Papers by Prakash Choudhary
American Journal of Ophthalmology, 2008
Purpose-To compare the prevalence of different types and densities of age-related cataract in sub... more Purpose-To compare the prevalence of different types and densities of age-related cataract in subjects with high myopia and emmetropia in the Indian urban clinic-based population. Design-Prospective observational clinic-based case-control study. Methods-An observational case-control study of 800 healthy eyes was undertaken at Iladevi Cataract & IOL Research Center, Ahmedabad, India. Subjects with high myopia (axial length [AXL] ≥ 26.0 mm) (n = 400 eyes) and those with emmetropia (AXL 21.0-23.99 mm) were examined (n = 400 eyes). The type of cataract was categorized as: nuclear, cortical, and posterior subcapsular cataract (PSC). Nuclear density was measured based on the Emery and Little classification. Results-In all the age groups (40+ years old), nuclear cataract was more often encountered in subjects with high myopia (Odds ratio: 3.8, 95% CI 2.9-5.2, P <0.001); PSC and mixed cataracts were frequently encountered in subjects with emmetropia (P <0.001). Prevalence of nuclear cataract was higher in subjects with high myopia with increasing AXL when compared with other types of cataract (P <0.001). In all the age groups, the nuclear density was significantly higher than grade 3 in subjects with high myopia when compared to those with emmetropia (P < 0.001 in <70 years of age, P = 0.003 in >70 years of age). Conclusion-Nuclear cataract was strongly associated with high axial myopia. The density of the cataract was higher in the high myopia group. No association was observed between PSC or cortical cataract and high axial myopia.
Biomedical Signal Processing and Control
Communications in Computer and Information Science
Stroke is a major life-threatening disease mostly occurs to a person of age 65 years and above bu... more Stroke is a major life-threatening disease mostly occurs to a person of age 65 years and above but nowadays also happen in younger age due to unhealthy diet. If we can predict a stroke in its early stage, then it can be prevented. In this paper, we evaluate five different machine learning techniques to predict stroke on Cardiovascular Health Study (CHS) dataset. We use Decision Tree (DT) with the C4.5 algorithm for feature selection, Principal Component Analysis (PCA) is used for dimension reduction and, Artificial Neural Network (ANN) and Support Vector Machine (SVM) are used for classification. The predictive methods discussed in this paper are tested on different data samples based on different machine learning techniques. From the different methods applied, the composite method of DT, PCA and ANN gives the optimal result.
2021 International Conference on Computational Performance Evaluation (ComPE)
COVID-19 was previously identified as 2019-nCoV, however it was reclassified as severe acute resp... more COVID-19 was previously identified as 2019-nCoV, however it was reclassified as severe acute respiratory syndrome coronavirus 2 by the International Committee on Taxonomy of Viruses (ICTV) (SARS-CoV-2). It was first discovered in Wuhan, China’s Hubei Province, and has since spread all over the world. The scientific community is working to develop COVID-19 detection technologies that are both quick and accurate. Chest x-ray imaging can aid in the early diagnosis of COVID-19 patients. In COVID-19 individuals, chest x-rays can indicate a variety of lung abnormalities, including lung consolidation, ground-glass opacity, and others. The COVID-19 biomarkers, however, must be identified by qualified and experienced radiologists. Each report must be inspected by the radiologist, which is a time-consuming procedure. The medical infrastructure is currently overburdened due to the huge volume of patients. In this study, we propose automatic COVID-19 identification in chest x-rays using a deep learning technique. COVID-19, pneumonia, and healthy x-rays are included in the dataset for the studies. The proposed model had an average accuracy and sensitivity of 97 percent. The obtained findings demonstrate that the model can compete with existing state-of-the-art models.
2017 13th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS), 2017
In this paper, we present compilation of Hindi handwritten text image Corpus and its linguistics ... more In this paper, we present compilation of Hindi handwritten text image Corpus and its linguistics perspective in the field of OCR and information retrieval from handwritten document. Devnagari script is little bit complicated to enter a single character; it requires a combination of multiples, due to use of modifier. A mixed approach is proposed and demonstrated for Hindi Corpus for OCR and Demographic data collection. Demographic part of database could be used to train a system to fetch the data automatically, which will be helpful to simplify existing manual data-processing task involved in the field of data collection such as input forms like AADHAR, driving license, Railway Reservation etc. This would increase the participation of Hindi language community in understanding and taking benefit of the government schemes. To make availability and applicability of database in a vast area of corpus linguistics, we propose a methodology for data collection, mark-up, digital transcription, and XML metadata information for benchmarking and ZipF' s law to analyze the distribution and behavior of words in the corpus.
Applied Intelligence, 2020
Recognition of handwritten characters in two Indic scripts Bangla and Meitei Mayek is one of the ... more Recognition of handwritten characters in two Indic scripts Bangla and Meitei Mayek is one of the challenging responsibilities due to intricate patterns and scarcity of standard datasets. Convolutional Neural Network (CNN) is one of the stablest well-known techniques for classifying objects in distinctive specialties as it has an extraordinary capability of discovering complex patterns. In this paper, we hook a different layout and obtain a unique CNN architecture from scratch, which has manifold advantages over classical machine learning (ML) approaches, and it has a unique ability to consolidate feature extraction and classification altogether. Further, we stretch our work to uncover the mathematical rationale for using non-linearity in the deep learning (DL) model. Our proposed CNN architecture consists of four layers, including convolutional layer (CL), nonlinear activation layer (AL), pooling layer (PL), and fully connected layer (FCL), which are used in the existing two accessible Bangla datasets named cMATERdb and ISI Bangla datasets. The identical model also validates on proposed Manipuri Character dataset, called "Mayek27". Moreover, we perform an in-depth comparison with different batch sizes and optimization techniques over all the datasets for understanding their functionality. We conceive a novel benchmark performance that has delivered state-of-the-art decisions on two regional handwritten character identifications.
The Visual Computer, 2022
Chest radiography (X-ray) is the most common diagnostic method for pulmonary disorders. A trained... more Chest radiography (X-ray) is the most common diagnostic method for pulmonary disorders. A trained radiologist is required for interpreting the radiographs. But sometimes, even experienced radiologists can misinterpret the findings. This leads to the need for computer-aided detection diagnosis. For decades, researchers were automatically detecting pulmonary disorders using the traditional computer vision (CV) methods. Now the availability of large annotated datasets and computing hardware has made it possible for deep learning to dominate the area. It is now the modus operandi for feature extraction, segmentation, detection, and classification tasks in medical imaging analysis. This paper focuses on the research conducted using chest Xrays for the lung segmentation and detection/classification of pulmonary disorders on publicly available datasets. The studies performed using the Generative Adversarial Network (GAN) models for segmentation and classification on chest X-rays are also included in this study. GAN has gained the interest of the CV community as it can help with medical data scarcity. In this study, we have also included the research conducted before the popularity of deep learning models to have a clear picture of the field. Many surveys have been published, but none of them is dedicated to chest X-rays. This study will help the readers to know about the existing techniques, approaches, and their significance.
Advances in Intelligent Systems and Computing, 2019
Handwritten character recognition is an exciting and challenging topic in the field of pattern re... more Handwritten character recognition is an exciting and challenging topic in the field of pattern recognition because of massive variation in writing style and similar looking characters. Combining two different scripts boost the challenge to another level as each language has a unique peculiarity. The choice of distinguishing feature enhances the accuracy and efficiency of a recognition system. In this paper, we present spatial features based recognition of handwritten Manipuri (Meitei Mayek) and English alphabets. Background directional distribution, projection histogram, and uniform local binary pattern features have been used for extracting distinct feature for recognition by KNN classifier. The highest accuracy achieved in the proposed methodology is 92.40%.
Sensors for Health Monitoring, 2019
Abstract MRI images are a very rich source of texture. Analysis of texture is crucial for the aut... more Abstract MRI images are a very rich source of texture. Analysis of texture is crucial for the automation of tumor detection from MRI images. By analyzing what information a specific texture analysis method exploits to construct a feature vector, this chapter presents various texture analysis methods, understanding the ingredients of these methods and recognizing their merits and limitations. For a comparison between the state-of-the-art texture analysis methods, this chapter presents a common experimental protocol. A comprehensive treatise of four texture analysis methods with three classifiers is implemented and evaluated. Thus, this study aims to provide panoramas of texture analysis methods for the detection of a brain tumor from MRI images and reviews related works of the past.
Advances in Intelligent Systems and Computing, 2018
A brain tumor is a fatal disease which takes thousands of lives each year. Thus, timely and accur... more A brain tumor is a fatal disease which takes thousands of lives each year. Thus, timely and accurate treatment planning is a critical stage to improve the quality of life. MRI is a very novel method of diagnosis of the brain which shows a fine level of details of the brain tumor. For the treatment of brain tumor, accurate segmentation of the tumor part is highly desirable. The manual tumor segmentation is a challenging and time-consuming process. Thus, automatic tumor segmentation can play a very important role in the process to speed up the treatment process. There exist different types of MRI sequences, each with its own merits and showing varying levels of information. We experimented with T2-weighted (T2), T1 with enhancing contrast (T1c), and FLAIR MRI images of the BRATS 2013 dataset and try to show that one particular MRI sequence is very useful for segmentation of one particular class of tumor than other. We have used thresholding and K-means algorithms along with a set of p...
International Ayurvedic Medical Journal, 2021
Ayurveda has described three basic physiological constituents of the human body, viz., Dosha, Dha... more Ayurveda has described three basic physiological constituents of the human body, viz., Dosha, Dhatu and Mala. In Ayurvedic Science, the one who has balanced Doshas, balanced Agni, properly formed Dhatus, proper elimina- tion of Malas, well-functioning of bodily processes and whose mind, soul, senses are full of bliss is called a healthy person1. So, the formation of Dhatu is also a good indicator of good health there are seven Dhatus ex- plained in Samhitas, those are Rasa, Rakta, Mamsa, Meda, Ashti, Majja and Shukra among all Dhatus, Shukra is considered as the sara of all other Dhatus2. Shukra Dhatu is one of the seven Dhatus in the body and Shukra is white, pure, excellent Dhatu, which is considered as best among all seven Dhatus. According to many Acharyas of Ayurveda, Garbhotpadana (reproduction) is the chief function of Shukra Dhatu, and the important fact is Shukra Dhatu also shows its effect all over the body in the form of Shukradhatusarata because Shukradhatuis located in ...
Handwritten character recognition is an essential field in pattern recognition. Its popularity is... more Handwritten character recognition is an essential field in pattern recognition. Its popularity is increasing with the potential to thrive in various applications such as banking, postal automation, form filling, etc. However, developing such a system is a challenging task with the diverse writing style of the same character, and present of visually similar characteristics. In this paper, a recognition system is proposed using a deep neural network. The performance of the network is investigated on a self-collected handwritten dataset of Manipuri script contributed by 90 different people of varying age and education. A total of 4900 sample images is considered for the experiment and recorded a recognition rate of 98.86%.
International Research Journal of Ayurveda & Yoga, 2021
In Ayurvedic works of art, Ahara (food) is referenced as one among the three Upasthambas (Submain... more In Ayurvedic works of art, Ahara (food) is referenced as one among the three Upasthambas (Submainstays of the body) which upholds the three principles Sthambas (Pillars) of the body. Ahara is viewed as indispensable for the human body as it gives the fundamental supplements, which are extremely crucial for complete the essential exercises of absorption and digestion. Viruddha Ahara is an interesting idea portrayed in Ayurveda. Ayurveda plainly characterizes that specific eating routine and its blends, which intrude on the digestion of tissue, which hinders the course of development of tissue and which have the contrary property to the tissue are called as Viruddha Anna or incongruent eating regimen. Diet ought to be healthy just as per Desh, Kala, Prakriti, and Vayah. Diet assumes a significant part in our life. Ayurveda has depicted Aahara exhaustively in their different Granthas. The food which isn't right in blend, which has gone through wrong handling, which is burned-through in mistaken portion, which is burned-through in the erroneous season of day and the wrong season, can prompt Viruddha Ahara. Food taken in legitimate technique feeds the individual genuinely and intellectually both and it is the food through which individual achieves positive wellbeing and development of the body. Food taken in ill-advised (Unbalanced) strategies can cause different sorts of sicknesses. Thusly Ayurveda has given sharp consideration to idea of healthy ahara and unwholesome ahara. Correspondingly admission of contrariness food is a lot of expansions in present time.
2017 8th Annual Industrial Automation and Electromechanical Engineering Conference (IEMECON), 2017
A stroke occurs when the blood supply to a person's brain is interrupted or reduced. The stro... more A stroke occurs when the blood supply to a person's brain is interrupted or reduced. The stroke deprives person's brain of oxygen and nutrients, which can cause brain cells to die. Numerous works have been carried out for predicting various diseases by comparing the performance of predictive data mining technologies. In this work, we compare different methods with our approach for stroke prediction on the Cardiovascular Health Study (CHS) dataset. Here, decision tree algorithm is used for feature selection process, principle component analysis algorithm is used for reducing the dimension and adopted back propagation neural network classification algorithm, to construct a classification model. After analyzing and comparing classification efficiencies with different methods and variation models accuracy, our work has the optimum predictive model for the stroke disease with 97.7% accuracy.
Journal of Ambient Intelligence and Humanized Computing, 2021
Zero-shot learning (ZSL) is a learning paradigm that tries to develop a recognition model to reco... more Zero-shot learning (ZSL) is a learning paradigm that tries to develop a recognition model to recognize mutually exclusive training and testing classes. To recognize mutually exclusive classes, some kind of correlation between training and testing classes are required. This paper proposed an inductive solution of the ZSL problem in two stages: (1) a supervised multiclass classifier is trained on the training set and further asked to classify the testing images to its nearest training class. (2) A mapping function, which maps training class to testing class is used to obtain the final class for each testing image. The correlation between seen classes and unseen classes are obtained using the mapping function. We have proposed a graphical mapping function based on a fully connected bipartite graph for mapping between training and testing classes. Each edge of the bipartite graph is assigned a weight calculated by exploiting the semantic space. The proposed model is evaluated over the three well-known ZSL datasets: AWA2, CUB and aPY and obtained 66.59%, 48.95%, and 32.91% mean accuracy respectively. The obtained f1 score of the proposed method is 0.675, 0.565 and 0.492 on AWA2, CUB and aPY dataset respectively.
Journal of Advances in Information Technology, 2018
This paper presents a new approach to improve the performance of palm vein-based identification s... more This paper presents a new approach to improve the performance of palm vein-based identification systems. Here palm vein is considered as a piece of texture and apply texture-based feature extraction technique. The palm vein from the database is used to identify the Region of Interest of that particular palm vein image. The extracted Region of Interest is than applied Gabor Filter for local features extraction. Extracted local features are then applied sequential feature reduction to reduce the features which are more unique. Reduced features vectors are then process using artificial neural network environment. Index Terms-palm vein, Gabor filter, ANN, Gaussian filter, ROI Rabikumar Meitram received his M.Tech degree from National Institute of Technology Manipur, India in 2016. His research area includes pattern recognition, feature extraction, artificial intelligence, deep learning, image annotation and retrieval, image processing, bio-medical image processing, and image classification.
Advances in Intelligent Systems and Computing, 2018
The normality & abnormality of the heart is normally monitored by ECG. Several algorithms are... more The normality & abnormality of the heart is normally monitored by ECG. Several algorithms are proposed to classify ECG signals. In this paper, discrete wavelet transform is used for extracting some statistical features and Multilayer perceptron (MLP) neural network with Back-propagation performs the classification task. Two types of ECG signals (normal and abnormal) can be detected in this work from each database. The records from MIT-BIH Arrhythmias and Apnea ECG database from physionet have been used for training and testing our neural network based classifier. 90% healthy and 100% abnormal are detected in MIT-BIH Arrhythmias database with the overall accuracy of 94.44%. In Apnea-ECG database, 96% normal and 95.6% abnormal ECG signals are detected and achieves 95.7% classification rate.
International Journal on Natural Language Computing, 2018
Recognition of Manipuri Script called Meitei Mayek is still in the infant stage due to its comple... more Recognition of Manipuri Script called Meitei Mayek is still in the infant stage due to its complex structure. In this paper, an attempt has been made to develop an offline Meitei Mayek handwritten character recognition model by exploiting the texture feature, Local Binary Pattern (LBP). The system has been developed and evaluated on a large dataset consisting of 3,780 characters which are collected from different people of varying age group. The highest recognition rate achieved by the proposed method is 93.33% using Support Vector Machine (SVM). So, the contribution of this paper is bi-fold: firstly, a collection of a large handwritten corpus of Meitei Mayek Script and secondly developing character recognition model on the collected dataset.
SAE Technical Paper Series, 2017
Uploads
Papers by Prakash Choudhary