Ensemble learning of machine vision (ML) models, particularly deep neural networks, has emerged as a powerful technique for enhancing the quality of medical image processing. In this research endeavor, we investigated the application of the Deep Convolutional Neural Network (DnCNN) to the denoising of medical images. The DnCNN model was trained using a diverse dataset comprising medical images encompassing various noise levels and types. The primary objective was to train the model to restore a clean version of the image from its noisy counterpart. To evaluate the performance of the proposed approach, we employed widely-used quality metrics, namely the Structural Similarity Index (SSIM) and Peak Signal-to-Noise Ratio (PSNR). The experimental results demonstrated the efficacy of the DnCNN model in effectively reducing noise while preserving crucial image details, thus offering a promising solution for denoising medical images.
Paper
Full text
MRI Denoising Neural Network Architecture Convolution
Semantic Scholar · Medicine · 2023
Abstract
Ensemble learning of machine vision (ML) models, particularly deep neural networks, has emerged as a powerful technique for enhancing the quality of medical image processing. In this research endeavor, we investigated the application of the Deep Convolutional Neural Network (DnCNN) to the denoising of medical images. The DnCNN model was trained using a diverse dataset comprising medical images encompassing various noise levels and types. The primary objective was to train the model to restore a clean version of the image from its noisy counterpart. To evaluate the performance of the proposed approach, we employed widely-used quality metrics, namely the Structural Similarity Index (SSIM) and Peak Signal-to-Noise Ratio (PSNR). The experimental results demonstrated the efficacy of the DnCNN model in effectively reducing noise while preserving crucial image details, thus offering a promising solution for denoising medical images.