Incorporating CT prior information in the robust fuzzy C-means algorithm for QSPECT image segmentation

摘要

Bones are a common site of metastases in a number of cancers including prostate and breast cancer. Assessingresponse or progression typically relies on planar bone scintigraphy. However, quantitative bone SPECT(BQSPECT) has the potential to provide more accurate assessment. An important component of BQSPECTis segmenting lesions and bones in order to calculate metrics like tumor uptake and metabolic tumor burden.However, due to the poor spatial resolution, noise, and contrast properties of SPECT images, segmentation ofbone SPECT images is challenging. In this study, we propose and evaluate a fuzzy C-means (FCM) clusteringbased semi-automatic segmentation method on quantitative Tc-99m MDP quantitative SPECT/CT. The FCMclustering algorithm has been widely used in medical image segmentation. Yet, the poor resolution and noiseproperties of SPECT images result in sub-optimal segmentation. We propose to incorporate information fromregistered CT images, which can be used to segment normal bones quite readily, into the FCM segmentationalgorithm. The proposed method modifies the objective function of the robust fuzzy C-means (RFCM) methodto include prior information about bone from CT images and spatial information from the SPECT image to allowfor simultaneously segmenting lesion and bone in BQSPECT/CT images. The method was evaluated usingrealistic simulated BQSPECT images. The method and algorithm parameters were evaluated with respect tothe dice similarity coefficient (DSC) computed using segmentation results. The effect of the number of iterationsused to reconstruct the BQSPECT images was also studied. For the simulated BQSPECT images studied, anaverage DSC value of 0.75 was obtained for lesions larger than 2 cm~3 with the proposed method.