脳腫瘍手術におけるコネクトミクス：大規模臨床的実現可能性と仮説生成トラクトメトリー所見

Maximal tumor resection with neurological preservation is central to brain tumor surgery. This study evaluates the integration of an artificial intelligence-based connectomics platform for surgical planning, with exploratory tractometry analysis of postoperative white matter changes.

We retrospectively reviewed 192 consecutive brain tumor surgeries performed between April 2023 and April 2025 using preoperative connectomic mapping (Quicktome, Omniscient Neurotechnology). Functional outcomes were assessed with Karnofsky Performance Status (KPS). In an opportunistic subgroup (n = 13), paired pre- and postoperative imaging enabled tractometry (Cleartome). Fractional anisotropy (FA) was measured across six major tracts, and three hemispheric FA asymmetry indices were calculated: difference in average FA (Diff FA), asymmetry index (AI), and percentage asymmetry (%Asym).

The cohort included intra-axial and extra-axial tumors. Median KPS remained stable at 3 months after surgery. In the tractometry subgroup, 12 of 13 patients showed postoperative shifts toward improved interhemispheric FA symmetry in at least one tract. Left-hemisphere cases most often showed changes in the uncinate fasciculus (UF) and inferior fronto-occipital fasciculi (IFOF) (median AI = - 0.68 and - 0.20, respectively), while right-hemisphere cases demonstrated alterations in the superior longitudinal fasciculus (SLF) (median AI = + 5.13; %Asym =-8.43). The corticospinal tract (CST) and IFOF improved in 61.5% of analyzed cases. Tract-based asymmetry indices captured subtle connectivity changes not evident in raw FA values alone.

Routine clinical use of AI-guided connectomics in brain tumor surgery was feasible, with stable short-term functional outcomes. Tract-based hemispheric FA asymmetry metrics suggested postoperative microstructural alterations; findings remain preliminary and warrant further prospective validation.