Related Experiment Video
Updated: Oct 20, 2025
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
MR Denoising Increases Radiomic Biomarker Precision and Reproducibility in Oncologic Imaging
Matías Fernández Patón1, Leonor Cerdá Alberich2, Cinta Sangüesa Nebot3
1Grupo de Investigación Biomédica en Imagen, Instituto de Investigación Sanitaria La Fe, Avenida Fernando Abril Martorell, 106 Torre A 7planta, 46026, Valencia, Spain. matias_fernandez@iislafe.es.
Abstract:
Several noise sources, such as the Johnson-Nyquist noise, affect MR images disturbing the visualization of structures and affecting the subsequent extraction of radiomic data. We evaluate the performance of 5 denoising filters (anisotropic diffusion filter (ADF), curvature flow filter (CFF), Gaussian filter (GF), non-local means filter (NLMF), and unbiased non-local means (UNLMF)), with 33 different settings, in T2-weighted MR images of phantoms (N = 112) and neuroblastoma patients (N = 25). Filters were discarded until the most optimal solutions were obtained according to 3 image quality metrics: peak signal-to-noise ratio (PSNR), edge-strength similarity-based image quality metric (ESSIM), and noise (standard deviation of the signal intensity of a region in the background area). The selected filters were ADFs and UNLMs. From them, 107 radiomics features preservation at 4 progressively added noise levels were studied. The ADF with a conductance of 1 and 2 iterations standardized the radiomic features, improving reproducibility and quality metrics.
Insights
Anisotropic diffusion filters (ADF) and unbiased non-local means (UNLM) effectively reduce noise in MRI scans. ADF with specific settings improved radiomic feature standardization, enhancing reproducibility and image quality for neuroblastoma patients.
Area of Science:
- Medical Imaging
- Radiology
- Image Processing
Background:
- Magnetic Resonance (MR) images are susceptible to noise, such as Johnson-Nyquist noise, which hinders visualization and radiomic data extraction.
- Effective noise reduction is crucial for accurate image analysis and diagnostic interpretation in medical imaging.
Purpose of the Study:
- To evaluate the performance of five denoising filters in T2-weighted MR images.
- To identify optimal filters and settings for preserving radiomic features and improving image quality.
- To assess the impact of noise levels on radiomic feature preservation.
Main Methods:
- Evaluated anisotropic diffusion filter (ADF), curvature flow filter (CFF), Gaussian filter (GF), non-local means filter (NLMF), and unbiased non-local means (UNLMF) with 33 settings.
- Utilized T2-weighted MR images from 112 phantoms and 25 neuroblastoma patients.
- Selected optimal filters based on peak signal-to-noise ratio (PSNR), edge-strength similarity-based image quality metric (ESSIM), and noise standard deviation.
Main Results:
- Anisotropic diffusion filters (ADF) and unbiased non-local means (UNLMF) were identified as the most effective denoising filters.
- ADF with a conductance of 1 and 2 iterations demonstrated superior performance in standardizing radiomic features.
- Preservation of 107 radiomic features was studied across four increasing noise levels, showing improved reproducibility and quality metrics with ADF.
Conclusions:
- ADF and UNLMF are promising for noise reduction in T2-weighted MR images.
- Specific ADF settings enhance radiomic feature standardization, leading to improved reproducibility and image quality.
- The findings support the use of optimized denoising techniques for robust radiomic analysis in neuroblastoma research.
