yellow color coding and magnetic resonance spectroscopy
            peaks. (B) The semi‑supervised KNFST algorithm procedure is shown to select data samples (on the left panel) and accept the unlabeled data match with
                                                                                   [19]
            confidence band conditions until all processed data are predicted (on the right panel). Modified from references .

            Volume 2 Issue 3 (2023)                         13                        https://doi.org/10.36922/gtm.337
   40   41   42   43   44   45   46   47   48   49   50