Sudipta Roy

Segmentation of Acute Brain Stroke from MRI of Brain Image Using Power Law Transformation with Accuracy Estimation

Segmentation of acute brain stroke and its position is very important task in medical community. Accurate segmentation of brain’s abnormal region by computer aided design (CAD) system is very difficult and challenging task due to its irregular shape, size, high degree of intensity and textural similarity between normal areas and abnormal regions areas. We developed a new method using power law transformation which gives very fine result visually and from quantifiable point of view. Our methods gives very accurate segmented tumor output with very low error rate and very high accuracy.

Contour Extraction and Segmentation of Cerebral Hemorrhage from MRI of Brain by Gamma Transformation Approach

Computer-aided diagnosis (CAD) systems have been the focus of several research endeavors and it based on the idea of processing and analyzing images of different hemorrhage of the brain for a quick and accurate diagnosis. We use a gamma transformation approach with a preprocessing step to segment and detect whether a brain hemorrhage exists or not in a MRI scans of the brain with the type and position of the hemorrhage. The implemented system consists of several stages that include artefact and skull elimination as an image preprocessing, image segmentation, and location identification. We compare the results of the conducted experiments with reference image which are very promising visually as well as mathematically.

Modified technique of insertion methods for data hiding using DNA sequences

Cryptographic applications require several biological techniques and hence they have become more popular recently. In one of the most interesting techniques data is hidden in Deoxyribo Nuclic Acid (DNA). In this paper we have proposed a Data Hiding Insertion Method based upon DNA sequence. In this method we hide information data into DNA sequence randomly using certain techniques. In this method we use several procedures as: random key generation, selection of the succeeding prime number of key value, cumulative XOR operation of key value, selection of look up table index mapping.

Abnormal regions detection and quantification with accuracy estimation from MRI of brain

Accurate segmentation of brain tumor is very difficult and challenging task due to irregular shape, size, high degree of intensity and textural similarity between normal areas and abnormal regions areas. An intensity based new methodology to segment and detect is proposed which gives more accurate segmentation and detection of abnormal regions from MRI of brain. The accuracy of segmentation methods determine by quantitative measurement from the segmented abnormal regions. Here location of the abnormal regions is clearly identified using distance calculation of several positions which is very helpful for diagnosis. From visually and quantifiable point of view our methods gives very accurate segmented tumor output with very low error rate and very high accuracy.

Segmentation of the Human Corpus Callosum Variability from T1 Weighted MRI of Brain

Corpus Callosum is an important part of the brain which works as major neural pathway that connects homologous cortical areas of the two cerebral hemispheres. The size of Corpus Callosum is affected by age, sex, neurodegenerative diseases and various lateralized behaviour in people. Here T1 weighted Magnetic Resonance Imaging (MRI) of brain, usually the sagittal sections is taken which is then followed by the automated segmentation of the MRI slide. This segmentation has an important application in neurology as the shape as the thickness, size and orientation of Corpus Callosum depends on the various characteristics of the person. Lobar connectivity based percolations of the corpus callosum can be computed by our proposed method which is very accurate segmentation.

A useful approach towards 3D representation of brain abnormality from its 2D MRI slides with a volumetric exclamation

In the three-dimensional (3D) construction of brain tumour using several slides of magnetic resonance imaging (MRI) has always been a keen interest for diagnosis and for research purpose. In this paper we propose a new approach for three-dimensional construction and its volume calculation from a series of two dimensional (2D) MRI images. Each of the abnormality detected MRI image are successively pushed into a stack to construct a 3-dimensional cube inside which it contains the 3-dimentional constructed brain abnormality. The volume of abnormality is calculated from area of each MRI slides with their inter slice distance. This computer aided diagnosis (CAD) system tool helps the neurosurgeon to take decision during their surgical planning.