TY - JOUR
T1 - Adaptive tuning noise estimation for medical images using maximum element convolution Laplacian
AU - Ting, Fung Fung
AU - Sim, Kok Swee
N1 - Funding Information:
We would like to acknowledge Mammographic Image Analysis Society (MIAS) [27] for providing the real patient digital mammogram dataset. Fundamental Research Grant Scheme (FRGS) funds this research; it is a research grant from the Ministry of Higher Education (MOHE), Malaysia.
Funding Information:
Acknowledgement. We would like to acknowledge Mammographic Image Analysis Society (MIAS) [27] for providing the real patient digital mammogram dataset. Fundamental Research Grant Scheme (FRGS) funds this research; it is a research grant from the Ministry of Higher Education (MOHE), Malaysia.
Publisher Copyright:
© 2020 ICIC International.
Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.
PY - 2020/2
Y1 - 2020/2
N2 - Noise in medical images can adversely affect the outcome of clinical diagnosis. In analyzing medical images, noise estimation is necessary to ensure consistency and performance quality of image processing techniques. In this study, we present a noise estimation method, namely Adaptive Tuning Noise Estimation (ATNE) that implements convolution Laplacian noise estimation. ATNE is based on subtraction of Gabor wavelet detected edges of images, and involves the relation element based on the parameters of the input image. This method allows a fast estimation of the image noise variance without a heavy computational cost. To assess the effectiveness of ATNE, 1000 mammograms are used. We pre-process these images to be Rician distributed with various noise variances. ATNE is used to estimate the noise level of the resulting images. We compare ATNE with other noise estimation methods, and the results show that ATNE outperforms other related methods with a lower percentage of error for noise variance estimation.
AB - Noise in medical images can adversely affect the outcome of clinical diagnosis. In analyzing medical images, noise estimation is necessary to ensure consistency and performance quality of image processing techniques. In this study, we present a noise estimation method, namely Adaptive Tuning Noise Estimation (ATNE) that implements convolution Laplacian noise estimation. ATNE is based on subtraction of Gabor wavelet detected edges of images, and involves the relation element based on the parameters of the input image. This method allows a fast estimation of the image noise variance without a heavy computational cost. To assess the effectiveness of ATNE, 1000 mammograms are used. We pre-process these images to be Rician distributed with various noise variances. ATNE is used to estimate the noise level of the resulting images. We compare ATNE with other noise estimation methods, and the results show that ATNE outperforms other related methods with a lower percentage of error for noise variance estimation.
KW - Image noise estimation
KW - Image processing
KW - Medical imaging
KW - Rician noise
UR - http://www.scopus.com/inward/record.url?scp=85079092973&partnerID=8YFLogxK
U2 - 10.24507/ijicic.16.01.1
DO - 10.24507/ijicic.16.01.1
M3 - Article
AN - SCOPUS:85079092973
VL - 16
SP - 1
EP - 14
JO - International Journal of Innovative Computing Information and Control
JF - International Journal of Innovative Computing Information and Control
SN - 1349-4198
IS - 1
ER -