Denoising of Scintillation Camera Images Using a Deep Convolutional Neural Network : A Monte Carlo Simulation Approach
(2020) In Journal of Nuclear Medicine 61(2). p.298-303- Abstract
Scintillation camera images contain a large amount of Poisson noise. We have investigated whether noise can be removed in whole-body bone scans using convolutional neural networks (CNNs) trained with sets of noisy and noiseless images obtained by Monte Carlo simulation. Methods: Three CNNs were generated using 3 different sets of training images: simulated bone scan images, images of a cylindric phantom with hot and cold spots, and a mix of the first two. Each training set consisted of 40,000 noiseless and noisy image pairs. The CNNs were evaluated with simulated images of a cylindric phantom and simulated bone scan images. The mean squared error between filtered and true images was used as difference metric, and the coefficient of... (More)
Scintillation camera images contain a large amount of Poisson noise. We have investigated whether noise can be removed in whole-body bone scans using convolutional neural networks (CNNs) trained with sets of noisy and noiseless images obtained by Monte Carlo simulation. Methods: Three CNNs were generated using 3 different sets of training images: simulated bone scan images, images of a cylindric phantom with hot and cold spots, and a mix of the first two. Each training set consisted of 40,000 noiseless and noisy image pairs. The CNNs were evaluated with simulated images of a cylindric phantom and simulated bone scan images. The mean squared error between filtered and true images was used as difference metric, and the coefficient of variation was used to estimate noise reduction. The CNNs were compared with gaussian and median filters. A clinical evaluation was performed in which the ability to detect metastases for CNN- and gaussian-filtered bone scans with half the number of counts was compared with standard bone scans. Results: The best CNN reduced the coefficient of variation by, on average, 92%, and the best standard filter reduced the coefficient of variation by 88%. The best CNN gave a mean squared error that was on average 68% and 20% better than the best standard filters, for the cylindric and bone scan images, respectively. The best CNNs for the cylindric phantom and bone scans were the dedicated CNNs. No significant differences in the ability to detect metastases were found between standard, CNN-, and gaussian-filtered bone scans. Conclusion: Noise can be removed efficiently regardless of noise level with little or no resolution loss. The CNN filter enables reducing the scanning time by half and still obtaining good accuracy for bone metastasis assessment.
(Less)
- author
- Minarik, David LU ; Enqvist, Olof LU and Trägårdh, Elin LU
- organization
- publishing date
- 2020-02-01
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- artificial intelligence, image enhancement, machine learning, Monte Carlo, nuclear medicine
- in
- Journal of Nuclear Medicine
- volume
- 61
- issue
- 2
- pages
- 6 pages
- publisher
- Society of Nuclear Medicine
- external identifiers
-
- pmid:31324711
- scopus:85079021134
- ISSN
- 0161-5505
- DOI
- 10.2967/jnumed.119.226613
- language
- English
- LU publication?
- yes
- id
- f3873e0f-a61b-40e3-b45e-f51e572ea4d6
- date added to LUP
- 2020-02-19 13:33:50
- date last changed
- 2024-04-17 04:09:21
@article{f3873e0f-a61b-40e3-b45e-f51e572ea4d6,
abstract = {{<p>Scintillation camera images contain a large amount of Poisson noise. We have investigated whether noise can be removed in whole-body bone scans using convolutional neural networks (CNNs) trained with sets of noisy and noiseless images obtained by Monte Carlo simulation. Methods: Three CNNs were generated using 3 different sets of training images: simulated bone scan images, images of a cylindric phantom with hot and cold spots, and a mix of the first two. Each training set consisted of 40,000 noiseless and noisy image pairs. The CNNs were evaluated with simulated images of a cylindric phantom and simulated bone scan images. The mean squared error between filtered and true images was used as difference metric, and the coefficient of variation was used to estimate noise reduction. The CNNs were compared with gaussian and median filters. A clinical evaluation was performed in which the ability to detect metastases for CNN- and gaussian-filtered bone scans with half the number of counts was compared with standard bone scans. Results: The best CNN reduced the coefficient of variation by, on average, 92%, and the best standard filter reduced the coefficient of variation by 88%. The best CNN gave a mean squared error that was on average 68% and 20% better than the best standard filters, for the cylindric and bone scan images, respectively. The best CNNs for the cylindric phantom and bone scans were the dedicated CNNs. No significant differences in the ability to detect metastases were found between standard, CNN-, and gaussian-filtered bone scans. Conclusion: Noise can be removed efficiently regardless of noise level with little or no resolution loss. The CNN filter enables reducing the scanning time by half and still obtaining good accuracy for bone metastasis assessment.</p>}},
author = {{Minarik, David and Enqvist, Olof and Trägårdh, Elin}},
issn = {{0161-5505}},
keywords = {{artificial intelligence; image enhancement; machine learning; Monte Carlo; nuclear medicine}},
language = {{eng}},
month = {{02}},
number = {{2}},
pages = {{298--303}},
publisher = {{Society of Nuclear Medicine}},
series = {{Journal of Nuclear Medicine}},
title = {{Denoising of Scintillation Camera Images Using a Deep Convolutional Neural Network : A Monte Carlo Simulation Approach}},
url = {{http://dx.doi.org/10.2967/jnumed.119.226613}},
doi = {{10.2967/jnumed.119.226613}},
volume = {{61}},
year = {{2020}},
}