Autism Spectrum Disorder Detection by Hybrid Convolutional Recurrent Neural Networks from Structural and Resting State Functional MRI Images.

This study aims to increase the accuracy of autism spectrum disorder (ASD) diagnosis based on cognitive and behavioral phenotypes through multiple neuroimaging modalities. We apply machine learning (ML) algorithms to classify ASD patients and healthy control (HC) participants using structural magnetic resonance imaging (s-MRI) together with resting state functional MRI (rs-f-MRI and f-MRI) data from the large multisite data repository ABIDE (autism brain imaging data exchange) and identify important brain connectivity features. The 2D f-MRI images were converted into 3D s-MRI images, and datasets were preprocessed using the Montreal Neurological Institute (MNI) atlas. The data were then denoised to remove any confounding factors. We show, by using three fusion strategies such as early fusion, late fusion, and cross fusion, that, in this implementation, hybrid convolutional recurrent neural networks achieve better performance in comparison to either convolutional neural networks (CNNs) or recurrent neural networks (RNNs). The proposed model classifies subjects as autistic or not according to how functional and anatomical connectivity metrics provide an overall diagnosis based on the autism diagnostic observation schedule (ADOS) standard. Our hybrid network achieved an accuracy of 96% by fusing s-MRI and f-MRI together, which outperforms the methods used in previous studies.

Evaluation of the Turkish translation of the Minimal Standard Terminology for Digestive Endoscopy by development of an endoscopic information system.

BACKGROUND/AIMS: There are very few evaluation studies for the Minimal Standard Terminology for Digestive Endoscopy. This study aims to evaluate the usage of the Turkish translation of Minimal Standard Terminology by developing an endoscopic information system. METHODS: After elicitation of requirements, database modeling and software development were performed. Minimal Standard Terminology driven forms were designed for rapid data entry. The endoscopic report was rapidly created by applying basic Turkish syntax and grammar rules. Entering free text and also editing of final report were possible. After three years of live usage, data analysis was performed and results were evaluated. RESULTS: The system has been used for reporting of all endoscopic examinations. 15,638 valid records were analyzed, including 11,381 esophagogastroduodenoscopies, 2,616 colonoscopies, 1,079 rectoscopies and 562 endoscopic retrograde cholangiopancreatographies. In accordance with other previous validation studies, the overall usage of Minimal Standard Terminology terms was very high: 85% for examination characteristics, 94% for endoscopic findings and 94% for endoscopic diagnoses. Some new terms, attributes and allowed values were also added for better clinical coverage. CONCLUSIONS: Minimal Standard Terminology has been shown to cover a high proportion of routine endoscopy reports. Good user acceptance proves that both the terms and structure of Minimal Standard Terminology were consistent with usual clinical thinking. However, future work on Minimal Standard Terminology is mandatory for better coverage of endoscopic retrograde cholangiopancreatographies examinations. Technically new software development methodologies have to be sought for lowering cost of development and the maintenance phase. They should also address integration and interoperability of disparate information systems.