A Comprehensive Study of Applying Object Detection Methods for Medical Image Analysis

Abstract

Medical imaging is a widely accepted technique for the early detection and diagnosis of disease within digital health. It includes different techniques such as Magnetic resonance imaging (MRI), X-ray, positron emission tomography (PET) scan. Human experts mostly perform the analysis of these images. However, recent advancement in the field of computer-assisted interventions shows the promising results for medical image analysis. With the availability of enormous data, sophisticated algorithms, and high computation power, deep neural networks are highly effective for image analysis and interpretation tasks. Medical image analysis can be performed using the object detection method, where a convolutional neural network (CNN) eliminates the need for manual feature extraction. Object detection using CNN able to extract features directly from images and provides good accuracy. This paper exhibits a detailed survey on applications of different object detection methods available for medical image analysis. This paper discusses the different techniques, state-of-the-art datasets, tools, techniques available, and performance metrics. It also presents the work carried out by various researchers for applying object detection methods for medical image analysis.