Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

From Empirical to Simulated Brains: Using Digital Twin Models to Identify Therapeutic Targets in MCI

Poster Session C - Sunday, March 8, 2026, 5:00 – 7:00 pm PDT, Fairview/Kitsilano Ballrooms

Jinyu Wang¹ (jinyu.wang.csu@gmail.com), Kim Cuong Tran Dang¹, Psyche Loui¹; ¹Northeastern University

Brain stimulation techniques offer promising non-pharmacological interventions for improving cognitive function in mild cognitive impairment (MCI), but it is unclear which target regions of stimulation would be most effective. While in vivo brain stimulation studies yield the most direct data, obtaining causal insights is often challenging due to practical and ethical concerns that arise from clinical experiments. Here, we built the digital twin brain(DTB) model, which integrates multimodal empirical neuroimaging data from MCI participants to simulate the dynamics and interactions inside their brains, and identify regions that are most influential for recovering healthy brains from MCI brains through virtual perturbation. A total of 30 MCI participants and 26 healthy participants were included. We used the Hopf model to model each brain region, where each region’s bifurcation parameter controls whether it’s noisy or oscillatory. The brain structural connectivity was used to provide information for simulating global dynamics. Probabilistic metastable substates (PMS) were used to estimate dynamic functional patterns. Model optimization was achieved by minimizing the difference between the simulated and empirical PMS. By perturbing each region independently, we found that increasing excitability in bilateral superior temporal gyri, bilateral rostral anterior cingulate gyri, and right inferior frontal gyrus causally shift the MCI DTB’s whole-brain metastable dynamics toward the healthy states. These results offer insight into the neural underpinnings of MCI in a more causal way and may inform future empirical brain stimulation therapies for MCI.

Topic Area: METHODS: Neuroimaging