Volumetric Pancreas Segmentation on Computed Tomography: Accuracy and Efficiency of a Convolutional Neural Network Versus Manual Segmentation in 3D Slicer in the Context of Interreader Variability of Expert Radiologists.

PURPOSE: This study aimed to compare accuracy and efficiency of a convolutional neural network (CNN)-enhanced workflow for pancreas segmentation versus radiologists in the context of interreader reliability. METHODS: Volumetric pancreas segmentations on a data set of 294 portal venous computed tomographies were performed by 3 radiologists (R1, R2, and R3) and by a CNN. Convolutional neural network segmentations were reviewed and, if needed, corrected ("corrected CNN [c-CNN]" segmentations) by radiologists. Ground truth was obtained from radiologists' manual segmentations using simultaneous truth and performance level estimation algorithm. Interreader reliability and model's accuracy were evaluated with Dice-Sorenson coefficient (DSC) and Jaccard coefficient (JC). Equivalence was determined using a two 1-sided test. Convolutional neural network segmentations below the 25th percentile DSC were reviewed to evaluate segmentation errors. Time for manual segmentation and c-CNN was compared. RESULTS: Pancreas volumes from 3 sets of segmentations (manual, CNN, and c-CNN) were noninferior to simultaneous truth and performance level estimation-derived volumes [76.6 cm 3 (20.2 cm 3 ), P < 0.05]. Interreader reliability was high (mean [SD] DSC between R2-R1, 0.87 [0.04]; R3-R1, 0.90 [0.05]; R2-R3, 0.87 [0.04]). Convolutional neural network segmentations were highly accurate (DSC, 0.88 [0.05]; JC, 0.79 [0.07]) and required minimal-to-no corrections (c-CNN: DSC, 0.89 [0.04]; JC, 0.81 [0.06]; equivalence, P < 0.05). Undersegmentation (n = 47 [64%]) was common in the 73 CNN segmentations below 25th percentile DSC, but there were no major errors. Total inference time (minutes) for CNN was 1.2 (0.3). Average time (minutes) taken by radiologists for c-CNN (0.6 [0.97]) was substantially lower compared with manual segmentation (3.37 [1.47]; savings of 77.9%-87% [ P < 0.0001]). CONCLUSIONS: Convolutional neural network-enhanced workflow provides high accuracy and efficiency for volumetric pancreas segmentation on computed tomography.

Giant fibrovascular polyp of the esophagus: report of a case.

A fibrovascular polyp is a peculiar nonepithelial tumor of the esophagus that invariably arises in the cervical esophagus at the level of the thoracic inlet and grows distally into a massive elongated, pedunculated, intraluminal lesion. Although it is a benign lesion that is eminently resectable, it is a dramatic entity owing to its tendency to cause bizarre complications such as asphyxia and sudden death when it regurgitates into the pharynx and causes laryngeal impaction. This report describes the multimodality imaging appearance of an archetypal case of a giant fibrovascular polyp in a patient with a seemingly innocuous presentation for the size of the lesion. The essential role of cross-sectional imaging in establishing a prompt diagnosis, defining the tissue elements of the mass, and delineation of the exact extent of the lesion in guiding the treatment approach is highlighted. The appearance of fibrovascular polyp in a single patient with a combination of barium swallow, multidetector computed tomography, and high-resolution contrast-enhanced magnetic resonance imaging has not been reported previously.

Disseminated cysticercosis in a child: whole-body MR diagnosis with the use of parallel imaging.

Cysticercosis is a parasitic disease caused by infestation with the encysted larval stage of the pork tapeworm, Taenia solium. Disseminated cysticercosis is an exceptional expression of this disease characterized by high morbidity due to massive symptomatic parasite burden in the central nervous system, striated muscles, subcutaneous tissues and other organs. Less than 50 such cases have been reported worldwide, and fewer than 10 children. We report on the whole-body MR diagnosis of extensively disseminated cysticercosis in a child. The critical role of whole-body MR as a stand-alone modality in the diagnosis and management of this pleomorphic disease is highlighted. Whole-body MR diagnosis of an infectious disease is unique.

MR-guided abdominal biopsy using a 1.5-Tesla closed system: a feasibility study.

BACKGROUND: MR-guided biopsy may aid to obtain samples in cases which are not feasible with conventional US or CT guidance. PURPOSE: To evaluate safety and efficacy of MRI-guided abdominal biopsy at 1.5-T closed MR system. METHODS AND MATERIALS: MRI guided abdominal biopsy was performed using Siemens Avanto 1.5-T closed MR system. Eighteen samples were obtained in 10 patients under local anesthesia using a novel technique to define skin entry site. None of the cases included in the study were amenable to biopsy using the conventional ultrasound (US) or computed tomography (CT) guidance. MR compatible 18G needle (US Biopsy) was used to obtain biopsy samples. Intravenous gadolinium was used in two patients for better delineation during biopsy. Patients were followed up for 3-5 h and were discharged on the same day. RESULTS: Technical success was achieved in all patients. Average number of biopsy passes were two (range 1-4). All biopsy samples were adequate for histopathological examination. Average size of the core biopsy specimen was 9 mm. CONCLUSION: 1.5-T closed MR system allows adequate biopsy sampling from various abdominal organs. This technique may help to sample lesions which are not otherwise amenable for biopsy under conventional CT or US guidance.

3-D MR imaging of ectopia vasa deferentia.

Ectopia vasa deferentia is a complex anomaly characterized by abnormal termination of the urethral end of the vas deferens into the urinary tract due to an incompletely understood developmental error of the distal Wolffian duct. Associated anomalies of the lower gastrointestinal tract and upper urinary tract are also commonly present due to closely related embryological development. Although around 32 cases have been reported in the literature, the MR appearance of this condition has not been previously described. We report a child with high anorectal malformation who was found to have ectopia vasa deferentia, crossed fused renal ectopia and type II caudal regression syndrome on MR examination. In addition to the salient features of this entity on reconstructed MR images, the important role of 3-D MRI in establishing an unequivocal diagnosis and its potential in facilitating individually tailored management is also highlighted.

Giant primary posterior mediastinal hydatid cyst in a child: report of a case and review of literature.

Primary posterior mediastinal cyst is an exceptionally uncommon manifestation of hydatid disease especially in pediatric age group. We herein present the account of a giant posterior mediastinal hydatid cyst in an 8-year-old boy that was peculiar due to the absence of typical clinical and radiological features of hydatid disease. The diagnosis was established perioperatively by a combination of surgical and pathological findings. The report depicts one of the myriad presentations of hydatid disease and also emphasizes the verity that it should be considered in the differential diagnoses of any mediastinal cystic lesion even in pediatric patients.