Search

Disease prediction using computer-based methods is now an established area of research. The importance of technological intervention is necessary for the better management of disease, as well as to optimize use of limited resources. Various AI-based methods for disease prediction have been documented in the literature. Validated AI-based systems support diagnoses and decision making by doctors/medical practitioners. The resource-efficient dissemination of the symptoms identified and the diagnoses undertaken is the requirement of the present-day scenario to support paperless, yet seamless, information sharing. The representation of symptoms using grammar provides a novel way for the resource-efficient encoding of disease diagnoses. Initially, symptoms are represented as strings, and, in terms of grammar, this is called a sentence. Moreover, the conversion of the generated string containing the symptoms and the diagnostic outcome to a QR code post encryption makes it portable. The code can be stored in a mobile application, in a secure manner, and can be scanned wherever required, universally. The patient can carry the medical condition and the diagnosis in the form of the QR code for medical consultations. This research work presents a case study based on two diseases, influenza and coronavirus, to highlight the proposed methodology. Both diseases have some common and overlapping symptoms. The proposed system can be implemented for any kind of disease detection, including clinical and diagnostic imaging.

Image segmentation is an application area of computer vision and digital image processing that partitions a digital image into multiple image regions or segments. This process involves extracting a set of contours from the input digital image so that pixels belonging to a region share some common characteristics or computed properties, such as color, texture, or intensity. The application domain of image segmentation is widespread and includes video surveillance, object detection, traffic control system, and medical imaging. The application of image segmentation techniques in the field of medical imaging can be further subcategorized into virtual surgery simulation, diagnosis, a study of anatomical structures, measurement of tissue volumes, location of tumours, and other pathologies. In this study, we have proposed two new Convolutional Neural Network (CNN)-based models: (a) S-Net and (b) SA-Net (S-Net with attention mechanism) to perform image segmentation tasks in the field of medical imaging, especially to generate segmentation masks for brain tumours if present in brain Medical Resonance Imaging (MRI) scans. Both proposed models were developed by considering U-Net as the base architecture. The newly proposed models have leveraged the concept of ‘Merge Block’ to infuse both the local and global context and ‘Attention Block’ to focus on the region of interest having a specific object. Additionally, it uses techniques, such as data augmentation to utilize the available annotated samples more efficiently. The proposed models achieved a Dice Similarity Coefficient (DSC) measures of 0.78 and 0.81 for the High-Grade Glioma (HGG) and Low-Grade Glioma (LGG) datasets, respectively.

Introduction: Corona Virus Infectious Disease (COVID-19) is the infectious disease. The COVID-19 disease came to earth in early 2019. It is expanding exponentially throughout the world and affected an enormous number of human beings starting from the last month. The World Health Organization (WHO) on March 11, 2020 declared COVID-19 was characterized as “Pandemic”. This paper proposed approach for confirmation of COVID-19 cases after the diagnosis of doctors. The objective of this study uses machine learning method to evaluate how much predicted results are close to original data related to Confirmed-Negative-Released-Death cases of COVID-19. Materials and methods: For this purpose, a verification method is proposed in this paper that uses the concept of Deep-learning Neural Network. In this framework, Long shrt-term memory (LSTM) and Gated Recurrent Unit (GRU) are also assimilated finally for training the dataset. The prediction results are tally with the results predicted by clinical doctors. Results: The results are obtained from the proposed method with accuracy 87 % for the “confirmed Cases”, 67.8 % for “Negative Cases”, 62% for “Deceased Case” and 40.5 % for “Released Case”. Another important parameter i.e. RMSE shows 30.15% for Confirmed Case, 49.4 % for Negative Cases, 4.16 % for Deceased Case and 13.72 % for Released Case. Conclusions: The outbreak of Coronavirus has the nature of exponential growth and so it is difficult to control with limited clinical persons for handling a huge number of patients within a reasonable time. So it is necessary to build an automated model, based on machine learning approach, for corrective measure after the decision of clinical doctors.

Purpose: To evaluate the efficacy, safety, and steroid-sparing effect of topical cyclosporine A (Cs A) 0.05% in patients with moderate to severe steroid dependent vernal keratoconjunctivitis (VKC). Methods: A prospective, comparative, placebo controlled study was carried out on 68 VKC patients, with 34 patients treated with topical Cs A 0.05% and the remaining 34 with topical carboxymethyl cellulose 0.5% (placebo). Both groups also received topical loteprednol etabonate 0.5%. Symptom (itching, photophobia, tearing, and discharge) score, sign (tarsal and limbal papillae, corneal involvement, and conjunctival hyperemia) score, and drug score (steroid drop usage/day/eye) were recorded at baseline and each follow-up visit. The intraocular pressure (IOP) measurement and evaluation of any ocular side effects were carried out. Results: Significant reduction in symptom score and sign score was seen in both groups. Cs A group significantly showed more reduction in symptom (P < 0.0001 in all follow-up visits) and sign (P < 0.0001 in all follow-up visits) scores compared to the placebo group. At day 7, mean steroid usage reduced from 4 to 3.44 ± 0.5 and 3.79 ± 0.4 in Cs A and placebo groups, respectively (P < 0.0001). Steroid drops completely stopped in 21 patients at day 60 in the Cs A group compared to none in the placebo group. No significant rise in IOP or any side effects were noted in either group. Conclusion: Topical Cs A 0.05% is effective and safe in patients with moderate to severe VKC with good steroid-sparing effect.

Objective: Accurate identification of brain tumor and its heterogeneity is a critical task in planning for proper therapy. A reliable fully automatic detection and analysis method for the brain tumor is necessary for an efficient measurement of the tumors and its extent. This paper presents a computerized approach to brain tumor-edema detection and analysis from the MRI of brain sequences. Method: Computer-aided diagnosis systems are focused on several research activities and the ideas of the study of brain images with the diverse modality of heterogeneity by applying better image analysis algorithms. The proposed automated modern approach includes several stages of image segmentation, area and volume calculation, and its location findings using statistical and unsupervised clustering prediction method. Result: The outcome of the proposed computerized method is compared with reference images and gives very promising results. Performance of our proposed methodology is also assessed with the gold standard recent comparable method and our method gives better results in context to accuracy and error metrics. Conclusion: The proposed method is capable of improving the overall detection, segmentation, and quantification of a variety of tumors for different cases from multiple standard datasets. Keywords: Abnormality detection, Brain tumor, Segmentation, Heterogeneity, Magnetic resonance imaging, Accuracy estimation

Purpose: To study the demographic profile, severity and causes of visual impairment among registered patients in a tertiary care hospital in north Kolkata, eastern India, and to assess the magnitude of under-registration in that population. Methods: This is a retrospective analytical study. A review of all visually impaired patients registered at our tertiary care hospital during a ten-year period from January 2005 to December 2014, which is entitled for certification of people of north Kolkata, eastern India (with a population denominator of 1.1 million), was performed. Overall, 2472 eyes of 1236 patients were analyzed in terms of demographic characteristics, cause of visual impairment, and percentage of visual disability as per the guidelines established by the government of India. Results: Male patients (844, 68.28%; 95% confidence interval [CI], 65.69-70.87) registered more often than female patients (392; 31.72%, P = 0.0004). The registration rate for visual impairment was 11.24 per 100,000 per annum; this is not the true incidence rate, as both new patients and those visiting for renewal of certification were included in the study. Optic atrophy was the most common cause of visual impairment (384 eyes, 15.53%; 95% CI, 14.1-16.96). Conclusion: Commonest cause of visual impairment was optic atrophy followed by microphthalmos. Under-registration is a prevalent problem as the registration system is voluntary rather than mandatory, and female patients are more likely to be unregistered in this area.

Data is growing at an enormous rate in the present world. One of the finest and most popular technologies available for handling and processing that enormous amount of data is the Hadoop ecosystem. Enterprises are increasingly relying on Hadoop for storing their valuable data and processing it. However, Hadoop is still evolving. There is much vulnerability found in Hadoop, which can question the security of the sensitive information that enterprises are storing on it. In this paper, security issues associated with the framework have been identified. We have also tried to give a brief overview of the currently available solutions and what are their limitations. At the end a novel method is introduced, which can be used to eliminate the found vulnerabilities in the framework. In the modern era, information security has become a fundamental necessity for each and every individual. However, not everyone can afford the specialized distributions provided by different vendors to their Hadoop cluster. This paper presents a cost-effective technique that anyone can use with their Hadoop cluster to give it 3-D security.

The dataset developed consists of 108 blood drip stains developed with fresh porcine blood, blood admixed with different dosage of Warfarin and Heparin, respectively. For each particular blood type (i.e. fresh blood, blood admixed with Warfarin at different dosage and blood admixed with Heparin at varied dosage) stain patterns were created by passive dripping of blood from a 2.5 cm3 subcutaneous syringe with needle filled to capacity, at 30°, 60° and 90° angle of impact with corresponding fall height of 20, 40 and 60 cm respectively. In the other dataset of 162 datapoints, 81 regular drip stains were formed from blood that had dripped passively from a subcutaneous syringe without needle at the aforementioned angle of impact and fall height, while the other stains were formed as a result of dripping of blood from a subcutaneous syringe with needle. In order to compare stains formed, all stains were recorded on the same representative, non-porous, smooth target surface under similar physical conditions. The interpretations relevant to the dataset are available in the article titled ‘2D Source Area prediction based on physical characteristics of a regular, passive blood drip stain’ (Basu and Bandyopadhyay, 2016) [7]. An image pre-processing algorithm for extracting ROI has also been incorporated in this article. Keywords: Drip stain, Bloodstain Pattern Analysis, Source Dimension prediction

Objective: Accurate identification of brain tumors and their heterogeneity is a critical task in planning for proper therapy. A reliable fully automatic detection and analysis method for the brain tumor is necessary for an efficient measurement of the tumors and their extent. This paper presents a computerized approach to brain tumor-edema detection and analysis from the MRI of brain sequences. Method: Computer-aided diagnosis systems are focused on several research activities, and the ideas of the study of brain images with the diverse modality of heterogeneity by applying better image analysis algorithms. The proposed automated modern approach includes several stages of image segmentation, area and volume calculation, and locational findings using statistical and an unsupervised clustering prediction method. Result: The outcome of the proposed computerized method is compared with reference images and gives very promising results. Performance of our proposed methodology is also assessed with the gold standard recent comparable method, and our method gives better results in the context of accuracy and error metrics. Conclusion: The proposed method is capable of improving the overall detection, segmentation, and quantification of a variety of tumors for different cases from multiple standard datasets. Keywords: Abnormality detection, Brain tumor, Segmentation, Heterogeneity, Magnetic resonance imaging, Accuracy estimation

Background: Accurate evaluation of an individuals' veracity is a fundamental aspect of social functioning that allows individuals to act in adaptive ways. The domain of deception detection ability is still young, and many components in this field are yet to be touched which demands more research in this field. Aims: The present study aims at deciphering the structural composition of face during felt, posed, and deceived emotions in facial expression unique to Indian culture, using Facial Action Coding System (FACS). Quantitative analysis of Euclidean distance has been done to complement qualitative FACS analysis. Methods: In this study, thirty female, young adults with age range of 23–27 years were chosen randomly for portraying their (felt, posed, and deceived) facial expression. All facial expressions were captured through instruction, and videos were converted into static images. The static images were coded on the basis of FACS to decipher the felt, posed, and deceived expressions. Quantitative analysis of the data has been done using MATLAB to meet the objectives of the study and to complement the qualitative analysis. Results: Felt and posed emotions differ in terms of intensity of the expression and subjective experience. Posed emotional and deceived expressions differ in intent. Facial asymmetry is an important indicator for detecting deception.