Showing 1–6 of 6 results for author: Dekker, A

DELTA-MRI: Direct deformation Estimation from LongiTudinally Acquired k-space data

Authors: Jens Renders, Banafshe Shafieizargar, Marleen Verhoye, Jan De Beenhouwer, Arnold J. den Dekker, Jan Sijbers

Abstract: Longitudinal MRI is an important diagnostic imaging tool for evaluating the effects of treatment and monitoring disease progression. However, MRI, and particularly longitudinal MRI, is known to be time consuming. To accelerate imaging, compressed sensing (CS) theory has been applied to exploit sparsity, both on single image as on image sequence level. State-of-the-art CS methods however, are gener… ▽ More Longitudinal MRI is an important diagnostic imaging tool for evaluating the effects of treatment and monitoring disease progression. However, MRI, and particularly longitudinal MRI, is known to be time consuming. To accelerate imaging, compressed sensing (CS) theory has been applied to exploit sparsity, both on single image as on image sequence level. State-of-the-art CS methods however, are generally focused on image reconstruction, and consider analysis (e.g., alignment, change detection) as a post-processing step. In this study, we propose DELTA-MRI, a novel framework to estimate longitudinal image changes {\it directly} from a reference image and subsequently acquired, strongly sub-sampled MRI k-space data. In contrast to state-of-the-art longitudinal CS based imaging, our method avoids the conventional multi-step process of image reconstruction of subsequent images, image alignment, and deformation vector field computation. Instead, the set of follow-up images, along with motion and deformation vector fields that describe their relation to the reference image, are estimated in one go. Experiments show that DELTA-MRI performs significantly better than the state-of-the-art in terms of the normalized reconstruction error. △ Less

Submitted 23 January, 2023; originally announced January 2023.

Comments: 5 pages, 4 figures, Submitted to ISBI 2023

Lung-Originated Tumor Segmentation from Computed Tomography Scan (LOTUS) Benchmark

Authors: Parnian Afshar, Arash Mohammadi, Konstantinos N. Plataniotis, Keyvan Farahani, Justin Kirby, Anastasia Oikonomou, Amir Asif, Leonard Wee, Andre Dekker, Xin Wu, Mohammad Ariful Haque, Shahruk Hossain, Md. Kamrul Hasan, Uday Kamal, Winston Hsu, Jhih-Yuan Lin, M. Sohel Rahman, Nabil Ibtehaz, Sh. M. Amir Foisol, Kin-Man Lam, Zhong Guang, Runze Zhang, Sumohana S. Channappayya, Shashank Gupta, Chander Dev

Abstract: Lung cancer is one of the deadliest cancers, and in part its effective diagnosis and treatment depend on the accurate delineation of the tumor. Human-centered segmentation, which is currently the most common approach, is subject to inter-observer variability, and is also time-consuming, considering the fact that only experts are capable of providing annotations. Automatic and semi-automatic tumor… ▽ More Lung cancer is one of the deadliest cancers, and in part its effective diagnosis and treatment depend on the accurate delineation of the tumor. Human-centered segmentation, which is currently the most common approach, is subject to inter-observer variability, and is also time-consuming, considering the fact that only experts are capable of providing annotations. Automatic and semi-automatic tumor segmentation methods have recently shown promising results. However, as different researchers have validated their algorithms using various datasets and performance metrics, reliably evaluating these methods is still an open challenge. The goal of the Lung-Originated Tumor Segmentation from Computed Tomography Scan (LOTUS) Benchmark created through 2018 IEEE Video and Image Processing (VIP) Cup competition, is to provide a unique dataset and pre-defined metrics, so that different researchers can develop and evaluate their methods in a unified fashion. The 2018 VIP Cup started with a global engagement from 42 countries to access the competition data. At the registration stage, there were 129 members clustered into 28 teams from 10 countries, out of which 9 teams made it to the final stage and 6 teams successfully completed all the required tasks. In a nutshell, all the algorithms proposed during the competition, are based on deep learning models combined with a false positive reduction technique. Methods developed by the three finalists show promising results in tumor segmentation, however, more effort should be put into reducing the false positive rate. This competition manuscript presents an overview of the VIP-Cup challenge, along with the proposed algorithms and results. △ Less

Submitted 2 January, 2022; originally announced January 2022.

Generative Models Improve Radiomics Performance in Different Tasks and Different Datasets: An Experimental Study

Authors: Junhua Chen, Inigo Bermejo, Andre Dekker, Leonard Wee

Abstract: Radiomics is an active area of research focusing on high throughput feature extraction from medical images with a wide array of applications in clinical practice, such as clinical decision support in oncology. However, noise in low dose computed tomography (CT) scans can impair the accurate extraction of radiomic features. In this article, we investigate the possibility of using deep learning gene… ▽ More Radiomics is an active area of research focusing on high throughput feature extraction from medical images with a wide array of applications in clinical practice, such as clinical decision support in oncology. However, noise in low dose computed tomography (CT) scans can impair the accurate extraction of radiomic features. In this article, we investigate the possibility of using deep learning generative models to improve the performance of radiomics from low dose CTs. We used two datasets of low dose CT scans -NSCLC Radiogenomics and LIDC-IDRI - as test datasets for two tasks - pre-treatment survival prediction and lung cancer diagnosis. We used encoder-decoder networks and conditional generative adversarial networks (CGANs) trained in a previous study as generative models to transform low dose CT images into full dose CT images. Radiomic features extracted from the original and improved CT scans were used to build two classifiers - a support vector machine (SVM) and a deep attention based multiple instance learning model - for survival prediction and lung cancer diagnosis respectively. Finally, we compared the performance of the models derived from the original and improved CT scans. Encoder-decoder networks and CGANs improved the area under the curve (AUC) of survival prediction from 0.52 to 0.57 (p-value<0.01). On the other hand, Encoder-decoder network and CGAN can improve the AUC of lung cancer diagnosis from 0.84 to 0.88 and 0.89 respectively (p-value<0.01). Moreover, there are no statistically significant differences in improving AUC by using encoder-decoder network and CGAN (p-value=0.34) when networks trained at 75 and 100 epochs. Generative models can improve the performance of low dose CT-based radiomics in different tasks. Hence, denoising using generative models seems to be a necessary pre-processing step for calculating radiomic features from low dose CTs. △ Less

Submitted 6 September, 2021; originally announced September 2021.

Recurrent Inference Machines as inverse problem solvers for MR relaxometry

Authors: E. R. Sabidussi, S. Klein, M. W. A. Caan, S. Bazrafkan, A. J. den Dekker, J. Sijbers, W. J. Niessen, D. H. J. Poot

Abstract: In this paper, we propose the use of Recurrent Inference Machines (RIMs) to perform T1 and T2 map**. The RIM is a neural network framework that learns an iterative inference process based on the signal model, similar to conventional statistical methods for quantitative MRI (QMRI), such as the Maximum Likelihood Estimator (MLE). This framework combines the advantages of both data-driven and model… ▽ More In this paper, we propose the use of Recurrent Inference Machines (RIMs) to perform T1 and T2 map**. The RIM is a neural network framework that learns an iterative inference process based on the signal model, similar to conventional statistical methods for quantitative MRI (QMRI), such as the Maximum Likelihood Estimator (MLE). This framework combines the advantages of both data-driven and model-based methods, and, we hypothesize, is a promising tool for QMRI. Previously, RIMs were used to solve linear inverse reconstruction problems. Here, we show that they can also be used to optimize non-linear problems and estimate relaxometry maps with high precision and accuracy. The developed RIM framework is evaluated in terms of accuracy and precision and compared to an MLE method and an implementation of the ResNet. The results show that the RIM improves the quality of estimates compared to the other techniques in Monte Carlo experiments with simulated data, test-retest analysis of a system phantom, and in-vivo scans. Additionally, inference with the RIM is 150 times faster than the MLE, and robustness to (slight) variations of scanning parameters is demonstrated. Hence, the RIM is a promising and flexible method for QMRI. Coupled with an open-source training data generation tool, it presents a compelling alternative to previous methods. △ Less

Submitted 8 June, 2021; originally announced June 2021.

Comments: 12 pages, 9 figures

Generative Models Improve Radiomics Reproducibility in Low Dose CTs: A Simulation Study

Authors: Junhua Chen, Chong Zhang, Alberto Traverso, Ivan Zhovannik, Andre Dekker, Leonard Wee, Inigo Bermejo

Abstract: Radiomics is an active area of research in medical image analysis, the low reproducibility of radiomics has limited its applicability to clinical practice. This issue is especially prominent when radiomic features are calculated from noisy images, such as low dose computed tomography (CT) scans. In this article, we investigate the possibility of improving the reproducibility of radiomic features c… ▽ More Radiomics is an active area of research in medical image analysis, the low reproducibility of radiomics has limited its applicability to clinical practice. This issue is especially prominent when radiomic features are calculated from noisy images, such as low dose computed tomography (CT) scans. In this article, we investigate the possibility of improving the reproducibility of radiomic features calculated on noisy CTs by using generative models for denoising.One traditional denoising method - non-local means - and two generative models - encoder-decoder networks (EDN) and conditional generative adversarial networks (CGANs) - were selected as the test models. We added noise to the sinograms of full dose CTs to mimic low dose CTs with two different levels of noise: low-noise CT and high-noise CT. Models were trained on high-noise CTs and used to denoise low-noise CTs without re-training. We also test the performance of our model in real data, using dataset of same-day repeat low dose CTs to assess the reproducibility of radiomic features in denoised images. The EDN and the CGAN improved the concordance correlation coefficients (CCC) of radiomic features for low-noise images from 0.87 to 0.92 and for high-noise images from 0.68 to 0.92 respectively. Moreover, the EDN and the CGAN improved the test-retest reliability of radiomic features (mean CCC increased from 0.89 to 0.94) based on real low dose CTs. The results show that denoising using EDN and CGANs can improve the reproducibility of radiomic features calculated on noisy CTs. Moreover, images with different noise levels can be denoised to improve the reproducibility using these models without re-training, as long as the noise intensity is equal or lower than that in high-noise CTs. To the authors' knowledge, this is the first effort to improve the reproducibility of radiomic features calculated on low dose CT scans. △ Less

Submitted 13 August, 2021; v1 submitted 30 April, 2021; originally announced April 2021.

Journal ref: Physics in Medicine & Biology 66.16 (2021) 165002

Lung Cancer Diagnosis Using Deep Attention Based on Multiple Instance Learning and Radiomics

Authors: Junhua Chen, Haiyan Zeng, Chong Zhang, Zhenwei Shi, Andre Dekker, Leonard Wee, Inigo Bermejo

Abstract: Early diagnosis of lung cancer is a key intervention for the treatment of lung cancer computer aided diagnosis (CAD) can play a crucial role. However, most published CAD methods treat lung cancer diagnosis as a lung nodule classification problem, which does not reflect clinical practice, where clinicians diagnose a patient based on a set of images of nodules, instead of one specific nodule. Beside… ▽ More Early diagnosis of lung cancer is a key intervention for the treatment of lung cancer computer aided diagnosis (CAD) can play a crucial role. However, most published CAD methods treat lung cancer diagnosis as a lung nodule classification problem, which does not reflect clinical practice, where clinicians diagnose a patient based on a set of images of nodules, instead of one specific nodule. Besides, the low interpretability of the output provided by these methods presents an important barrier for their adoption. In this article, we treat lung cancer diagnosis as a multiple instance learning (MIL) problem in order to better reflect the diagnosis process in the clinical setting and for the higher interpretability of the output. We chose radiomics as the source of input features and deep attention-based MIL as the classification algorithm.The attention mechanism provides higher interpretability by estimating the importance of each instance in the set for the final diagnosis.In order to improve the model's performance in a small imbalanced dataset, we introduce a new bag simulation method for MIL.The results show that our method can achieve a mean accuracy of 0.807 with a standard error of the mean (SEM) of 0.069, a recall of 0.870 (SEM 0.061), a positive predictive value of 0.928 (SEM 0.078), a negative predictive value of 0.591 (SEM 0.155) and an area under the curve (AUC) of 0.842 (SEM 0.074), outperforming other MIL methods.Additional experiments show that the proposed oversampling strategy significantly improves the model's performance. In addition, our experiments show that our method provides an indication of the importance of each nodule in determining the diagnosis, which combined with the well-defined radiomic features, make the results more interpretable and acceptable for doctors and patients. △ Less

Submitted 12 February, 2022; v1 submitted 29 April, 2021; originally announced April 2021.