K. Prabhakar Nayak

Automated classification of glaucoma stages using higher order cumulant features

Glaucoma is a group of disease often causing visual impairment without any prior symptoms. It is usually caused due to high intra ocular pressure (IOP) which can result in blindness by damaging the optic nerve. Hence, diagnosing the glaucoma in the early stage can prevent the vision loss. This paper proposes a novel automated glaucoma diagnosis system using higher order spectra (HOS) cumulants extracted from Radon transform (RT) applied on digital fundus images. In this work, the images are classified into three classes: normal, mild glaucoma and moderate/severe glaucoma. The 3rd order HOS cumulant features are subjected to linear discriminant analysis (LDA) to reduce the number of features and then these clinically significant linear discriminant (LD) features are fed to the support vector machine (SVM) and Naive Bayesian (NB) classifiers for automated diagnosis. This work is validated using 272 fundus images with 100 normal, 72 mild glaucoma and 100 moderate/severe glaucoma images using ten-fold cross validation method. The proposed system can detect the early glaucoma stage with an average accuracy of 84.72%, and the three classes with an average accuracy of 92.65%, sensitivity of 100% and specificity of 92% using NB classifier. This automated system can be used during the mass screening of glaucoma.

Decision support system for the glaucoma using Gabor transformation

Abstract Increase in intraocular pressure (IOP) is one of the causes of glaucoma which can lead to blindness if not detected and treated at an early stage. Glaucoma symptoms are not always obvious; hence patients seek treatment only when the condition progressed significantly. Early detection and treatment will decrease the chances of vision loss due to glaucoma. This paper proposes a novel automated glaucoma diagnosis method using various features extracted from Gabor transform applied on digital fundus images. In this work, we have used 510 images to classify into normal and glaucoma classes. Various features namely mean, variance, skewness, kurtosis, energy, and Shannon, Renyi, and Kapoor entropies are extracted from the Gabor transform coefficients. These extracted features are subjected to principal component analysis (PCA) to reduce the dimensionality of the features. Then these features are ranked using various ranking methods namely: Bhattacharyya space algorithm, t-test, Wilcoxon test, Receiver Operating Curve (ROC), and entropy. In this work, t-test ranking method yielded the highest performance with an average accuracy of 93.10%, sensitivity of 89.75% and specificity of 96.20% using 23 features with Support Vector Machine (SVM) classifier. Also, we have proposed a Glaucoma Risk Index (GRI) developed using principal components to classify the two classes using just one number.

Quantification of spinal deformities using combined SCP and geometric 3D reconstruction

Abstract Three dimensional reconstruction is essential in accurate diagnosis of numerous spinal deformities which are 3D in nature. The stereo-radiographic reconstruction involving bi-planar X-rays is one of the most commonly used methods. Algorithms with stereo-corresponding point (SCP) or non-stereo-corresponding point (NSCP) can be used to achieve this 3D reconstruction. But, the NSCP method is slower and needs manual identification of many anatomical landmarks. Hence it suffers from observer variability and has restricted usage in normal clinical setup. Thus, a hybrid method is proposed in which the SCP reconstructed model is refined using the geometric features from the X-rays to achieve the accuracy closest to NSCP method. The SCP model is constructed using the X-rays on a calibration bench from the scoliotic subject. From these X-rays the vertebral features are extracted automatically. The SCP model structure is refined using the geometric transformations according to the extracted features. The 3D model thus formed is called combined SCP and geometric (CSCPG) reconstruction. By considering the NSCP model as a reference, both qualitative and quantitative approaches are followed to validate the proposed model. The CSCPG method has lesser observer variability as it needs only six anatomical landmarks per vertebra. Further, it is faster and the reconstruction error is within the acceptable limits. To quantify the deformities like axial vertebral rotation and spinal curvature novel methods have been proposed. The axial vertebral rotation is measured using simple vertebra vector parametric computations. It needs identification of only two landmarks per vertebra for angle measurement. The apical vertebra gives the plane of maximum curvature. The actual spinal curvature has to be computed on this plane. A semi-automatic method is proposed to compute this curvature using a new projection technique. The deformity quantification methods are validated using manual measurements as well as the results from standard approaches. Hence, a fast, simple and economic 3D diagnostic method is developed for quantification of the spinal deformities.

A Trusted Ubiquitous Healthcare Monitoring System for Hospital Environment

WirelessBodySensorNetworkwithwearableandimplantablebodysensorshavebeengrabbinglotof interestsamongtheresearchersandhealthcareserviceproviders.Thesesensorsforwardphysiological datatothepersonnelatthehospital,doctororcaretakeranytime,anywhere;hencethenameofthe networkisUbiquitoushealthmonitoringsystem.ThetechnologyhasbroughtInternetofThingsinto thissystemmakingittogetconnectedtothecloudbasedinternet.Thishasmadetheretrievalof informationtotheexpertandthusimprovingthehappinessofelderlypeopleandpatientssuffering fromchronicdiseases.Thispaperfocusesoncreatinganandroidbasedapplicationformonitoring patientsinhospitalenvironment.Thenecessityofsharinghospitaldatatotheexpertsaroundtheglobe hasbroughtthenecessityoftrustinHealthcaresystems.ThedatasharingintheIOTenvironmentis secured.Theenvironmentistestedinreal-timecloudenvironment.Theproposedandroidapplication servestobebetterarchitectureforhospitalmonitoring. KEywoRdS Internet of Things, Trust, Wireless Body Sensor Network (WBSN), ZigBee

Effect of automatic landmark identification on CSCPG reconstruction

Scoliosis is a 3D deformation of the human spine. It is evaluated using 3D stereo-radiographic reconstruction from the biplanar x-rays. The combined stereo-corresponding point and geometric (CSCPG) reconstruction algorithm is used for this purpose. The CSCPG reconstruction requires six stereo-corresponding landmarks per vertebra to be identified on the biplanar x-rays. Currently, these landmarks are semi automatically identified. In this paper, the landmark identification procedure is automated and the effect of automated procedures on the accuracy of 3D reconstruction is analyzed. The statistical significance test is performed to compare the accuracies of these reconstructions. The benefits of automated procedure are twofold. It is able to give better reconstruction accuracy and at the same time it also reduces observer variability.

Improved Performance Evaluation of Frequency Hopped Spread Spectrum using QAM/FSK Modulation Techniques

theoretical foundation has been provided that expands upon QAM and FSK modulation in the application of Frequency Hopped spread spectrum systems. This paper mainly focuses on the Bit Error Estimation in Frequency Hop Spread Spectrum System using Quadrature Amplitude modulation and Frequency shift keying which are used in defense applications. The main aim here is to calculate the bit error performance of a frequency hopping spread spectrum model in the presence of AWGN channel. This paper provides a systematic approach for evaluating the performance of FHSS operating with coherent M-ary FSK demodulation. There have been investigations into the frequency hop spread spectrum systems employing different modulation schemes to decrease the bit error ratios. There has been much work done on computing BER of FHSS systems with error control coding using industry standard convolutional coding. KeywordsHopped spread spectrum,FSK,QAM.

Combined SCP and Geometric Reconstruction of Spine from Biplanar Radiographs

Three-dimensional reconstruction of the spine is necessary in proper diagnosis of various spinal deformities. This is normally achieved using stereo-radiographic techniques involving biplanar (frontal and lateral) radiographs. Either stereo-corresponding point (SCP) algorithm or non-stereo-corresponding point (NSCP) algorithm is used for this purpose. The NSCP technique suffers from observer variability. Moreover, it is time consuming. Hence, it has restricted usage in clinical environment. Here, a hybrid approach is proposed in which a 3D spine model is reconstructed by applying geometric features to the SCP reconstructed model. The vertebral orientation features are automatically extracted from the calibrated radiographs. The 3D model thus produced is successfully validated. The proposed method has lesser observer variability due to the limited number of anatomical landmarks. Also, the reconstruction errors are within the acceptable limits available in the literature. Thus, the proposed technique can be used in clinical practices for the diagnosis of spinal deformities.