The goal of this observational study is to develop and validate a machine learning-based model for predicting pain recurrence risk after percutaneous balloon compression (PBC) in adult patients with primary trigeminal neuralgia (TN) who had their first PBC treatment. The main questions it aims to answer are:
Can the machine learning-based model accurately predict pain recurrence after PBC in these primary TN patients? What key factors (like patient baseline traits, imaging parameters, surgical operation data) affect PBC post-operative pain recurrence? Do machine learning algorithms perform better than traditional Cox proportional hazards regression in predicting such recurrence? Participants (with existing PBC treatment records) will have their past data-including clinical info from the hospital's electronic medical record system, imaging data from the image archiving system, surgical data from the surgical anesthesia system, and follow-up data from the outpatient system-collected and analyzed to build and validate the prediction model.
Aged ≥ 18 years.
Meet the diagnostic criteria for primary trigeminal neuralgia according to the International Classification of Headache Disorders, 3rd edition (ICHD-3).
Undergo the first PBC treatment.
Have complete preoperative clinical data, imaging data, and intraoperative records.
Secondary trigeminal neuralgia (e.g., caused by cerebellopontine angle tumors, multiple sclerosis, etc.).
Missing rate of key predictor variables (e.g., balloon shape) or outcome variables > 15%.
Postoperative loss to follow-up (defined as no follow-up records available).
Second Xiangya Hospital of Central South University
Lead sponsor
Yueyang Hospital of Traditional Chinese Medicine
Collaborator