Related Experiment Video
Updated: Oct 10, 2025
06:48
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
9.1K
Automatic Segmentation of Intracochlear Anatomy in MR Images Using a Weighted Active Shape Model
Summary
A new automated method accurately segments cochlear anatomy in MRI scans for predicting hearing loss after acoustic neuroma surgery. This technique aids large studies and clinical use.
Area of Science:
- Medical Imaging
- Neurosurgery
- Otolaryngology
Background:
- Cochlear MR signal intensity shows potential for predicting hearing loss risk after middle cranial fossa (MCF) resection of acoustic neuroma (AN).
- Manual segmentation of intra-cochlear anatomy in MR images is challenging and error-prone, limiting retrospective studies and clinical application.
Purpose of the Study:
- To develop and validate a fully automated method for segmenting intra-cochlear anatomy in MR images.
Main Methods:
- Utilized a weighted active shape model, previously validated for CT image segmentation.
- Applied the method to 132 ears from 66 high-resolution T2-weighted MR images, using CT segmentation as ground truth.
Main Results:
- Achieved a mean Dice Similarity Coefficient (DSC) of 0.81 for the scala tympani (ST) and 0.79 for the scala vestibuli (SV).
- Demonstrated accurate and fully automated segmentation of intra-cochlear structures in MR images.
Conclusions:
- The automated method provides accurate segmentation of intra-cochlear anatomy in MR images.
- This technique can support large retrospective studies correlating preoperative MR signal with patient outcomes.
- Facilitates routine clinical use of MR imaging for predicting hearing loss after acoustic neuroma surgery.
