Variations in cochlear duct shape revealed on clinical CT images with an automatic tracing method

Area of Science:

• Otolaryngology
• Medical Imaging
• Biomedical Engineering

Background:

• Cochlear size and shape significantly impact cochlear implant (CI) electrode insertion and final position.
• Understanding individual cochlear morphology is crucial for predicting surgical outcomes and speech perception.
• Variations in cochlear anatomy can pose challenges during CI surgery.

Purpose of the Study:

• To develop and validate an automatic tracing method for assessing individual cochlear duct shapes from clinical CT images.
• To model cochlear wall morphology using a logarithmic spiral function.
• To identify regions at risk for CI-related insertion trauma based on cochlear shape.

Main Methods:

• Automatic tracing of cochlear walls (outer, inner, scala tympani floor) from pre-operative CT scans of 479 inner ears.
• Interpolation of voxel intensities along radial and perpendicular lines in multiplanar reconstructions.
• Modeling of cochlear wall shapes using a logarithmic spiral function with an offset value.

Main Results:

• Successfully traced the outer cochlear wall in all 479 cochleas and the inner wall/scala tympani floor in 192 cochleas.
• Modeled cochlear wall shapes with a logarithmic spiral function, revealing non-monotonous spiral slopes.
• Identified three distinct slope categories for the scala tympani, highlighting regions susceptible to CI insertion trauma.

Conclusions:

• The automated tracing method provides a detailed description of cochlear morphology.
• This technique is valuable for individual cochlear characterization and large-scale evaluation of cochlear implant patients.
• The findings can inform surgical planning and potentially improve CI outcomes by mitigating insertion risks.