Evolution of Deep Learning Algorithms for MRI-Based Brain Tumor Image Segmentation.

Brain tumor textures are among the most challenging features for neuroradiologists to extract from magnetic resonance images (MRIs). Exceptionally high-grade tumors such as gliomas require quick and precise diagnosis and medical intervention due to their infiltrative and fast-spreading nature. Therefore, they require computer assistance instead of manual methods. Deep learning (DL) methods are currently on the rise and have become an active field of research in several domains varying from stock market analysis to deep space object detection. They have very promising potential in brain tumor feature extraction from MRIs. Convolutional neural network (CNN) architectures, one of the most influential families of DL algorithms, have undergone a profound transformation since their first successes. This has led to increasing feature extraction quality and algorithm generalizability over various brain tumor types and grades. This review paper presents an explanatory and comparative survey on MRI-based brain tumor image segmentation. First, it provides the survey background and the typical process chain for brain MRI segmentation using CNNs. Second, it details the typical CNN architecture structure and its advantages over other machine learning algorithms. CNN architectures proposed for this purpose are enumerated and classified corresponding to their complexity, and then compared using specific metrics that consider the datasets they use.

Anorectal manometry with and without ketamine for evaluation of defecation disorders in children.

INTRODUCTION: Anorectal manometry (ARCM) provides valuable information in children with chronic constipation and fecal incontinence but may not be tolerated in the awake child. This study aimed to evaluate the effect of ketamine anesthesia on the assessment of anorectal function by manometry and to evaluate defecation dynamics and anal sphincter resting pressure in the context of pathophysiology of chronic functional (idiopathic) constipation and soiling in children. METHODS: This was a prospective study of children who were investigated for symptoms of chronic constipation and soiling between April 2001 and April 2004. We studied 52 consecutive children who had awake ARCM, biofeedback training and endosonography (awake group) and 64 children who had ketamine anesthesia for ARCM and endosonography (ketamine group). We age matched 31 children who had awake anorectal studies with 27 who had ketamine anesthesia. RESULTS: The children in awake and ketamine groups were comparable for age, duration of bowel symptoms and duration of laxative treatments. ARCM profile was comparable between the awake and the ketamine groups with regard to anal sphincter resting pressure, rectal capacity, amplitude of rectal contractions, frequency of rectal and IAS contractions and functional length of anal canal. Of 52 children who had awake ARCM, dyssynergia of the EAS muscles was observed in 22 (42%) and median squeeze pressure was 87mm Hg (range 25-134). The anal sphincter resting pressure was non-obstructive and comparable to healthy normal children. Rectoanal inhibitory reflex was seen in all children excluding diagnosis of Hirschsprung disease. CONCLUSIONS: Ketamine anesthesia does not affect quantitative or qualitative measurements of autonomic anorectal function and can be used reliably in children who will not tolerate the manometry while awake. Paradoxical contraction of the EAS can only be evaluated in the awake children and should be investigated further as the underlying cause of obstructive defecation in patients with chronic functional constipation and soiling.

Cryptorchidism and the value of ultrasonography.

INTRODUCTION: The aim of this study was to evaluate the usefulness and accuracy of ultrasonography in investigating cryptorchidism. METHODS: Case notes were reviewed of all children with cryptorchidism who had undergone ultrasonography for their testes between February and October 2010. Preoperative ultrasonography findings for presence or absence and position of testes were noted and compared with the operative findings. The sensitivity and specificity of ultrasonography were calculated. RESULTS: A total of 50 testes in 42 boys were examined by ultrasonography preoperatively during the study period. The median age was 6 years. All had single B-mode greyscale ultrasonography using greyscale and Doppler imaging. Of the 50 testes, 42 were visualised intracanalicularly or in the superficial inguinal pouch, 2 were intra-abdominal and 6 could not be detected. However, on examination under anaesthesia, 37 out of the 50 testes were palpable in the groin and 13 were impalpable. Two out of the remaining thirteen were found to be atrophic on diagnostic laparoscopy. The sensitivity and specificity of ultrasonography in determining the presence of testes was 85% and 25% respectively. CONCLUSIONS: Ultrasonography has low sensitivity and specificity in determining the presence of testes and localisation of their position. When performing an examination under anaesthesia and using a laparoscopic technique, ultrasonography is not recommended in cases of cryptorchidism.

Duodenal atresia: associated anomalies, prenatal diagnosis and outcome.

BACKGROUND: The diagnosis of duodenal atresia is commonly made prenatally, either as an isolated lesion or due to its association with other chromosomal abnormalities (Robertson et al. in Semin Perinatol 18:182-195, 1994; Hemming and Rankin in J Prenat Diagn 27:1205-1211, 2007). The aim of this study was to describe the prevalence of associated anomalies, prenatal diagnostic accuracy and survival of cases of congenital duodenal atresia in our institution. METHODS: All cases of duodenal atresia registered with our local congenital anomaly register over a 10-year period, 1995-2004 inclusive, were studied, including those resulting in termination of pregnancies, stillbirths, intrauterine deaths and neonatal deaths. To ensure high-case ascertainment, data were cross checked with prenatal ultrasound, cytogenetic laboratory, pathology department and neonatal surgical data base. Data were analysed for associated anomalies, accuracy of prenatal diagnosis and neonatal outcomes. RESULTS: A total of 65 patients were initially diagnosed as having duodenal atresia, of these 4 were subsequently excluded (1 postnatal normal bowel and 3 high jejunal atresias). In the remaining 61 cases, 35 (57%) had an association with other congenital abnormalities and 26 (43%) were isolated anomalies. Thirty-five were male and 26 female (M:F = 1.4:1). Twenty-one out of 29 (72%) patients prenatally diagnosed, compared with 14 out of 32 (44%) patients diagnosed postnatally had associated anomalies. Duodenal atresia was suspected on routine prenatal ultrasonography at 20-week gestation in 33 cases and confirmed in 29 (48%) cases with 4 false-positive diagnoses (1 normal bowel and 3 high jejunal atresias). No prenatal diagnosis was made in 32 (52%) babies. Of the 61 cases, 53 were live births with 2 early neonatal deaths (1 cardiac and 1 VACTERL), 5 terminations, 2 intrauterine deaths and 1 stillbirth (Fig. 3). Overall neonatal survival was 96% (51 cases). Mortality in the group diagnosed prenatally was 34 % (10 cases). CONCLUSION: This study shows an overall increased association of duodenal atresia with Down's syndrome. In the group diagnosed prenatally, mortality as well as the association with other congenital anomalies was found to be higher. We have demonstrated a greater prenatal diagnostic accuracy, but confirm postnatal outcomes similar to previous studies.