An accomplished academician with over 20 years of experience in teaching, research, and academic
leadership. I have held significant roles such as Head of Department (8 years), Dean of Academics (4
years), and Assistant Director (5 years) in AICTE and UGC-approved institutions. Under my leadership,
institutions have achieved ISO, NBA, and NAAC accreditation, and I have facilitated numerous MOUs with
international universities and industries, as well as established new projects
Deep Learning Models Performance Analysis For Cardiovascular Disease Using An Ecg Based Dataset Saroj Kumari, Meena Chaudhary, Raghav Mehra International Journal of Drug Delivery Technology, 2026 Because CVD is among the leading causes of death worldwide, early risk prediction is particularly important. This work employs a mixed dataset and differentiates between three types of variables, clinical/lifestyle, and ECG-based ones, aiming at analyzing four DLMs namely- MLP, LSTM, CNN, and ViT. The case for the relevance to PTB Using the ECG portion of the PTB-XL and combining it with UCI Heart Disease features; this resulted in 22 multimodal predictors down the first level. We evaluated the model performance using accuracy, precision, recall, F1-score, and AUC. The results indicate that all architectures perform well (Accuracy ≥ 0.97, AUC ≥ 0.99). LSTM obtained the optimal overall balance (Accuracy = 0.99, Recall = 0.98, F1 = 0.98), though CNN had slightly lower recall, (Recall = 0.97), and ViT reached perfect accuracy (1.00) with slightly lower recall (0.90). SVM joint analysis of three modalities as an example, the hybrid method has shown the feasibility of deep learning in early CVD diagnosis and decision support by means of boosting robustness and clinical relevance as compared to single- modality studies.
A Machine Learning Approach to Nadi Pariksha: Detecting Dosha Imbalances Raghav Mehra, Nandani Agarwal, Justina J 2025 IEEE 7th International Conference on Computing Communication and Automation Iccca 2025, 2025 This research introduces PulseVision: AI-based Nadi Pariksha Health Diagnosis, an AI-based system for automated and upgraded traditional Ayurvedic diagnostic process of Nadi Pariksha. The analysis uses a panel-based dataset containing pulse-derived diagnostic attributes that have been preprocessed by the Random Forest classifier to reliably identify the major dosha, i.e., Vata, Pitta, Kapha, or mixture. The 5-fold cross-validation method guarantees the robustness of the model and leads to both high accuracy and reliability of the classification.The model’s performance is evaluated using parameters such as accuracy, precision, recall and f1 score. The accuracy of the random forest was found to be 98.67% with 98.71% precision, 98.67% recall and 98.63 f1 score. The framework is also equipped with interpretability mechanisms, and these are implemented through data visualization methods, specifically, feature distribution histograms, correlation heatmaps, and class distribution plots.The paper illustrates the possibility of combining AI with mainstream medical expertise, hence suggesting a scalable and objective diagnostic tool. Results indicate that PulseVision may be used by Ayurvedic physicians and persons to support data-driven healthcare decisions in Ayurvedic integrative medicine with the potential impact of computational intelligence in the treatment of personalized Ayurvedic medicine.
Towards Next-Generation Sustainable Cancer Diagnosis: A Hybrid Transformer-Enhanced Texture-Attention Framework with Multi-Objective Optimization Kanishk Agrawal, Javed Wasim, Raghav Mehra 2025 IEEE International Conference on Modern Electronics Devices and Intelligent Communication Systems Medcom 2025, 2025 Sustainable cancer diagnosis with histopathological images has not yet been achieved as high computational cost, class imbalance, and lack of model interpretability are present. The currently available CNN and transformer-based models are computationally expensive, resulting into high performance, but limit their clinical applicability. The proposed paper suggests using Hybrid Transformer-Enhanced Texture Attention (TETA) framework, which combines the handcrafted Histogram Gray-Level Co-occurrence Matrix (HGLCM) texture features with transformer-based global attention. Enzyme Action Optimization (EAO) and Multi-Objective Optimization (MOO) are used in the optimization of the model to compromise between the diagnostic accuracy, energy consumption, and the speed of inferences. The novel TextureAware Attention Module is a biologically significant region detector and enhances the interpretability and diagnostic reliability. The CHAVI histopathology dataset was evaluated experimentally with 98.8% accuracy and 98.5% F1-score and used 60 J of energy and 18 ms of inference time - better than CNN, ResNet-50, ViT, and hybrid baselines. The suggested framework identifies a scalable, explainable, and energy-efficient cancer detection method based on AI, which is in line with the vision of Green AI of sustainable healthcare.
The Moderator AI: Classifying Harmful Language in Real-Time Raghav Mehra, Kartik Rupal, Rishi Raj 2025 IEEE 7th International Conference on Computing Communication and Automation Iccca 2025, 2025 The proliferation of toxic content on online platforms requires the development of robust, scalable, and automated solutions for content moderation. This paper presents a comprehensive, full-stack web application designed for real-time hate speech detection and classification. The system leverages a long-short-term memory (LSTM) based Recurrent Neural Network (RNN) to categorize user-provided text into one of three distinct classes: "Hate Speech," "Offensive Language," or "Neither." The methodology encompasses a rigorous preprocessing pipeline applied to a public dataset, including advanced techniques to address severe class imbalance through the synthetic minority oversampling technique (SMOTE). The trained deep learning model is encapsulated and served via a Flask back-end API, which communicates with a lightweight and responsive HTML/JavaScript front-end. The resulting application serves as a functional proof-of-concept that demonstrates a practical, end-to-end approach to utilizing deep learning for the critical task of automated content moderation, addressing both technical implementation and ethical considerations.
AI Personalized Mantra Recommender Meenakshy Shiju, B.J Gouri, Raghav Mehra 2nd International Conference on Intelligent Systems for Cybersecurity Iscs 2025, 2025 Increasing levels of stress and digital fatigue underscore the need for personalized wellness interventions. Though mantra meditation has been shown to have psychological and physiological benefits, wellness platforms currently in existence are generally generic. This paper introduces an AI-powered conversational agent that suggests mantras tailored to users’ personality traits and emotional states. The platform combines the Big Five Inventory for personality assessment with real-time emotion detection using natural language processing, voice, and optional facial recognition. A hybrid recommendation system, integrating rule-based reasoning and adaptive machine learning, translates user states to mantras that induce calmness, concentration, or energy. Deployed in Python (Flask) with text and voice-based interactivity, the prototype leverages a curated Vedic mantra database annotated with desired outcomes. Early results indicate context-sensitive and successful recommendations that increase user engagement and emotional resonance, supporting the potential of AI-personalized spiritual health systems.
Providing Security Solutions to IoT Devices Using Micro Version of Security Protocols over Wireless Networks C Satisha, Raghav Mehra, M. Giri 2025 4th International Conference on Power Control and Computing Technologies Icpc2t 2025, 2025 In this research paper proposed micro version of Security protocol for IoT environment, it consist of three phases, in first phase all IoT devices must be registered with server to get user login credentials, in second phase the IoT devices which are really wants to access with server must be authenticated by server, and in third phase data is transferred through secure IoT channel by providing integrity, security and confidentiality. Data security is provided with twenty rounds of sub keys. In security context, to hide the probabilistic relation between transmitted data to cipher text we can use diffusion matrix. Diffusion matrix is used to protect from attacks and in our study we calculated diffusion matrix by varying p and q values. We observed that change of one bit in a single block may change many bits in cipher text.
An Energy-Aware Hybrid Adaptive Preprocessing Framework for Attention-Guided Histopathological Cancer Detection Kanishk Agrawal, Javed Wasim, Raghav Mehra 2025 International Conference on Emerging Technologies and Innovation for Sustainability Emergin 2025, 2025 The demand for histopathological cancer diagnosis to be reproducible and high-quality spurs the need for preprocessing frameworks that increase image interpretability and optimize computational cost. However, current algorithmic approaches are mainly limited by noise sensitivity, staining variability, and energy load. To that end, we present the Energy Aware Hybrid Adaptive Pre-processing Framework for cancer detection in histopathological images (HAPA-D) by using hybrid adaptive contrast, attention guided denoising, and staining normalization, and optimizing for energy efficiency while maintaining quality detection of cancer. As an early step, the input images for HAPA-D methods were collected from the CHAVI dataset. The input histopathological images are pre-processed as per the proposed HAPA-D pipeline to improve interpretability and computational costs through incorporating multiple preprocessing and optimization module into a framework. Adaptive Contrast Enhancement (ACE) enhances local tissue contrast, with Attention-Guided Denoising (AGD) selectively removing noise and retaining significant diagnostic regions. A Stain Normalization Module (SNM) addresses inter-slide discoloration by making sure datasets are as similar a color as possible. Energy-Aware Feature Optimization (EAFO) helps improve energy efficiency by removing redundant features and retaining clinically necessary diagnostic information. The features optimized using EAFO are then classified using the Efficient Focal Attention-Based Extreme Learning Classifier (EFA-ELC), which offers multi-class cancer diagnosis with accuracy and reliability. All techniques are implemented in a Python framework and are compared with the many existing methods based on their accuracy (99.4%) and energy consumption (45J). In conclusion, EFA-ELC provides accurate, reliable, and computationally sustainable multi-class cancer detection from histopathological images.
Advanced Ensemble Learning Approaches Based Energy Consumption Dhruv Aggarwal, Vijay Mohan Shrimal, Raghav Meha, Manoj Wadhwa 2nd International Conference on Research Methodologies in Knowledge Management Artificial Intelligence and Telecommunication Engineering Rmkmate 2025, 2025
Enhancing Waste Separation and Management Through IoT System Parveen Badoni, Ranjan Walia, Raghav Mehra Proceedings 2024 1st International Conference on Innovative Sustainable Technologies for Energy Mechatronics and Smart Systems Istems 2024, 2024
Wearable IoT Technology: Unveiling the Smart Hat Parveen Badoni, Ranjan Walia, Raghav Mehra Proceedings 2024 1st International Conference on Innovative Sustainable Technologies for Energy Mechatronics and Smart Systems Istems 2024, 2024
Enhancing Education with Augmented Reality: A Prototype-Based Approach Girish Paliwal, Kanta Prasad Sharma, Raghav Mehra, Vijay Mohan Shrimal, Manoj Kumar Pandey, Harsh Vijay Proceedings of International Conference on Emerging Technologies and Innovation for Sustainability Emergin 2024, 2024
FarmSaarthi: A Vernacular-Enabled Digital Agricultural Ecosystem with Multimodal AI for Precision Farming LGBECSEACUIAJBECSEACUIMKDRMBECSEABECSEACUCII com jetir 1 , 2026 2026
A blockchain-based self-sovereign identity in adaptive access control framework and differential privacy for real-time privacy-preserving EHR management SS Palkar, R Mehra, L Hadimani OPSEARCH, 1-38 , 2026 2026
Real Time Fire and Smoke Detection Using AI and Computer Vision S Bhardwaj, H Kaur, M Jethasri, K Verma, R Mehra 2025 IEEE 5th International Conference on ICT in Business Industry … , 2025 2025
An Optimized & Data-Driven Approach for Real-Time Slot Allocation and User-Friendly Access in Smart Parking System T Nagpal, S Kumar, VM Shrimal, G Paliwal, KP Sharma, R Mehra 2025 IEEE 5th International Conference on ICT in Business Industry … , 2025 2025
Optimizing Cloud Migration: A Case Study Approach to Application Modernization T Nagpal, S Kumar, VM Shrimal, G Paliwal, KP Sharma, R Mehra 2025 IEEE 5th International Conference on ICT in Business Industry … , 2025 2025
Towards Next-Generation Sustainable Cancer Diagnosis: A Hybrid Transformer-Enhanced Texture-Attention Framework with Multi-Objective Optimization K Agrawal, J Wasim, R Mehra 2025 Modern Electronics Devices and Intelligent Communication Systems … , 2025 2025
An Energy-Aware Hybrid Adaptive Preprocessing Framework for Attention-Guided Histopathological Cancer Detection K Agrawal, J Wasim, R Mehra 2025 International Conference on Emerging Technologies and Innovation for … , 2025 2025
Towards Sustainable Cancer Diagnosis: An Efficient Texture-Attention Framework with Enzyme-Inspired Optimization K Agrawal, J Wasim, R Mehra 2025 IEEE 7th International Conference on Computing, Communication and … , 2025 2025
The Moderator AI: Classifying Harmful Language in Real-Time R Mehra, K Rupal, R Raj 2025 IEEE 7th International Conference on Computing, Communication and … , 2025 2025
Evaluating Robustness of Neural Text Detectors in Generative AI Detection MK Verma, R Mehra, K Soni, M Kumar 2025 IEEE 7th International Conference on Computing, Communication and … , 2025 2025
Interpretability in Cardiovascular Diagnostics: Classical Machine Learning vs. Deep Learning S Kumari, M Chaudhary, R Mehra 2025 IEEE 7th International Conference on Computing, Communication and … , 2025 2025
A Machine Learning Approach to Nadi Pariksha: Detecting Dosha Imbalances R Mehra, N Agarwal 2025 IEEE 7th International Conference on Computing, Communication and … , 2025 2025
Multimodal Emotion Recognition Using Acoustic Features and Deep Language Models R Raghav, A Tomar, T Kaur, A Chauhan, R Mehra 2025 7th International Conference on Artificial Intelligence and Speech … , 2025 2025
Real-Time Patient Monitoring using AIML V Dhull, R Mehra Available at SSRN 5818162 , 2025 2025
Artificial Intelligence and Machine Learning Approaches for Healthcare Fraud Detection: A Review, Case Study, and Framework G Arora, R Mehra, S Kaswan, D Upadhaya, K Soni 2025 2nd Global AI Summit-International Conference on Artificial … , 2025 2025
AI Personalized Mantra Recommender M Shiju, BJ Gouri, R Mehra 2025 2nd International Conference on Intelligent Systems for Cybersecurity … , 2025 2025
AI-based Multi model Misinformation Detection using NLP and CNN Models J Justina, N Agarwal, H Bhardwaj, R Mehra 2025 2nd International Conference on Intelligent Systems for Cybersecurity … , 2025 2025
Adaptive Smart Valve Controlling System for Indian Farms based on Soil, Crop and Climatic Data M Wadhwa, K Thirupathi, P Badoni, B Kaur, N Kumar, R Mehra 2025 Third International Conference on Emerging Applications of Material … , 2025 2025
AI-powered adaptive traffic light system R Mehra, S Shri, H Agarwal, S Dutta Engineering Science and Technology: Innovations for the Future, 192-198 , 2025 2025
Advances in Chromatographic Techniques for Drug Purity Determination R Mehra, MS Kapoor Journal of Pharmaceutical Analysis and Drug Research 7 (2) , 2025 2025
MOST CITED SCHOLAR PUBLICATIONS
Detection of Various Security Attacks on IoT Devices Using Multi-Layer Neural Network Model Over Sensor Networks C Satisha, R Mehra, M Giri 2023 International Conference on Computational Intelligence, Networks and … , 2023 2023.0 Citations: 17
Enhancing Waste Separation and Management Through IoT System P Badoni, R Walia, R Mehra IEEE - 2024 1st International Conference on Innovative Sustainable … , 2024 2024.0 Citations: 12
Wearable IoT Technology: Unveiling the Smart Hat P Badoni, R Walia, R Mehra ieee - 2024 1st International Conference on Innovative Sustainable … , 2024 2024.0 Citations: 10
Column based NoSQL database, scope and future R Mehra, N Lodhi, R Babu International Journal of Research and Analytical Reviews 2 (4), 105-113 , 2015 2015.0 Citations: 9
Novel approach to automated test data generation for AOP A Misra, R Mehra, M Singh, J Kumar, S Mishra International Journal of Information and Education Technology 1 (2), 179 , 2011 2011.0 Citations: 9
Towards EOS-04 ARD normalized radar backscatter (NRB) product R Mehra, W Akram, KM Agrawal, VM Ramanujam 2023 IEEE India Geoscience and Remote Sensing Symposium (InGARSS), 1-4 , 2023 2023.0 Citations: 6
Design of neural network based approaches for land usage land cover classification VK MR, R Mehra, HPR Kunadharaju 2024 Third International Conference on Electrical, Electronics, Information … , 2024 2024.0 Citations: 4
MIDAS: a software for radiometric and polarimetric processing of EOS-04 SAR data. D Putrevu, T Maganti, T Chakraborty, M Kumar, C Sanid, P Arora, ... Current Science (00113891) 126 (9) , 2024 2024.0 Citations: 4
Supervised SVM classification of rainfall datasets KHP Raju, N Sandhya, R Mehra Indian Journal of Science and Technology 10 (15), 1-6 , 2017 2017.0 Citations: 4
Speckle Reduction in SAR Images using CNN V Santhi, D Mohandass, J Jayanthi, P Arulmozhivarman, R Mehra 2021 3rd International Conference on Signal Processing and Communication … , 2021 2021.0 Citations: 3
Intelligent method for cost estimation during software maintenance D Puja, R Mehra, BD Mazumdar International Journal Of Engineering And Computer Science , 2017 2017.0 Citations: 3
Redesign of e-learning development in India J Wasim, R Mehra, IA Khan International Journal of Research in Engineering and Science (IJRES) 4 (10 … , 2016 2016.0 Citations: 3
Detection of Brain Tumor Using Unsupervised Enhanced K-Means, PCA and Supervised SVM Machine Learning Algorithms HPR Kunadharaju, N Sandhya, R Mehra International Research Journal on Advanced Science Hub 2, 62-67 , 0 Citations: 3
Radar Imaging With India’s Earth Observation Satellite-04 : Earth resource monitoring using C-band synthetic aperture radar system CVN Rao, D Putrevu, VM Ramanujam, BK Bhattacharya, C Patnaik, ... IEEE Geoscience and Remote Sensing Magazine 13 (2), 152-181 , 2025 2025.0 Citations: 2
Enhancing Education with Augmented Reality: A Prototype-Based Approach G Paliwal, KP Sharma, R Mehra, VM Shrimal, MK Pandey, H Vijay 2024 International Conference on Emerging Technologies and Innovation for … , 2024 2024.0 Citations: 2
Assessment of EOS-04 (RISAT-1A) data calibration. S Sharma, S Tripathi, B Sowkhya, P Arora, S Tyagi, C Sanid, R Agrawal, ... Current Science (00113891) 126 (9) , 2024 2024.0 Citations: 2
WITHDRAWN: An efficient detection of micro aneurysms from fundus images HPR Kunadharaju, N Sandhya, R Mehra Materials Today: Proceedings , 2021 2021.0 Citations: 1
Study of factors impacting the successful establishment of enterprise architecture implementation capability in Indian public sector N Sharma, R Mehra Asian Journal of Research in Business Economics and Management 11 (6), 1-11 , 2021 2021.0 Citations: 1
Role of Indian remote sensing imaging satellites for the Antarctic monitoring and mapping: a case study around Indian Antarctic research stations P Jayaprasad, R Mehra, S Chawla, DR Rajak, SR Oza Land Surface and Cryosphere Remote Sensing III 9877, 151-160 , 2016 2016.0 Citations: 1
Multi Sensor Image Matching using Super Symmetric Affinity Tensors based HyperGraph Matching HPR Kunadharaju, N Sandhya, R Mehra Citations: 1