Our goal is to strengthen selective extero-interoceptive attention (SEIA), motor planning (MP) and working memory (WM) by targeting networks that regulate visuospatial perception (VP) as a function of motion direction and coherence discrimination through our individualized fMRI neuromodulation (iNM; U.S. Patent No.16/954,256). Our goal is to decrease deficits in subjective cognitive impairment (SCI), the prodromal phase to Mild-Cognitive-Impairment. SCI lacks objective diagnostic criteria and affects 11.2% >45years; 50.6% of this cohort reports functional limitations. Low vision impairment (LVI) can serve as a diagnostic predictor, as 18% reported LVI and SCI, compared with only 4% with SCI but no LVI.

fMRI measures the magnitude and spatial extent of oxy-(O₂)-to-deoxy-genated hemoglobin [Hb]. iNM: 1. is non-invasive with 1mm anatomical and functional precision; and 2. is guided by reinforcement of the HbO2 intensity of each patient’s brain network. We assessed the feasibility of iNM to strengthen the magnitude of the signal in VP, SEIA, WM, and MP networks. 

Eight participants underwent iNM and control-NO iNM to discriminate up and down direction, at full and subthreshold coherences. iNM targeted each participant’s individualized VP network. We conducted: 1. an encoding model via a GLM, which determined the HbO2 magnitude area under the curve (AUC) for each network’s area; and 2. a decoding model via SVM, which predicted the stimulus from the brain maps. 

The increase in the AUCs’ HbO2 magnitude under iNM across directions and coherences ranged from: 1. 48-76% in the SEIA; 2. 26-59% in the MP; 3. 20-47% in the WM; and 4. 100% in the VP for strong coherences. SVM resulted in statistically significant greater classification accuracies under iNM compared to control (p<0.001).

iNM enhances the HbO2 magnitude of networks. Encoding and decoding modeling allows to validate the results and allude to a causal inference of the mechanisms induced via iNM.