Abbreviations & AcronymsObjectives: To present an overview of our surgical experience in the management of growing teratoma syndrome. Methods: A retrospective analysis of all patients undergoing post-chemotherapy retroperitoneal lymphadenectomy between November 2005 and February 2012 revealed 15 patients who met the criteria for growing teratoma syndrome. Their clinical data, imaging characteristics, and surgical and oncological outcomes were reviewed. Results:The median age at diagnosis was 23 years. Primary testis tumors included nonseminomatous germ cell tumor in 12 of 15 patients, seminoma in two of 15 patients and hemorrhagic mass in one patient. Mature teratoma was present in just six (40%) of the orchiectomy specimens. All patients received preoperative chemotherapy. On imaging, the median size of the largest retroperitoneal mass was 7 cm (range 3.9-24.5 cm). The median rate of linear growth was 0.5 cm/month (range 0.03-2.9), and the increase in volume was 9.2 cm 3 /month. All tumors were found to have cystic and necrotic components. Median operative time was 6.2 h (range 4.2-15.2 h). Estimated blood loss was 600 mL (range 100-7000 mL), and median length of stay was 5 days (range 3-19 days). Four patients required resection of non-retroperitoneal growing teratoma masses after post-chemotherapy retroperitoneal lymphadenectomy to achieve tumor-free status. There were two minor (Clavien I-II) and two major postoperative complications (Clavien ≥III). All patients are alive and disease free with a median duration of follow-up of 8 months (range 1-64 months). Conclusions: Growing teratoma syndrome tumors vary in their growth rate, but they all appear to have cystic features with necrosis elements on radiographic evaluation. Aggressive surgical excision is associated with excellent outcomes.

Our objective was to quantitatively compare maternal reactions to viewing a three-dimensional (3D) ultrasound image of the foetal face to a traditional two-dimensional (2D) sonographic image. One hundred and twelve pregnant women were asked to evaluate their excitement, relief, amazement and satisfaction reactions to 2D and 3D ultrasound images of their foetus' face. A weighted kappa Cochran-Armitage trend test, Fisher exact test and generalised estimating equations were used to analyse the data obtained. 3D imaging was found to result in significantly more favourable reactions than 2D imaging of the foetal face. 3D imaging was significantly better than 2D in regard to the clearness of the image, feeling closer to the baby, experiencing amazement and recognising specific facial features. We hypothesised that 3D images might evoke more feelings of fear than 2D, but our analysis found there were no significant differences in this regard. 3D ultrasonography of the foetal face is well received by patients. It has been shown in this study to elicit stronger and more positive maternal reactions regarding excitement, amazement and satisfaction than experienced with 2D ultrasound.

The purpose of this pilot study is to determine whether a deep convolutional neural network can be trained with limited image data to detect high-grade small bowel obstruction patterns on supine abdominal radiographs. Grayscale images from 3663 clinical supine abdominal radiographs were categorized into obstructive and non-obstructive categories independently by three abdominal radiologists, and the majority classification was used as ground truth; 74 images were found to be consistent with small bowel obstruction. Images were rescaled and randomized, with 2210 images constituting the training set (39 with small bowel obstruction) and 1453 images constituting the test set (35 with small bowel obstruction). Weight parameters for the final classification layer of the Inception v3 convolutional neural network, previously trained on the 2014 Large Scale Visual Recognition Challenge dataset, were retrained on the training set. After training, the neural network achieved an AUC of 0.84 on the test set (95% CI 0.78-0.89). At the maximum Youden index (sensitivity + specificity-1), the sensitivity of the system for small bowel obstruction is 83.8%, with a specificity of 68.1%. The results demonstrate that transfer learning with convolutional neural networks, even with limited training data, may be used to train a detector for high-grade small bowel obstruction gas patterns on supine radiographs.

No abstract

Kidney segmentation (KS) system is an objective and reproducible radiologic method of universally reporting tumour location according to 12 renal segments. By adding descriptive information on tumour characteristics in candidates for nephron-sparing surgery, this novel KS system could serve as an adjunct to current nephrometry systems.