Abstract: A robust and fast method was proposed to extract the inferior alveolar nerve (IAN) canals. First of all, the center points of an IAN canal on a series of cross-sectional images were identified by the division and the feature analysis of the connected regions on these images. Then, the centerline of this IAN canal was computed by using a spline interpolation. Finally, the IAN canal was extracted by adding the voxels with similar characteristics to the centerline region. The real dental CT images evaluation results showed that based on the proposed method all the IAN canals in the evaluation datasets could be successfully extracted with high accuracy.