Background This paper attempts to identify suitable Machine Learning (ML) approach for image denoising of radiology based medical application. The Identification of ML approach is based on (i) Review of ML approach for denoising (ii) Review of suitable Medical Denoising approach. Discussion The review focuses on six application of radiology: Medical Ultrasound (US) for fetus development, US Computer Aided Diagnosis (CAD) and detection for breast, skin lesions, brain tumor MRI diagnosis, X-Ray for chest analysis, Breast cancer using MRI imaging. This survey identifies the ML approach with better accuracy for medical diagnosis by radiologists. The image denoising approaches further includes basic filtering techniques, wavelet medical denoising, curvelet and optimization techniques. In most of the applications, the machine learning performance is better than the conventional image denoising techniques. For fast and computational results the radiologists are using the machine learning methods on MRI, US, X-Ray and Skin lesion images. The characteristics and contributions of different ML approaches are considered in this paper. Conclusion The problem faced by the researchers during image denoising techniques and machine learning applications for clinical settings have also been discussed.

/pdf/a-review-of-denoising-medical-images-using-machine-learning-4rhj429a77.pdf

A Review of Denoising Medical Images Using Machine Learning Approaches

To suppress the heavy noise and keep the distinct edges of the images in the low light condition, we propose a denoising model based on the combination of total variation (TV) and nonlocal similarity in the wavelet domain. The TV regularization in the wavelet domain effectively suppresses the heavy noise with the biorthogonal wavelet function; the nonlocal similarity regularization improves the fine image details. Denoising experiments on artificially degraded and low light images show that in the heavy noise condition, the proposed denoising model can suppress the heavy noise effectively and preserve the detail of images than several state-of-the-art methods.

Wavelet-Based Total Variation and Nonlocal Similarity Model for Image Denoising

Feature extraction is one of the challenging problems in fault diagnosis, and it has a direct bearing on the accuracy of fault diagnosis. Therefore, in this paper, a new method based on ensemble empirical mode decomposition (EEMD), wavelet semi-soft threshold (WSST) signal reconstruction, and multi-scale entropy (MSE) is proposed. First, the EEMD method is applied to decompose the vibration signal into intrinsic mode functions (IMFs), and then, the high-frequency IMFs, which contain more noise information, are screened by the Pearson correlation coefficient. Then, the WSST method is applied for denoising the high-frequency part of the signal to reconstruct the signal. Secondly, the MSE method is applied for calculating the MSE values of the reconstructed signal, to construct an eigenvector with the complexity measure. Finally, the eigenvector is input to a support vector machine (SVM) to find the fault diagnosis results. The experimental results prove that the proposed method, with a better classification performance, can better solve the problem of the effective signal and noise mixed in high-frequency signals. Based on the proposed method, the fault types can be accurately identified with an average classification accuracy of 100%.

https://www.mdpi.com/1099-4300/22/3/290/pdf

A Rolling Bearing Fault Diagnosis Method Based on EEMD-WSST Signal Reconstruction and Multi-Scale Entropy.

Image compression plays a key role in the transmission of an image and storage capacity. Image compression aims to reduce the size of the image with no loss of significant information and no loss o...

https://journals.sagepub.com/doi/pdf/10.1177/0020294019877508

An efficient JPEG image compression based on Haar wavelet transform, discrete cosine transform, and run length encoding techniques for advanced manufacturing processes

The image often contains noises due to several factors such as a problem in devices or due to an environmental problem etc. Noise is mainly undesired information, which degrades the quality of the picture. Therefore, image denoising method is adopted to remove the noises from the degraded image which in turn improve the quality of the image. In this paper, image denoising has been done by wavelet transform using Visu thresholding techniques for different wavelet families. PSNR (Peak signal to noise ratio) and RMSE (Root Mean Square Error) value is also calculated for different wavelet families.

/pdf/image-denoising-based-on-wavelet-transform-using-visu-1t4ha4hbzn.pdf

Image Denoising Based on Wavelet Transform using Visu Thresholding Technique

Mammogram is an easy and affordable means of diagnosis of breast cancer. Like other medical data, it is also affected by noise during acquisition. Therefore, it is a challenge for the researchers to denoise the mammograms for clear data extraction. This paper aims at denoising the mammogram by the wavelet and the curvelet transform with a motive to investigate the role of the “embedded” thresholding algorithm. As the thresholding technique is a key factor for the noise reduction, a comprehensive study on the employment of the different types of the thresholding techniques with the transforms have been presented methodically. A standard mammogram from the Mammographic Image Analysis Society database is supplemented with different types of noise and then denoised by the wavelet and the curvelet transforms using the three commonly used thresholding techniques to compare the denoising performance of the thresholding algorithms along with the transforms. This investigation renders a clear insight to th...

Soft, Hard and Block Thresholding Techniques for Denoising of Mammogram Images

We present here an experimental work on mammogram denoising by the mathematical tool called the curvelet transform. The infiltration of noise in mammogram during the X-ray screening is a common and inevitable phenomenon. And such noise is normally reduced by the curvelet transform based on conventional thresholding strategy called the hard thresholding (HT). Therefore, the motive of this experimentation is to suggest an alternate but efficient mechanism of mammogram denoising by the same transform but with different algorithms purely based on the information of the neighbouring coefficients. It is found that the curvelet transform applied with the precept of surrounding curvelet coefficients is visually and statistically better than the conventional approach based on HT.

Mammogram denoising by curvelet transform based on the information of neighbouring coefficients

This chapter introduces two mathematical transforms—wavelet and curvelet—in the field of biomedical imaging. Presenting the theoretical background with relevant properties, the applications of the two transforms are presented. The biomedical applications include heart sound analysis, electrocardiography (ECG) characterization, positron emission tomography (PET) image analysis, medical image compression, mammogram enhancement, magnetic resonance imaging (MRI) and computer tomography (CT) image denoising, diabetic retinopathy detection. The applications emphasize the development of algorithms to diagnose human diseases, thereby rendering fast and reliable support to the medical personnel. The transforms—one classical (wavelet) and another contemporary (curvelet)—are selected to focus the difference in architecture, limitation, evolution, and application of individual transform. Two joint applications are addressed to compare their performance. This survey is also supplemented by a case study: mammogram denoising using wavelet and curvelet transforms with the underlying algorithms.

Wavelet and Curvelet Transforms for Biomedical Image Processing

We present an experimental work on the denoising of mammogram with Poisson noise. Reviewing the literature, it is found that the denoising performance of the multiresolution tools like wavelet, contourlet and curvelet implemented on mammogram with Poisson noise is unique. The first part of the investigation deals with the confirmation of this exceptional performance with our result. The later half implements the recently developed denoising approach called the Poisson Unbiased Risk Estimation-Linear Expansion of Thresholds (PURE-LET) to the Poisson noise corrupted mammogram with an objective to improve the peak signal to noise ratio (PSNR) further. The PURE-LET successfully removes Poisson noise better than the traditional mathematical transforms already mentioned. The computation time and PSNR are also evaluated in the perspective of the cycle spinning technique. This validates the applicability and efficiency of the novel denoising strategy in the field of digital mammography.

Poisson Noise Removal from Mammogram Using Poisson Unbiased Risk Estimation Technique

Over the last few years, the wavelet transform has played a significant role in the field of multiresolution image analysis. The shortcomings of the wavelet transform laid the foundation of many advanced wavelets. This review paper brings together ten advanced wavelets on a common platform to discuss their importance, concept, architecture, merits and demerits in various fields of image processing. The relationships among the different advanced wavelets are also illustrated here. A comparison table serves as a catalog to know the recent trends and applications of the advanced wavelets.

Manas Saha

Papers

Soft, Hard and Block Thresholding Techniques for Denoising of Mammogram Images

Mammogram denoising by curvelet transform based on the information of neighbouring coefficients

Wavelet and Curvelet Transforms for Biomedical Image Processing

Poisson Noise Removal from Mammogram Using Poisson Unbiased Risk Estimation Technique

Advanced Wavelet Transform for Image Processing—A Survey