Treffer: A Clinical Decision Support System of Brain Tumor Classification Using Deep Learning Based on Image Processing.

Background: The use of intelligent medical diagnosis systems in diagnosing brain diseases, as an assistant alongside physicians and radiologists, in addition to helping them, paves the way for accurate and error-free identification to identify and distinguish these diseases from other similar diseases. Objective: The main goal of this study is to design a clinical decision-making system based on deep learning. Material and Methods: The brain tumor images dataset was downloaded from the Kaggle database in this study. A total of 3064 MRI images of brain tumor patients were used. The classification of the images was done in two stages; these stages include the training stage (equal to 90% of the images) and the testing stage (equal to 10% of images). We developed a 2D Convolutional Neural Network (2D-CNN). This network consists of eight convolution layers and four pooling layers, and after all convolution layers, batch-normalization layers were applied. The clinical diagnostic system was created using the Java programming language. Results: The training accuracy of the 2D convolutional neural network is 96.47%, and the testing accuracy is 93.44%. The average precision, recall, and F-measure are 94.75%, 95.75%, and 95%, respectively. The developed net was used as the processing core in the decision support system. Conclusion: The use of a deep neural network in this system by automatically extracting features from the image and increasing the final accuracy of the system made it possible for neurologists to solve very complex problems to a great extent. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Biomedical Physics & Engineering is the property of Shiraz University of Medical Sciences and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)