Shiru Sharma

Segmentation and classification of medical images using texture-primitive features: Application of BAM-type artificial neural network

The objective of developing this software is to achieve auto-segmentation and tissue characterization. Therefore, the present algorithm has been designed and developed for analysis of medical images based on hybridization of syntactic and statistical approaches, using artificial neural network (ANN). This algorithm performs segmentation and classification as is done in human vision system, which recognizes objects; perceives depth; identifies different textures, curved surfaces, or a surface inclination by texture information and brightness. The analysis of medical image is directly based on four steps: 1) image filtering, 2) segmentation, 3) feature extraction, and 4) analysis of extracted features by pattern recognition system or classifier. In this paper, an attempt has been made to present an approach for soft tissue characterization utilizing texture-primitive features with ANN as segmentation and classifier tool. The present approach directly combines second, third, and fourth steps into one algorithm. This is a semisupervised approach in which supervision is involved only at the level of defining texture-primitive cell; afterwards, algorithm itself scans the whole image and performs the segmentation and classification in unsupervised mode. The algorithm was first tested on Markov textures, and the success rate achieved in classification was 100%; further, the algorithm was able to give results on the test images impregnated with distorted Markov texture cell. In addition to this, the output also indicated the level of distortion in distorted Markov texture cell as compared to standard Markov texture cell. Finally, algorithm was applied to selected medical images for segmentation and classification. Results were in agreement with those with manual segmentation and were clinically correlated.

Segmentation of medical images using Simulated Annealing Based Fuzzy C Means algorithm

Accurate segmentation is desirable for analysis and diagnosis of medical images. This study provides methodology for fully automated simulated annealing based fuzzy c-means algorithm, modelled as graph search method. The approach is unsupervised based on pixel clustering using textural features. The virtually training free algorithm needs initial temperature and cooling rate as input parameters. Experimentation on more than 180 MR and CT images for different parameter values, has suggested the best-suited values for accurate segmentation. An overall 97% correct segmentation has been achieved. The results, evaluated by radiologists, are of clinical importance for segmentation and classification of Region of Interest.

Simulated annealing-based particle swarm optimisation with adaptive jump strategy for modelling of dynamic cerebral pressure autoregulation mechanism

This paper proposes a new particle swarm optimisation (PSO) algorithm based on simulated annealing (SA) with adaptive jump strategy to alleviate some of the limitations of the standard PSO algorithm. In this algorithm, swarm particles jump into the space to find new solutions. The jump radius is selected adaptively based on the particle velocity and its distance from the global best position. The designed algorithm has been tested on benchmark optimisation functions and on known autoregressive exogenous (ARX) model design problem. The results are superior as compared to the existing PSO methods. Finally, the designed algorithm has been applied for the analysis of the dynamic cerebral autoregulation mechanism.

Five days daily sessions of noninvasive blood glucose level predictions based on amplitude modulated ultrasound and infrared technique over a healthy and diabetic subject.

A noninvasive method based blood glucose predicting device with medically acceptable readings could transform the diabetic management protocols. Noninvasive blood glucose monitoring devices will increase patient compliances along with reduction of medical burden and its related expenses. Here, we had utilized the indigenously developed amplitude modulated ultrasound and infrared technique based noninvasive glucometer for this purpose. A lab based study had been performed to measure the performance of the modulated ultrasound and infrared technique based noninvasive glucometer. A total of 02 subjects (01 healthy normal subject and 01 diabetic subject) were engaged in this study. The working accuracy of noninvasive glucometer readings had been compared with the invasive glucometer readings and plotted over the Clarke Error Grid for its critical analysis. The experimental result depicts a good relationship exists between the predicted (noninvasive) and reference (invasive) blood glucose levels. The Clarke Error Grid Analysis depicts that results of the predictions as well as the reference measurement values occupies the medically significant A and B domains. This experimental result directs towards the potential applications of amplitude modulated ultrasound and infrared technique for continuous noninvasive blood glucose level predictions.

Multithreaded Approach for Registration of Medical Images using Mutual Information in Multicore Environment and its Applications in Medical Imaging

Image Registration plays very crucial role in case of medical imaging to register different modalities of images like CT (Computed Tomography) and PET (Positron Emission Tomography) registration. CT is essential for structural information of anatomic and PET (Positron Emission Tomography) is for functional information. Basically it is the procedure of transforming dissimilar sets of data into one coordinate system. These sets of data can be acquired from multiple image modalities, different viewpoints, similar or dissimilar sensors. MI based image registration has been found to be reasonably useful methods of image registration. However, it is found to be quite computationally intensive and time consuming process for enormous size images and for different data sets of images. It involves steps for computation of joint histogram, marginal entropies, calculation and probability distribution. Main motive of this paper is to provide an intelligent method for image registration based on Mutual Information using multi core environment with maintaining the synchronization between different activated cores and processors. Proposed Method has been able to execute with different number of threads to achieve all the remuneration of the processors and gives significant speedup working with verity of images like gray scale, RGB and Dicom images with different size. Finally the designed algorithm has been used to register medical images of different modalities.

A non-linear complex diffusion based filter adapted to Rayleigh’s speckle noise for de-speckling ultrasound images

In this paper, an edge and structure preserving non-linear complex diffusion based filter adapted to Rayleigh’s speckle noise for speckle reduction from 2D ultrasound images is proposed and implemented in MATLAB 7.0. The initial condition of the proposed filter is the speckle noised ultrasound image and the speckle-reduced image is obtained after certain iterations of the filter till its convergence. For digital implementations, the proposed scheme has been discretised using the finite difference scheme. The performance of the proposed complex diffusion-based filter has been evaluated both qualitatively and quantitatively. A comparative study of the proposed scheme with other standard speckle reduction schemes such as the homomorphic Wiener filter, the Lee filter, the Frost filter, the Kuan filter and the Speckle Reducing Anisotropic Diffusion (SRAD) filter for several ultrasound images with varying amounts of speckle noise variance is also presented. The obtained results show that the proposed non-linear complex diffusion-based scheme performs better than all other schemes in consideration and is also well capable of preserving edges and fine structures from ultrasound images during speckle reduction.

A fourth-order PDE-based non-linear filter for speckle reduction from Optical Coherence Tomography images

In this paper, a fourth-order PDE-based non-linear filter for speckle reduction from Optical Coherence Tomography (OCT) images is proposed in a variational framework. The speckle noise pattern in OCT images is distributed according to Rayleigh’s Probability Distribution Function (pdf). The initial condition of the proposed PDE-based filter is the speckle noised OCT images and the output is the better quality speckle reduced image. For digital implementations, the proposed scheme is discretised using the finite difference scheme. The performance comparison of the proposed scheme with other standard speckle reduction techniques such as the Lee filter, the Frost filter, the Kuan filter, the SRAD filter, the homomorphic Wiener filter and homomorphic versions of the anisotropic diffusion based filter, the non-linear complex diffusion-based filter and the fourth order PDE-based filter is also presented. The obtained results justify the applicability of the proposed scheme.

Measurement of glucose by using modulating ultrasound with optical technique in normal and diabetic human blood serum

This paper describes the effect of various glucose concentrations in normal and diabetic human blood serums mixed with intralipid phantoms. Amplitude modulated ultrasound and infrared techniques were utilized here. Modulated ultrasound causes signature specific molecular vibrations through the area of its propagation. The infrared light of specific wavelength is used here to detect the target glucose molecules. The infrared photo sensor detects the resultant signal and undergoes series of signal processing steps to yield the output value. The trend obtained from the result indicates that human diabetic blood serum occupies highest peak values in Fast Fourier Transform domain as compared to the normal human blood serum peak values. The proposed method provides a new dimension for noninvasive blood glucose measurement techniques. Noninvasive method for blood glucose detection is the need of hour. Tight control regimen with noninvasive glucometer will change the way of living of the diabetes affected population.

Non-rigid registration of multimodal images (Ultrasound and CT) of Liver using gradient orientation information

Image registration is considered as a highly challenging task which is used in various medical applications such as diagnosis and image guided interventions. Registration is performed with medical images captured via different modalities and labeled as moving and fixed images. The transformation of the moving image is achieved by minimizing an objective function through updating the parameters of transformation. The existing techniques have some drawbacks in terms of speed, performance level and accuracy. Considering the limits, a new algorithm for non-rigid registration is proposed in this paper which is executed using the Ultrasound (US) and Computed Tomography (CT) images of Liver. The algorithm includes segmentation of liver surface, selection of best matched slice using similarity measure, calculation of objective function and estimation of transformation. The proposed method is applied to three clinical datasets and quantitative evaluations are conducted. Visual examinations and experimental results verifies a lower level of registration error and a higher level of accuracy which makes the algorithm acceptable for clinical applications.

In-vitro Measurement of Glucose Concentration in Human Blood Plasma Mixed Intralipid Phantom Samples by Using Modulated Ultrasound and Infrared Light

Non-invasive blood glucose measurement is one of most innovative domain in Biomedical Engineering. Multiple methodologies have-been introduced over last few decades to fulfil the clinical requirement for non-invasive glucose measurement in human beings, however, without a successful breakthrough. This research article uses modulated ultrasound with infrared light-based technique to study glucose-induced variations in human blood plasma mixed Intralipid TM phantom samples using infrared light of 940 nm and 40 kHz central frequency based ultrasonic transmitter unit. The test uses blood samples of 30 study subjects during oral glucose tolerance test and fasting, postprandial and random stages based blood glucose tests respectively. The result as obtained from oral glucose tolerance tests and fasting, postprandial and random stages blood Original Research Article