Faculty Publications
Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736
Publications by NITK Faculty
Browse
10 results
Search Results
Item A new non-local maximum likelihood estimation method for Rician noise reduction in magnetic resonance images using the Kolmogorov-Smirnov test(Elsevier, 2014) Rajan, J.; den Dekker, A.J.; Sijbers, J.Denoising algorithms play an important role in the enhancement of magnetic resonance (MR) images. Effective denoising is vital for proper analysis and accurate quantitative measurements from MR images. Maximum Likelihood (ML) estimation methods were proved to be very effective in denoising MR images. Among the ML based methods, the recently proposed non-local maximum likelihood (NLML) approach gained much attention. In the NLML method, the samples for the ML estimation of the true underlying intensity are selected in a non-local way based on the intensity similarity of the pixel neighborhoods. This similarity is generally measured using the Euclidean distance. A drawback of this approach is the usage of a fixed sample size for the ML estimation resulting in over- or under-smoothing. In this work, we propose an NLML estimation method for denoising MR images in which the samples are selected in an adaptive and statistically supported way using the Kolmogorov-Smirnov (KS) similarity test. The method has been tested both on simulated and real data, showing its effectiveness. © 2013 Elsevier B.V.Item Nonlocal linear minimum mean square error methods for denoising MRI(Elsevier Ltd, 2015) Sudeep, P.V.; Ponnusamy, P.; Kesavadas, C.; Rajan, J.The presence of noise results in quality deterioration of magnetic resonance (MR) images and thus limits the visual inspection and influence the quantitative measurements from the data. In this work, an efficient two stage linear minimum mean square error (LMMSE) method is proposed for the enhancement of magnitude MR images in which data in the presence of noise follows a Rician distribution. The conventional Rician LMMSE estimator determines a closed-form analytical solution to the aforementioned inverse problem. Even-though computationally efficient, this approach fails to take advantage of data redundancy in the 3D MR data and hence leads to a suboptimal filtering performance. Motivated by this observation, we put forward the concept of nonlocal implementation with LMMSE estimation method. To select appropriate samples for the nonlocal version of the LMMSE estimation, the similarity weights are computed using Euclidean distance between either the gray level values in the spatial domain or the coefficients in the transformed domain. Assuming that the signal dependent component of the noise is optimally suppressed by this filtering and the rest is a white and uncorrelated noise with the image, we adopt a second stage LMMSE filtering in the principal component analysis (PCA) domain to further enhance the image and the noise variance is adaptively adjusted. Experiments on both simulated and real data show that the proposed filters have excellent filtering performance over other state-of-the-art methods. © 2015 Elsevier Ltd. All rights reserved.Item Magnetic resonance image denoising using nonlocal maximum likelihood paradigm in DCT-framework(John Wiley and Sons Inc, 2015) Kumar, P.K.; Darshan, P.; Kumar, S.; Ravindra, R.; Rajan, J.; Saba, L.; Suri, J.S.The data acquired by magnetic resonance (MR) imaging system are inherently degraded by noise that has its origin in the thermal Brownian motion of electrons. Denoising can enhance the quality (by improving the SNR) of the acquired MR image, which is important for both visual analysis and other post processing operations. Recent works on maximum likelihood (ML) based denoising shows that ML methods are very effective in denoising MR images and has an edge over the other state-of-the-art methods for MRI denoising. Among the ML based approaches, the Nonlocal maximum likelihood (NLML) method is commonly used. In the conventional NLML method, the samples for the ML estimation of the unknown true pixel are chosen in a nonlocal fashion based on the intensity similarity of the pixel neighborhoods. Euclidean distance is generally used to measure this similarity. It has been recently shown that computing similarity measure is more robust in discrete cosine transform (DCT) subspace, compared with Euclidean image subspace. Motivated by this observation, we integrated DCT into NLML to produce an improved MRI filtration process. Other than improving the SNR, the time complexity of the conventional NLML can also be significantly reduced through the proposed approach. On synthetic MR brain image, an average improvement of 5% in PSNR and 86%reduction in execution time is achieved with a search window size of 91 × 91 after incorporating the improvements in the existing NLML method. On an experimental kiwi fruit image an improvement of 10% in PSNR is achieved. We did experiments on both simulated and real data sets to validate and to demonstrate the effectiveness of the proposed method. © 2015 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 25, 256-264, 2015 © 2015 Wiley Periodicals, Inc.Item Iterative bilateral filter for Rician noise reduction in MR images(Springer London, 2015) Riji, R.; Rajan, J.; Sijbers, J.; Nair, M.S.Noise removal from magnetic resonance images is important for further processing and visual analysis. Bilateral filter is known for its effectiveness in edge-preserved image denoising. In this paper, an iterative bilateral filter for filtering the Rician noise in the magnitude magnetic resonance images is proposed. The proposed iterative bilateral filter improves the denoising efficiency, preserves the fine structures and also reduces the bias due to Rician noise. The visual and diagnostic quality of the image is well preserved. The quantitative analysis based on the standard metrics like peak signal-to-noise ratio and mean structural similarity index matrix shows that the proposed method performs better than the other recently proposed denoising methods for MRI. © 2014, Springer-Verlag London.Item GPU implementation of non-local maximum likelihood estimation method for denoising magnetic resonance images(Springer Verlag service@springer.de, 2017) Upadhya, A.H.K.; Talawar, B.; Rajan, J.Magnetic resonance imaging (MRI) is a widely deployed medical imaging technique used for various applications such as neuroimaging, cardiovascular imaging and musculoskeletal imaging. However, MR images degrade in quality due to noise. The magnitude MRI data in the presence of noise generally follows a Rician distribution if acquired with single-coil systems. Several methods are proposed in the literature for denoising MR images corrupted with Rician noise. Amongst the methods proposed in literature for denoising MR images corrupted with Rician noise, the non-local maximum likelihood methods (NLML) and its variants are popular. In spite of the performance and denoising quality, NLML algorithm suffers from a tremendous time complexity O(m3N3) , where m3 and N3 represent the search window and image size, respectively, for a 3D image. This makes the algorithm challenging for deployment in the real-time applications where fast and prompt results are required. A viable solution to this shortcoming would be the application of a data parallel processing framework such as Nvidia CUDA so as to utilize the mutually exclusive and computationally intensive calculations to our advantage. The GPU-based implementation of NLML-based image denoising achieves significant speedup compared to the serial implementation. This research paper describes the first successful attempt to implement a GPU-accelerated version of the NLML algorithm. The main focus of the research was on the parallelization and acceleration of one computationally intensive section of the algorithm so as to demonstrate the execution time improvement through the application of parallel processing concepts on a GPU. Our results suggest the possibility of practical deployment of NLML and its variants for MRI denoising. © 2016, Springer-Verlag Berlin Heidelberg.Item A nonlocal maximum likelihood estimation method for enhancing magnetic resonance phase maps(Springer London, 2017) Sudeep, P.V.; Ponnusamy, P.; Kesavadas, C.; Sijbers, J.; den Dekker, A.J.; Rajan, J.A phase map can be obtained from the real and imaginary components of a complex valued magnetic resonance (MR) image. Many applications, such as MR phase velocity mapping and susceptibility mapping, make use of the information contained in the MR phase maps. Unfortunately, noise in the complex MR signal affects the measurement of parameters related to phase (e.g, the phase velocity). In this paper, we propose a nonlocal maximum likelihood (NLML) estimation method for enhancing phase maps. The proposed method estimates the true underlying phase map from a noisy MR phase map. Experiments on both simulated and real data sets indicate that the proposed NLML method has a better performance in terms of qualitative and quantitative evaluations when compared to state-of-the-art methods. © 2016, Springer-Verlag London.Item Automatic detection and localization of Focal Cortical Dysplasia lesions in MRI using fully convolutional neural network(Elsevier Ltd, 2019) Bijay Dev, K.M.; Pawan, P.S.; Niyas, S.; Vinayagamani, S.; Kesavadas, C.; Rajan, J.Focal cortical dysplasia (FCD) is the leading cause of drug-resistant epilepsy in both children and adults. At present, the only therapeutic approach in patients with drug-resistant epilepsy is surgery. Hence, the quantification of FCD via non-invasive imaging techniques helps physicians to decide on surgical interventions. The properties like non-invasiveness and capability to produce high-resolution images makes magnetic resonance imaging an ideal tool for detecting the FCD to an extent. The FCD lesions vary in size, shape, and location for different patients and make the manual detection time consuming and sensitive to the experience of the observer. Automatic segmentation of FCD lesions is challenging due to the difference in signal strength in images acquired with different machines, noise, and other kinds of distortions such as motion artifacts. Most of the methods proposed in the literature use conventional machine learning and image processing techniques in which their accuracy relies on the trained features. Hence, feature extraction should be done more precisely which requires human expertise. The ability to learn the appropriate features/representations from the training data without any human interventions makes the convolutional neural network (CNN) the suitable method for addressing these drawbacks. As far as we are aware, this work is the first one to use a CNN based model to solve the aforementioned problem using only MRI FLAIR images. We customized the popular U-Net architecture and trained the proposed model from scratch (using MRI images acquired with 1.5T and 3T scanners). FCD detection rate (recall) of the proposed model is 82.5 (33/40 patients detected correctly). © 2019Item An improved nonlocal maximum likelihood estimation method for denoising magnetic resonance images with spatially varying noise levels(Elsevier B.V., 2020) Sudeep, P.V.; Ponnusamy, P.; Kesavadas, C.; Rajan, J.Magnetic resonance images (MRI) reconstructed with parallel MRI (pMRI) techniques generally have spatially varying (non-stationary) noise levels. However, most of the existing MRI denoising methods rely on a stationary noise model and end with suboptimal results when applied to pMRI images. To address this problem, this paper proposes an improved nonlocal maximum likelihood (NLML) estimation method. In the proposed method, a noise map is computed with a robust noise estimator before the ML estimation of the underlying signal. Also, a similarity measure based on local frequency descriptors (LFD) is introduced to find the nonlocal samples for ML estimation. The experiments on simulated and real magnetic resonance (MR) data demonstrate that the proposed technique has superior filtering capabilities in terms of subjective and quantitative assessments when compared with other state-of-the-art methods. © 2018 Elsevier B.V.Item Multi-Res-Attention UNet: A CNN Model for the Segmentation of Focal Cortical Dysplasia Lesions from Magnetic Resonance Images(Institute of Electrical and Electronics Engineers Inc., 2021) Thomas, E.; Pawan, S.J.; Kumar, S.; Horo, A.; Niyas, S.; Vinayagamani, S.; Kesavadas, C.; Rajan, J.In this work, we have focused on the segmentation of Focal Cortical Dysplasia (FCD) regions from MRI images. FCD is a congenital malformation of brain development that is considered as the most common causative of intractable epilepsy in adults and children. To our knowledge, the latest work concerning the automatic segmentation of FCD was proposed using a fully convolutional neural network (FCN) model based on UNet. While there is no doubt that the model outperformed conventional image processing techniques by a considerable margin, it suffers from several pitfalls. First, it does not account for the large semantic gap of feature maps passed from the encoder to the decoder layer through the long skip connections. Second, it fails to leverage the salient features that represent complex FCD lesions and suppress most of the irrelevant features in the input sample. We propose Multi-Res-Attention UNet; a novel hybrid skip connection-based FCN architecture that addresses these drawbacks. Moreover, we have trained it from scratch for the detection of FCD from 3 T MRI 3D FLAIR images and conducted 5-fold cross-validation to evaluate the model. FCD detection rate (Recall) of 92% was achieved for patient wise analysis. © 2013 IEEE.Item Segmentation of focal cortical dysplasia lesions from magnetic resonance images using 3D convolutional neural networks(Elsevier Ltd, 2021) Niyas, S.; Chethana Vaisali, S.; Show, I.; Chandrika, T.G.; Vinayagamani, S.; Kesavadas, C.; Rajan, J.Computer-aided diagnosis using advanced Artific ial Intelligence (AI) techniques has become much popular over the last few years. This work automates the segmentation of Focal Cortical Dysplasia (FCD) lesions from three-dimensional (3D) Magnetic Resonance (MR) images. FCD is a type of neuronal malformation in the brain cortex and is the leading cause of intractable epilepsy, irrespective of gender or age differences. Since the neuron related abnormalities are usually resistant to drug therapy, surgical resection has been the main treatment approach for patients with intractable epilepsy. Automating the identification and segmentation of FCD is useful for neuroradiologists in pre-surgical evaluations. Convolutional Neural Networks (CNNs) have the ability to learn appropriate features from the training data without any human intervention. But, most of the state-of-the-art FCD segmentation approaches use two-dimensional (2D) CNN models despite the availability of 3D Magnetic resonance imaging (MRI) volumes, and hence fail to leverage the inter-slice information present in the MRI volumes. The major hurdles in considering a 3D CNN model are the need for a large 3D dataset, big memory, and high computation cost. A deep 3D CNN segmentation model, which can extract inter-slice information and overcomes the drawbacks of conventional 3D CNN methods to an extent, is proposed in this paper. The model uses a 3D version of U-Net with residual blocks that works on shallow depth 3D sub-volumes generated from MRI volumes. The proposed method shows superior performance over the state-of-the-art FCD segmentation methods in both qualitative and quantitative analysis. © 2021 Elsevier Ltd