Prostate cancer (PCa) is the most common cancer in men in the United States. Multiparametic magnetic resonance imaging (mp-MRI) has been explored by many researchers to targeted prostate biopsies and radiation therapy. However, assessment on mp-MRI can be subjective, development of computer-aided diagnosis systems to automatically delineate the prostate gland and the intraprostratic lesions (ILs) becomes important to facilitate with radiologists in clinical practice. In this paper, we first study the implementation of the Mask-RCNN model to segment the prostate and ILs. We trained and evaluated models on 120 patients from two different cohorts of patients. We also used 2D U-Net and 3D U-Net as benchmarks to segment the prostate and compared the model's performance. The contour variability of ILs using the algorithm was also benchmarked against the interobserver variability between two different radiation oncologists on 19 patients. Our results indicate that the Mask-RCNN model is able to reach state-of-art performance in the prostate segmentation and outperforms several competitive baselines in ILs segmentation.Automated segmentation of the prostate and screening of prostate cancer from MR images is critical for computer-aided clinical diagnosis, treatment planning and prognosis. However, the development of automatic algorithms remains challenging in several reasons. First of all, there are large variations in image quality caused by several factors at the time of image acquisition (e.g. patient motion, signal-to-noise ratio, use of an endorectal coil, Gadolinium enhancement, etc.). Second, the normal anatomy of the prostate is highly variable across patients and at different time points; and the boundaries between the prostate and surrounding structures (e.g. neurovascular bundles, bladder, rectum, seminal vessels and other soft tissues) are not always immediately clear. The prostate also shows a large variation in size and shape among different patients due to individual differences and physiological changes. Third, the presence of benign conditions such as benign prostatic hyperplasia and prostatitis may mimic the radiographic presentation of a malignancy. Contrast and pixel value of MRI also highlight a large variability in both tissue and texture information.Over the past few years, progress in image segmentation tasks has been exclusively driven by convolutional neural network (CNN) based models. Many segmentation models fall into two classes. The first class does not rely on the region proposal algorithm. Typical models in this class usually apply an encoder-decoder framework (Liou et al., 2014). The encoder network extracts representations of the image, and the decoder network reconstructs segmentation mask from the learned image representations produced by the encoder network. U-Net (Ronneberger et al., 2015), for instance, is a classic algorithm widely used in biomed-ical image segmentation tasks. Another class of models have their underlying fundamentals on region proposals such as the Mask-RCNN model,...
Background & Aims Many cancers in the proximal colon develop via from sessile serrated adenomas (SSAs), which have flat, subtle features that are difficult to detect with conventional white-light colonoscopy. Many SSA cells have the V600E mutation in BRAF. We investigated whether this feature could be used with imaging methods to detect SSAs in patients. Methods We used phage display to identify a peptide that binds specifically to SSAs, using subtractive hybridization with HT29 colorectal cancer cells containing the V600E mutation in BRAF and Hs738.St/Int cells as a control. Binding of fluorescently labeled peptide to colorectal cancer cells was evaluated with confocal fluorescence microscopy. Rats received intra-colonic 0.0086 mg/kg, 0.026 mg/kg, or 0.86 mg/kg peptide or vehicle and morbidity, mortality, and injury were monitored twice daily to assess toxicity. In the clinical safety study, fluorescently labeled peptide was topically administered, using a spray catheter, to the proximal colon of 25 subjects undergoing routine outpatient colonoscopies (3 subjects were given 2.25 µmol/L and 22 patients were given 76.4 µmol/L). We performed blood cell count, chemistry, liver function, and urine analyses approximately 24 hrs after peptide administration. In the clinical imaging study, 38 subjects undergoing routine outpatient colonoscopies, at high risk for colorectal cancer, or with a suspected unresected proximal colonic polyp, were first evaluated by white-light endoscopy, to identify suspicious regions. The fluorescently labeled peptide (76.4 µmol/L) was administered topically to proximal colon, unbound peptide was washed away, and white-light, reflectance, and fluorescence videos were recorded digitally. Fluorescence intensities of SSAs were compared with those of normal colonic mucosa. Endoscopists resected identified lesions, which were analyzed histologically by gastrointestinal pathologists (reference standard). We also analyzed the ability of the peptide to identify SSAs vs adenomas, hyperplastic polyps, and normal colonic mucosa in specimens obtained from the tissue bank at the University of Michigan. Results We identified the peptide sequence KCCFPAQ, and measured an apparent dissociation constant of kd = 72 nM and an apparent association time constant of k = 0.174 min−1 (5.76 min). During fluorescence imaging of patients during endoscopy, regions of SSA had 2.43-fold higher mean fluorescence intensity than that for normal colonic mucosa. Fluorescence labeling distinguished SSAs from normal colonic mucosa with 89% sensitivity and 92% specificity. The peptide had no observed toxic effects in animals or patients. In the analysis of ex vivo specimens, peptide bound to SSAs had significantly higher mean fluorescence intensity than to hyperplastic polyps. Conclusions We have identified a fluorescently labeled peptide that has no observed toxic effects in animals or humans and can be used for wide-field imaging of lesions in the proximal colon. It distinguishes SSAs from normal colonic mucosa with 89%...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.