Data-driven modeling of trigeminal nerve shape reveals pain-associated morphology in trigeminal neuralgia
Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Although the leading theory for the cause of trigeminal neuralgia (TN) is neurovascular compression, compression alone does not fully explain which nerves are painful. Recent work has demonstrated that pain in TN is associated with morphological changes of the trigeminal nerve, which may be a result of specific forms of compression or other disease processes. These changes have so far been quantified by hand-crafted geometric measurements, which can miss higher-order characteristics of the anatomy that are difficult to visualize. In this study, we use a cohort of 479 patients with TN who underwent microvascular decompression and had high-resolution pre-operative MRI scans to determine whether data-driven 3D shape features of the trigeminal nerve capture pain-associated morphology better than predefined metrics. We find that a predictive model using our shape features can distinguish nerves implicated in pain from those without pain in a held-out test cohort with an area under the receiver operating characteristic curve (AUROC) of 0.719, compared with 0.626 for a model using existing hand-crafted measurements (p=0.005). Our study identifies candidate nerve morphology biomarkers of TN pain through data-driven shape analysis and demonstrates that these higher-order features relate better to symptoms of TN than conventional shape metrics.
Source: medRxiv
Language: English
Subfield: Clinical Neurology
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7226012378
