Automatic Segmentation of the Cisternal Segment of Trigeminal Nerve on MRI Using Deep Learning

Area of Science:

• Medical imaging
• Artificial intelligence in medicine
• Neurosurgery

Background:

• Accurate segmentation of the cisternal trigeminal nerve is crucial for diagnosing and treating trigeminal nerve disorders like trigeminal neuralgia (TN).
• Manual segmentation is time-consuming and subject to interobserver variability, limiting its clinical utility.

Purpose of the Study:

• To develop and evaluate an automated deep learning-based approach, U-Net, for segmenting the cisternal trigeminal nerve.
• To improve the accuracy and efficiency of trigeminal nerve segmentation compared to manual methods.

Main Methods:

• A U-Net deep learning model was trained and validated on healthy control MRI data.
• The model was tested on a separate dataset of patients with trigeminal neuralgia (TN).
• Segmentation performance was assessed using Dice, Jaccard, PPV, SEN, CMD, and Hausdorff distance metrics.

Main Results:

• The U-Net model achieved high accuracy in segmenting the cisternal trigeminal nerve.
• The automated segmentation performance was comparable to that of participating radiologists.
• The method demonstrated robust performance across different datasets.

Conclusions:

• The U-Net deep learning approach offers a promising solution for accurate and efficient cisternal trigeminal nerve segmentation.
• This is the first fully automated segmentation method for the trigeminal nerve in anatomical MRI.
• The technique has the potential to significantly aid in the diagnosis and treatment of trigeminal nerve disorders, including TN.