A Bibliometric Review: Brain Tumor Magnetic Resonance Imagings Using Different Convolutional Neural Network Architectures

Abstract

Background: Numerous scientists and researchers have been developing advanced procedures and methods for diagnosing the kind and phase of a human tumor. Brain tumors, which are neoplastic and abnormal developments of brain cells, are one of the most prominent causes of death. Brain tumors, also known as lesions or neoplasia, may be roughly classified as either primary or metastatic. Primary brain tumors arise from brain tissue and its surrounding environment. The recognition of brain tumors using magnetic resonance images via a deep learning technique such as convolutional neural network (CNN) has garnered significant academic interest over the last few decades.

Methods: In this study, a detailed evaluation based on bibliometrics is considered in order to synthesize and organize the available academic literature and to identify current research trends and hotspots. We used bibliometric methodologies and a literature review for the CNN-based brain tumor to synthesize and evaluate prior studies.

Results: For this bibliometric analysis, we applied the Visualization of Similarity Viewer program to classify the major publications, notable journals, financial sponsors, and affiliations.

Conclusions: In conclusion, we suggest that one of the next paths of study will be the incorporation of other databases to advance CNN-based brain tumor identification from magnetic resonance images. No drug dosages are applied in this work.

Keywords: Brain tumor; CNN; Deep learning; MRI.