CNN: A review of models, application of IVD segmentation Silvoster M. Leena, R. Mathusoothana, S. Kumar Computational Analysis and Deep Learning for Medical Care Principles Methods and Applications, 2021 The widespread publicity of Convolutional Neural Network (CNN) in various domains such as image classification, object recognition, and scene classification has revolutionized the research in machine learning, especially in medical images. Magnetic Resonance Images (MRIs) are suffering from severe noise, weak edges, low contrast, and intensity inhomogeneity. Recent advances in deep learning with fewer connections and parameters made their training easier. This chapter presents an in-depth review of the various deep architectures as well as its application for segmenting the Intervertebral disc (IVD) from the 3D spine image and its evaluation. The first section deals with the study of various traditional architectures of deep CNN such as LeNet, AlexNet, ZFNet, GoogleNet, VGGNet, ResNet, Inception model, ResNeXt, SENet, MobileNet V1/V2, and DenseNet. It also deals with the study of the parameters and components associated with the models in detail. The second section discusses the application of these models to segment IVD from the spine image. Finally, theoretically performance and experimental results of the state-of-art of the literature shows that 2.5D multi-scale FCN performs the best with the Dice Similarity Index (DSC) of 90.64%.
Efficient segmentation of lumbar intervertebral disc from MR images Leena Silvoster M, Retnaswami Mathusoothana S. Kumar Iet Image Processing, 2020 : Segmentation of spine Magnetic Resonance Images (MRIs) has become an indispensable process in the diagnosis of lumbar disc degeneration, causing low back pain. Over the last decade of years, computer-directed diagnosis of disease, as well as computer-guided spine surgery, is based on the two-dimensional (2D) analysis of mid-sagittal slice of MRI. This work proposes an automatic strategy to extract the 3D segmentation of the normal disc as well as degenerated lumbar intervertebral discs (IVDs) from T2-weighted Turbo Spin Echo MRI of the spine using Connected Component (CC) analysis algorithm and statistical shape analysis. The challenges faced by the IVD segmentation includes (i) partial volume effects (ii) intensity inhomogeneity (iii) grey level overlap of different soft tissues. The proposed method first pre-processes the dataset and enables it for the application of the CC algorithm. The CC (subsets of pixels of the disc) of the spine MRI is extracted and apply statistical shape analysis for the refinement of the segmentation results to detect IVDs. Experimental results of the proposed method show a robust segmentation, accomplishing the dice similarity index of 92.4% and thus achieving a low error rate. Other performance measures such as Precision, Accuracy, JaccardIdx, JaccardDist, Global Consistency Error, Variation of Information, etc were also evaluated. The algorithm is evaluated quantitatively using adequate experiments on a dataset of 15 MRI scans, of different scenarios such as healthy and degenerate disc and this proposed method is verified as a promising accurate method for the automatic segmentation of IVD
Automatic segmentation, size determination and classification of IVD Journal of Advanced Research in Dynamical and Control Systems, 2019
Efficient identification of desiccated lumbar IVD from MRI Iioab Journal, 2016
Graph colouring based image similarity M Leena Silvoster 2012 International Conference on Power Signals Controls and Computation Epscicon 2012, 2012 Region based segmentation approaches have been applied to several different problems in computer vision, like image classification, image registration, image retrieval, objects recognition, video indexing etc. The attraction of this algorithm is independent of the user defined seed point. The generality of region based segmentation of real-world problems in the light of regions has made it popular for large range of images. However in the prior methods, the computational effort increases as the complexity of the problem increases. This paper is based on an efficient graph theoretical approach. A maximal planar graph is constructed from the region segmented image. Primary advantage of this approach is one can naively analyze the resulting graph and determine the number and position of a particular object in the region of interest. The images are matched (with N points) in O(N2 log N) and segment a given image of N pixels in O(N log N).
Enhanced CNN based electron microscopy image segmentation M. Leena Silvoster, V. K. Govindan Cybernetics and Information Technologies, 2012 Detecting the neural processes like axons and dendrites needs high quality SEM images. This paper proposes an approach using perceptual grouping via a graph cut and its combinations with Convolutional Neural Network (CNN) to achieve improved segmentation of SEM images. Experimental results demonstrate improved computational efficiency with linear running time.
