Detection of arachnoid cysts in the brain using machine learning

Abstract

Cysts are sacs filled with fluid that can form in various organs, such as the kidneys, liver, breast, and brain. Treatment of these sacs may require surgical intervention. The importance of machine learning in detecting abnormal tissues in medical imaging is increasingly evident. This study specifically focuses on using deep learning structures to detect arachnoid cysts in the brain. The study employed Logistic Regression, InceptionV3, Kernel DVM, and Googlenet algorithms to detect arachnoid cysts. The accuracy rates achieved were 98.63% for Logistic Regression, 92.83% for InceptionV3, 92.38% for Kernel DVM, and 91.42% for Googlenet. Logistic Regression was the most successful algorithm. The study utilized data obtained from a 1.5T GE Magnetic Resonance Imaging (MRI) device.