Abstract: An algorithm for simultaneous truth and performance estimation of various approaches for human cortical surface parcellation was proposed. The probabilistic true segmentation was estimated as a weighted combination of the segmentations resulted from multiple methods. Afterward,an Expectation-Maximization algorithm was used to optimize the weighting depending on the estimated performance level of each method. Furthermore, a spatial homogeneity constraint modeled by the Hidden Markov Random Field theory was incorporated to refine the estimated true segmentation into a spatially homogenous decision. The proposed method was evaluated using both synthetic and real data. Also, it was used to generate reference sulci regions to perform a comparison study of three methods for cortical surface parcel lation. The experimental results demonstrate the validity of the proposed method.

Key words: Cortical surface parcellation, Optimal combination, Performance estimation