Sex, Age and Body Side Influences in Cochlear Morphology from 500 pairs of CT Images

BACKGROUND: The cochlea shows a great variability in size and morphology. An improved knowledge of the cochlea shape is not only essential for diagnosis of shape abnormality, atraumatic surgery planning and postoperative insertion assessment but also provides better anatomical understanding and can suggest electrode design improvements for cochlear implant manufacturers. OBJECTIVE: 1.Demonstrate the important contribution of medical imaging for the study of cochlear morphology and 2.Study the variability and the correlation between cochlear size measurements and sex, age and body side. METHODS: The database consists of 429 males and 554 females, between 10 and 97 years of age with 448 left and 535 right sides. Automatic segmentation methods were used to analyzed the CT images and the variation in cochlear base (typical A and B-axis), height and volume were assessed. RESULTS: The average segmentation error between automatically segmented low-resolution CT images and corresponding manually segmented high-resolution µCT images is equal to 0.12mm while CT images have a voxel size of 0.2x0.2x0.25mm³. Large variation in cochlear size were found with, amongst others, A-axis ranging from 7 to 9.5mm and volume from 60 to 100mm³ (and both are correlated, r=0.48). Average cochlear volume is 3% larger among males than females (p<<0.05). No statistically significant (p>0.05) linear correlation between age or ear-side and cochlear size were found. Dissimilarity measure is non-zero but smaller with intrapatient bilateral data (0.61) than with interpatient data (0.71). CONCLUSIONS: This study demonstrates a large variability in cochear anatomical measurements. The bilateral symmetry hypothesis is far from being observed.