Conference Papers
Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506
Browse
5 results
Search Results
Item A parallel segmentation of brain tumor from magnetic resonance images(2012) Dessai, V.S.; Arakeri, M.P.; Guddeti, G.Medical image segmentation is nowadays at the core of medical image analysis and supports computer-aided diagnosis, surgical planning, intra-operative guidance or postoperative assessment. Large amounts of research efforts have been made in developing effective brain MR (magnetic resonance) image tumor segmentation methods in the past years. However algorithms proposed so far are time consuming because it involves lot of mathematical computations. Also serial segmentation of multiple MRI slices (usually required for 3D visualization) takes exponential time. This results in need for improvement in performance as far as the time complexity is concerned. This paper proposes a methodology that incorporates the K-means clustering and morphological operation for parallel segmentation of multiple MRI slices corresponding to single patient. Segmentation of multiple MRI slices for tumor extraction plays major role in 3D (Three Dimensional) visualization and serves as an input for the same. The proposed framework follows SIMD (Single Instruction Multiple Data) model and since the segmentation of individual slice is independent of each other and can be performed in parallel and multithreading definitely speeds up the entire process. Also the framework does not involve any kind of inter-process communication thus the time is saved here as well. © 2012 IEEE.Item A new nonlocal maximum likelihood estimation method for denoising magnetic resonance images(2013) Rajan, J.; den Dekker, A.J.; Juntu, J.; Sijbers, J.Denoising of Magnetic Resonance images is important for proper visual analysis, accurate parameter estimation, and for further preprocessing of these images. Maximum Likelihood (ML) estimation methods were proved to be very effective in denoising Magnetic Resonance (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 and, as a result, optimal results cannot be achieved because of 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 way using the Kolmogorov-Smirnov (KS) similarity test. The method has been tested both on simulated and real data, showing its effectiveness. © Springer-Verlag 2013.Item A semi-automatic method for carotid artery wall segmentation in MR images(Institute of Electrical and Electronics Engineers Inc., 2017) Kumar, P.K.; Kesavadas, C.; Rajan, J.The quantification of carotid artery stenosis via imaging techniques guides the physicians to take a decision regarding surgical interventions. The measurement of wall thickness from magnetic resonance (MR) images is a promising approach to measure the degree of carotid stenosis. Manual tracing of the carotid vessel walls is time consuming and is sensitive to observer variability. Further, the existing segmentation techniques are limited by the poor contrast and presence of noise in MR images. The objective this paper is to present a novel segmentation strategy for carotid lumen and outer wall from MR images. The segmentation has been carried out in two stages which starts with a user assisted region of interest selection. In the first stage, an active contour based global segmentation has been applied to classify the lumen region. In the second stage, morphological gradient of the region of interest has been computed. This is followed by particle swarm optimization based localized segmentation to separate the wall region. The results demonstrate excellent correspondence between the automatic and manual tracings for lumen and outer walls of the carotid artery. © 2016 IEEE.Item 3D AttU-NET for Brain Tumor Segmentation with a Novel Loss Function(Institute of Electrical and Electronics Engineers Inc., 2023) Roy, R.; Annappa, B.; Dodia, S.In the United States of America (USA), every year 150,000 patients are registered with a secondary brain tumor that is not generated in the brain. This necessitates the need for early brain tumor detection, which in turn will help patients to live longer. For clinical evaluation and treatment, precise segmentation of brain tumors in MRI images is required. This process can be aided by machine learning and efficient image processing, but manual imaging can be time-consuming. In this study, we aim to develop an 3D automated segmentation algorithm with a novel loss function. A 3D attention UNET CNN model was trained using the novel loss function, which was calculated by taking the weighted average of dice loss and focal loss to overcome the class imbalance. Results show the enhancement in the segmentation performance of attention UNET model with an average increase of 5% in the Dice coefficient for all three classes. However, the model's performance was not as strong for enhanced and core tumors. Further research may be needed to optimize performance in these areas. . © 2023 IEEE.Item 3D-Conditional Generative Adversarial Networks for Brain Tumour Segmentation(Institute of Electrical and Electronics Engineers Inc., 2024) Magar, P.K.; Naik, D.Gliomas, the most common primary brain tumor, exhibit significant heterogeneity in their prognosis, aggressiveness, and histological composition, encompassing areas such as necrotic cores, enhancing and non-enhancing tumor cores, and peritumoral edema. While multimodal MRI is invaluable for brain tumor detection, precise tumor segmentation remains challenging. To overcome this, a novel 3D volume-to-volume GAN, termed the 3D-Conditional Generative Adversarial Network (3D-cGAN), was developed for the brain tumor segmentation, leveraging data from the 2020 BraTS Challenge. This model utilizes multi channel 3D MRI images to accurately segment core, whole, and enhancing tumor regions. Employing a batch size of 4 and an alpha value of 2, the model demonstrates remarkable accuracy on the BraTS 2020 dataset, achieving a Dice score of 0.8286 and IoU score of 0.7111. © 2024 IEEE.