Padmavathi Kora

In this paper we proposed a prostate segmentation and also tumour detection using deep neural networks. The cutting-edge deep learning techniques are useful compared to the challenges of machine learning based feature extraction techniques. Here we proposed a strategy that contains an FCN model that incorporates data from several MRI images, allowing for faster convergence and more accurate segmentation. T1 and DWI volumes may be used together to delineate the prostate boundary, according to this study. Second, we investigated whether this method might be utilized to provide voxel-level prostate tumor forecasts. The cascaded learning method and performed tests to demonstrate its effectiveness.

In this paper, firefly optimization is applied to balance robustness and imperceptibility in image watermarking. First of all, the original image is partitioned into 4 × 4 blocks, and later Redundant Discrete Wavelet Transform (RDWT) and Schur applied for obtaining transform coefficients. The highly correlated pixels in the first column are collected from each 4 × 4 block and are amended by blindly concealing the watermark using quantization index modulation. These quantized values are amended with the randomized matrix and quantization factor randomly selected from the firefly algorithm to obtain the best performance in terms of robustness without compromising the quality of the image. Experimental results show that the proposed scheme maintains improved imperceptibility, and the watermark can still extract from a seriously distorted image.

This paper suggests a new approach for improving the accuracy of retinal OCT frames. This is referred to as the QIROCT (Quality Enhancement of Retinal Optical Coherence Tomography) procedure. Retinal optical coherence tomography (OCT) image is a layered structure. A mixture model, combination of multiple distributions, is used to represent retinal OCT image. A Gaussian-Mixture-Model (GMM), mixture of Gaussians, is proposed to represent the retinal OCT image as retina is a layered structure. Expectation maximization (EM) is an algorithm that fits the Gaussian mixture model (GMM). Gaussian components are obtained using the Expectation Maximization (EM) algorithm to match the Gaussian Mixture Model (GMM) to the retinal OCT results. Adaptive Gamma Correction with Weighting Distribution (AGCWD) is used to improve Gaussian components of the retinal OCT image. To understand the superiority of QIROCT method for OCT image processing, 30 healthy retinal OCT images are tested. The QIROCT approach is opposed to the contrast limited adaptive histogram equalization (CLAHE) method, and the difference between the two methods is visually and numerically illustrated. And segmentation is done for retinal OCT image using the QIROCT method and results are shown.

Direction-of-arrival (DOA) evaluation involves the procedure by which multiple electromagnetic (EM) waves from the outputs of various receiving antennas that construct a sensor array retrieve the direction information. There are many algorithms for the direction of arrival estimation in the literature like MUSIC, ESPRIT, first-order forward prediction, Capon, etc. These algorithms have heavy calculation operations. This situation could cause lags in the algorithm’s response time and may pose an essential disadvantage in real time applications. To overcome this problem, artificial neural network (ANN) can be used. The training stage of an ANN needs significant time and sources, but after training, the estimation of Direction of arrival using ANN is very fast. An ANN method for the direction of arrival estimation in uniform linear array antennas has been projected in this project. In training, the whole pseudo spectrum is scanned by 10-degree steps. A uniform linear array with 2,3,4, and 5 isotropic antenna elements and one source signal is considered in the simulations. Tests of the trained ANN have been done for various arrival angles, and satisfactory results have been obtained

In medical image anatomy, brain tumor extraction occupies an important role. This is used for medical diagnosis and thus leading towards the treatment of disease. The objective of the paper is to present various MR brain image segmentation techniques ranging from elementary threshold methods to complicated methods such as deformable methods, hybrid methods. The motivation of this study is early detection and identification of the brain tumor. In this paper, the main focus is on gliomas tumor which is the most recurrent type of malignant brain tumors. This work presents various research studies in brain tumor segmentation along with deep learning techniques such as Convolutional Neural Network.