Transformative Noise Reduction: Leveraging a Transformer-Based Deep Network for Medical Image Denoising

被引:18
作者
Naqvi, Rizwan Ali [1 ]
Haider, Amir [1 ]
Kim, Hak Seob [2 ]
Jeong, Daesik [3 ]
Lee, Seung-Won [4 ]
机构
[1] Sejong Univ, Dept AI & Robot, 209 Neungdong Ro, Seoul 05006, South Korea
[2] Korea Agcy Educ Promot & Informat Serv Food Agr Fo, Sejong 30148, South Korea
[3] Sangmyung Univ, Div Software Convergence, Seoul 03016, South Korea
[4] Sungkyunkwan Univ, Sch Med, Suwon 16419, South Korea
基金
新加坡国家研究基金会;
关键词
medical image denoising; deep-wider residual block; multi-head attention; multi-modal denoising; deep learning; GENERATIVE ADVERSARIAL NETWORK; SPARSE; CNN;
D O I
10.3390/math12152313
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
Medical image denoising has numerous real-world applications. Despite their widespread use, existing medical image denoising methods fail to address complex noise patterns and typically generate artifacts in numerous cases. This paper proposes a novel medical image denoising method that learns denoising using an end-to-end learning strategy. Furthermore, the proposed model introduces a novel deep-wider residual block to capture long-distance pixel dependencies for medical image denoising. Additionally, this study proposes leveraging multi-head attention-guided image reconstruction to effectively denoise medical images. Experimental results illustrate that the proposed method outperforms existing qualitative and quantitative evaluation methods for numerous medical image modalities. The proposed method can outperform state-of-the-art models for various medical image modalities. It illustrates a significant performance gain over its counterparts, with a cumulative PSNR score of 8.79 dB. The proposed method can also denoise noisy real-world medical images and improve clinical application performance such as abnormality detection.
引用
收藏
页数:21
相关论文
共 48 条
[1]  
Bhonsle D., 2012, INT J IMAGE GRAPH SI, V4, P36, DOI [10.5815/ijigsp.2012.06.06, DOI 10.5815/IJIGSP.2012.06.06]
[2]   Association of genomic subtypes of lower-grade gliomas with shape features automatically extracted by a deep learning algorithm [J].
Buda, Mateusz ;
Saha, Ashirbani ;
Mazurowski, Maciej A. .
COMPUTERS IN BIOLOGY AND MEDICINE, 2019, 109 :218-225
[3]   Low-Dose CT With a Residual Encoder-Decoder Convolutional Neural Network [J].
Chen, Hu ;
Zhang, Yi ;
Kalra, Mannudeep K. ;
Lin, Feng ;
Chen, Yang ;
Liao, Peixi ;
Zhou, Jiliu ;
Wang, Ge .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 2017, 36 (12) :2524-2535
[4]   Single Low-Dose CT Image Denoising Using a Generative Adversarial Network With Modified U-Net Generator and Multi-Level Discriminator [J].
Chi, Jianning ;
Wu, Chengdong ;
Yu, Xiaosheng ;
Ji, Peng ;
Chu, Hao .
IEEE ACCESS, 2020, 8 :133470-133487
[5]  
Cho KYHY, 2014, Arxiv, DOI [arXiv:1406.1078, DOI 10.48550/ARXIV.1406.1078]
[6]   Image denoising by sparse 3-D transform-domain collaborative filtering [J].
Dabov, Kostadin ;
Foi, Alessandro ;
Katkovnik, Vladimir ;
Egiazarian, Karen .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2007, 16 (08) :2080-2095
[7]   Feature-Guided CNN for Denoising Images From Portable Ultrasound Devices [J].
Dong, Guanfang ;
Ma, Yingnan ;
Basu, Anup .
IEEE ACCESS, 2021, 9 (09) :28272-28281
[8]   Deep CNN Model for Multimodal Medical Image Denoising [J].
El-Shafai, Walid ;
Mahmoud, Amira A. ;
Ali, Anas M. ;
El-Rabaie, El-Sayed M. ;
Taha, Taha E. ;
Zahran, Osama F. ;
El-Fishawy, Adel S. ;
Soliman, Naglaa F. ;
Alhussan, Amel A. ;
Abd El-Samie, Fathi E. .
CMC-COMPUTERS MATERIALS & CONTINUA, 2022, 73 (02) :3795-3814
[9]   Efficient Deep-Learning-Based Autoencoder Denoising Approach for Medical Image Diagnosis [J].
El-Shafai, Walid ;
Abd El-Nabi, Samy ;
El-Rabaie, El-Sayed M. ;
Ali, Anas M. ;
Soliman, Naglaa F. ;
Algarni, Abeer D. ;
Abd El-Samie, Fathi E. .
CMC-COMPUTERS MATERIALS & CONTINUA, 2022, 70 (03) :6107-6125
[10]   Image denoising via sparse and redundant representations over learned dictionaries [J].
Elad, Michael ;
Aharon, Michal .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2006, 15 (12) :3736-3745