We present a novel method for the joint segmentation of anatomical and functional images. Our proposed methodology unifies the domains of anatomical and functional images, represents them in a product lattice, and performs simultaneous delineation of regions based on random walk image segmentation. Furthermore, we also propose a simple yet effective object/background seed localization method to make the proposed segmentation process fully automatic. Our study uses PET, PET-CT, MRI-PET, and fused MRI-PET-CT scans (77 studies in all) from 56 patients who had various lesions in different body regions. We validated the effectiveness of the proposed method on different PET phantoms as well as on clinical images with respect to the ground truth segmentation provided by clinicians. Experimental results indicate that the presented method is superior to threshold and Bayesian methods commonly used in PET image segmentation, is more accurate and robust compared to the other PET-CT segmentation methods recently published in the literature, and also it is general in the sense of simultaneously segmenting multiple scans in real-time with high accuracy needed in routine clinical use.

BACKGROUND Increasing evidence supports the use of MRI-ultrasound fusion-targeted prostate biopsy (MRF-TB) to improve the detection of clinically significant prostate cancer (PCa) while limiting detection of indolent disease compared to systematic 12-core biopsy (SB). OBJECTIVE We report results of MRF-TB and SB and the relationship between biopsy outcomes and pre-biopsy MRI in 601 men presenting to our center. DESIGN/SETTING/PARTICIPANTS Retrospective analysis of a prospectively acquired cohort of men presenting for prostate biopsy over a 26-month period. A total of 601 of 803 consecutively eligible men were included. INTERVENTIONS All men were offered pre-biopsy MRI and assigned a maximum MRI suspicion score (mSS). Men with an MRI abnormality underwent combined MRF-TB and SB. OUTCOMES Detection rate of all PCa and high-grade PCa (Gleason score (GS)≥7) were compared by McNemar's test. RESULTS MRF-TB detected fewer GS6 PCa (75 vs 121, p<0.001) and more GS≥7 PCa (158 vs 117, p<0.001) than SB. Increasing mSS was associated with increasing detection of GS≥7 PCa (p<0.001), but had no relationship to the detection of GS6 PCa. The prediction of GS≥7 disease by mSS varied according to biopsy history. Compared to SB, MRF-TB identified more GS≥7 cancer in men with no prior biopsy (88 vs 72, p=0.012), with prior negative biopsy (28 vs 16, p=0.010), and with prior cancer diagnosis (42 vs 29, p=0.043). MRF-TB detected fewer GS6 cancers in men with no prior biopsy (32 vs 60, p<.001) and prior cancer (30 vs 46, p=0.034). Limitations include retrospective design and potential for selection bias given a referral population. CONCLUSIONS MRI-US fusion-targeted biopsy detects more high-grade cancer than systematic biopsy while limiting detection of GS6 cancer in men presenting for prostate biopsy. These findings suggest that pre-biopsy mpMRI and MRF-TB should be considered in all men undergoing prostate biopsy and, in conjunction with biopsy indication, mSS may ultimately help identify a select group of men at low risk of high-grade cancer in whom prostate biopsy may not be warranted.

A negative prebiopsy mpMRI confers an overall NPV of 82% on 12-core biopsy for all cancer and 98% for GS ≥7 cancer. Based on biopsy indication, these findings assist in prebiopsy risk stratification for detection of high-risk disease and may provide guidance in the decision to pursue biopsy.

In men selected for multiple repeat biopsies clinically significant cancer is found at each sampling round. Given the continued likelihood of cancer detection even by the fifth biopsy, early consideration of saturation or image guided biopsy may be warranted in the repeat biopsy population.