CMM-Net: Contextual multi-scale multi-level network for efficient biomedical image segmentation

Mohammed A Al-Masni,Dong-Hyun Kim

doi:10.1038/s41598-021-89686-3

Mohammed A Al-Masni, Dong-Hyun Kim

Open Access

https://doi.org/10.1038/s41598-021-89686-3

Copy DOI

Journal: Scientific Reports	Publication Date: May 13, 2021
Citations: 31	License type: open-access

Affiliation: Yonsei University

Abstract

Medical image segmentation of tissue abnormalities, key organs, or blood vascular system is of great significance for any computerized diagnostic system. However, automatic segmentation in medical image analysis is a challenging task since it requires sophisticated knowledge of the target organ anatomy. This paper develops an end-to-end deep learning segmentation method called Contextual Multi-Scale Multi-Level Network (CMM-Net). The main idea is to fuse the global contextual features of multiple spatial scales at every contracting convolutional network level in the U-Net. Also, we re-exploit the dilated convolution module that enables an expansion of the receptive field with different rates depending on the size of feature maps throughout the networks. In addition, an augmented testing scheme referred to as Inversion Recovery (IR) which uses logical “OR” and “AND” operators is developed. The proposed segmentation network is evaluated on three medical imaging datasets, namely ISIC 2017 for skin lesions segmentation from dermoscopy images, DRIVE for retinal blood vessels segmentation from fundus images, and BraTS 2018 for brain gliomas segmentation from MR scans. The experimental results showed superior state-of-the-art performance with overall dice similarity coefficients of 85.78%, 80.27%, and 88.96% on the segmentation of skin lesions, retinal blood vessels, and brain tumors, respectively. The proposed CMM-Net is inherently general and could be efficiently applied as a robust tool for various medical image segmentations.

Highlights

Medical image segmentation of tissue abnormalities, key organs, or blood vascular system is of great significance for any computerized diagnostic system
Computer-Aided Diagnosis (CAD) system is an indispensable tool that aims to provide assistance to clinicians through interpretations of the abnormalities existing in the medical images such as brain tumors in MR images[1], liver nodules and pulmonary lung nodules in Computed Tomography (CT) images[2,3], breast masses in mammograms[4], and skin lesions in dermoscopy images[5]
We present a version of Time Augmentation (TTA) for image segmentation tasks called Inversion Recovery (IR), which is performed as a post-processing step applied to augmented test data

Summary

Introduction

Medical image segmentation of tissue abnormalities, key organs, or blood vascular system is of great significance for any computerized diagnostic system. Medical imaging is an approach that generates interior visual representations of the hidden internal structures inside the human body This process could be applied noninvasively such as Magnetic Resonance Imaging (MRI), Computed Tomography (CT), X-ray, Ultrasound (US), endoscope, ophthalmoscopy, and dermoscopy modalities. The annotation process itself is extremely laborious, subjective, prone to error, and time-consuming at large biomedical data In this regard, efficient automated segmentation algorithms are highly demanded in clinical applications for more accurate analysis and diagnosis. The proposed method achieves state-of-the-art segmentation performance on three different medical imaging modalities (i.e., skin lesions in dermoscopy images, retinal blood vessels in fundus images, and brain glioma tumor in MR images)

Methods

Results

Conclusion