ClinPatch-AD: Token-Level Fusion of 3D MRI Patch Transformers and Clinical Scores for Subject-Wise Alzheimer's Disease Screening Harsh Vardhan Bansal, Tajinder Kumar, Purushottam Sharma, Xiaochun Cheng IEEE Access, 2026 Alzheimer’s disease (AD) is a leading cause of dementia, and early subject-level screening is important for timely clinical support. However, automated MRI-based AD diagnosis remains difficult because many deep learning studies rely on 2D slices or non-subject-wise splits, which can overestimate performance, and clinical variables are often fused only at the final stage. This study aims to develop a more reliable multimodal framework that uses full 3D MRI volumes and clinical information under strict subject-wise evaluation.We propose ClinPatch-AD, a 3D patch-based Transformer that converts each MRI scan into volumetric patch tokens and embeds age, sex, and MMSE as a dedicated clinical token. The clinical token is inserted into the same Transformer sequence as the MRI tokens, allowing token-level fusion through self-attention. An auxiliary MMSE regression head is also used during training to guide the shared representation with cognitive severity information. Experiments on the OASIS-1 cross-sectional cohort with a stratified subject-wise split achieved an AUC of 0.951, accuracy of 0.898, precision of 0.857, recall of 0.818, and F1-score of 0.840 on the held-out test set. Ablation results further showed that both the clinical token and the auxiliary MMSE objective improved performance over the MRI-only Transformer baseline. These findings show that token-level multimodal fusion improves realistic patient-level AD screening and supports the use of clinical context during MRI-based diagnosis.
Attention-Guided Lightweight CNN-Transformer Fusion for Real-Time Traffic Sign Recognition in Adverse Environments: HACTNet Mandeep Singh Devgan, Gurvinder Singh, Purushottam Sharma, Tajinder Kumar, Xiaochun Cheng, Deepak Ahlawat Iet Intelligent Transport Systems, 2026 Autonomous driving is also impossible without traffic sign recognition (TSR; also known as traffic sign‐on‐road), which limits its reliability to domain changes, unfavourable weather, obstruction and hardware capacity. This paper proposes HACTNet, a low‐complexity CNN‐Transformer hybrid model that pushes the state‐of‐art in TSR by making a noteworthy set of contributions including (i) efficient convaps to model parts of the image, (ii) transformer encoder to capture the global context and (iii) an attention‐based fusion block to dynamically combine the two complementary sets of features. This synergy facilitates strong recognition in presence of blur and occlusion and in varying illumination. In addition to accuracy, HACTNet achieves high robustness (52.8%) against strong PGD adversarial attacks (8/255), but is still efficient (7.9 M parameters and 22.1 FPS) on the NVIDIA Jetson Nano. Moreover, the comparative analysis between the hybrid models (EATFormer, local‐ViT) and HACTNet proves that HACTNet has a better accuracy‐efficiency ratio. The extraordinary capability to counteract adverse weather conditions, fog, night, rain, snow etc., which is proven by the extensive testing of the real‐world ACDC adverse conditions data set, supports the viability of the proposed solutions in the real world. It is plug and play modularity with on‐going learning via elastic weight consolidation (3.3% less forgetting) and unsupervised domain adaptation via MMD loss (5.3% better on TT100K with no labels). Moreover, INT8 quantization with quantization‐aware training (QAT) incurs little accuracy loss (less than 0.5 percent) and much lower energy (0.27 J/sample) usage, which forms an edge deployment preparedness. Additionally, when adjusting to new traffic signs over time, the model shows compatibility with continuous learning, achieving a low forgetting rate (3.3%), highlighting its practical viability for long‐term autonomous deployment. Overall, HACTNet produces a versatile and expandable solution for next‐generation intelligent transportation systems by striking a balance between accuracy, robustness and efficiency.
LSTM guided homomorphic encryption for threat-resistant IoT networks Sanjeev Kumar, Sukhvinder Singh Deora, Tajinder Kumar, Purushottam Sharma, Xiaochun Cheng, Vishal Garg Discover Computing, 2025 The rapid growth of the Internet of Things (IoT) has led to revolutionary innovations in many fields; however, it has also resulted in significant security and privacy issues due to the resource limitations and distributed nature of IoT networks. Traditional cryptographic techniques or machine learning-based anomaly detection systems do not jointly provide data privacy and resilience to threats in real time. The existing methods, such as Homomorphic Encryption (HE), offer a high computation cost for performing encryption. Furthermore, Long Short-Term Memory (LSTM) networks can predict an anomaly profile instead of performing encryption. To address these shortcomings, this paper proposes NeuroCrypt. This new hybrid system combines Fully Homomorphic Encryption (FHE) with LSTM-based encrypted anomaly detection and supplements it with blockchain-based dynamic key management and multi-factor authentication. The architecture targets edge and fog computing settings using, among other techniques, ciphertext packing, model quantisation, and parallelised encrypted operations. The performance of the proposed framework has been evaluated on a real dataset. The results show that the accuracy in the proposed framework is 99.2% compared to existing techniques such as HE-based DNN, FL-based models, and LSTM IDS. Conclusively, NeuroCrypt provides a privacy-preserving, effective, and scalable solution to real-time threat abatement in IoT networks.
Middleware architecture performance analysis for vehicular ad hoc network Rajender Kumar, Punit Soni, Purushottam Sharma, Tajinder Kumar, Xiaochun Cheng, Mandeep Singh, Mrinal Paliwal Eurasip Journal on Wireless Communications and Networking, 2025 Implementing Intelligent Transportation Systems (ITS) raises serious safety concerns, directly contributing to good health and well-being by enhancing road safety. Intelligent transport technologies are used by vehicular ad hoc networks (VANET) to enhance traffic flow and safety on the roads aligning with sustainable cities and communities. For this purpose, a variety of techniques have been examined in this article. The design and analysis of middleware architecture for VANET are also covered in this paper, promoting industry, innovation, and infrastructure. Since the automobiles travel deliberately rather than carelessly to connect with roadside equipment by limiting the range of motion, the first implementation of this relies on VANET networks, middleware, and heuristic technique. The VANET network is replacing wireless telephony, mobile nodes are evolving into vehicle nodes, and the transport system is changing to an intelligent transport system. In terms of latency (high 10.14%), power dissipation (less 2.46%), throughput (high 2.82%), and overall cumulative performance (high 3.12%) on different nodes ranging from 100 to 500, the experimentation results show that the middleware and VANET architecture are superior to the heuristic approach, contributing to responsible consumption and production through improved efficiency.
Energy Efficient VM Selection Using CSOA-VM Model in Cloud Data Centers Mandeep Singh Devgan, Tajinder Kumar, Purushottam Sharma, Xiaochun Cheng, Shashi Bhushan, Vishal Garg Caai Transactions on Intelligence Technology, 2025 The cloud data centres evolved with an issue of energy management due to the constant increase in size, complexity and enormous consumption of energy. Energy management is a challenging issue that is critical in cloud data centres and an important concern of research for many researchers. In this paper, we proposed a cuckoo search (CS)‐based optimisation technique for the virtual machine (VM) selection and a novel placement algorithm considering the different constraints. The energy consumption model and the simulation model have been implemented for the efficient selection of VM. The proposed model CSOA‐VM not only lessens the violations at the service level agreement (SLA) level but also minimises the VM migrations. The proposed model also saves energy and the performance analysis shows that energy consumption obtained is 1.35 kWh, SLA violation is 9.2 and VM migration is about 268. Thus, there is an improvement in energy consumption of about 1.8% and a 2.1% improvement (reduction) in violations of SLA in comparison to existing techniques.
Measuring the impact of predictive analytics on patient satisfaction Tajinder Kumar, Sachin Lalar, Sarbjit Kaur, Vinay Goyal Advancing Healthcare Through Decision Intelligence Machine Learning Robotics and Analytics in Biomedical Informatics, 2025 A critical factor in determining the quality of healthcare is patient satisfaction , it is the patients’ compliance to their treatment plans and overall well-being and success of the health care organizations. It has emerged as a strong weapon in the recent years to overall enhance the patient centeredness with patient-specific care needs. This may lead to improvement of the satisfaction level of the patients. This chapter is devoted to the analysis of the theoretical framework that links patient satisfaction and predictive analytics , which shows the possibilities of the data-driven methods in the healthcare domain. Big data is used to proactively identify patients’ needs and tailor patient care and support by analyzing, among others, demographic data, treatment records, patient’s electronic health records , and even social factors that may influence health status . Risk analysis and possible problems can be identified with more precision by health care services through the use of predictive analytics ; thus, the business can allocate resources in the most effective manner, while service to the clients becomes much more proactive and client oriented. This chapter focuses on patient satisfaction of the applied predictive analytics in the field of healthcare and discusses it in detail. Some specific examples of specific cases as well as factual information describing the degree of the changes that affect the patient satisfaction index and caused by the application of predictive analytics are described in the chapter. Some examples of benefits include, patient’s wait time, coordinated care, and better management of chronic diseases . Ethical issues of data patient use including informed consent and privacy in the usage of the patient data are also asserted. This, in turn, strengthens the role that was taken by strong data management frameworks and open data sharing procedures for developing and maintaining patients’ trust. These technologies can revolutionize patient experience from care providers by providing the interventions and treatment regimens that have the likelihood of yielding the best results as well as identifying potential adverse outcomes so that health care providers can react proactively. Analytics is going to make patient satisfaction, superior medical outcomes, and better health care delivery systems predictable.
Generative Artificial Intelligence (GAI) for Accurate Financial Forecasting Tajinder Kumar, Sachin Lalar, Vishal Garg, Pooja Sharma, Ravi Dutt Mishra Generative Artificial Intelligence in Finance Large Language Models Interfaces and Industry Use Cases to Transform Accounting and Finance Processes, 2025 Prudent financial management relies heavily on accurate financial forecasting, which helps companies reduce risks, allocate resources wisely, and make well-informed decisions. In the dynamic world of finance, incorporating generative artificial intelligence (GAI) into financial forecasting has become a game-changer, potentially improving forecast accuracy and dependability. This study examines the use of GAI in economic forecasting, emphasizing how revolutionary it can be in improving predictive accuracy. To create a dynamic and adaptive forecasting framework, GAI combines the capabilities of generative models and artificial intelligence with a large dataset, historical financial data, market trends, and macroeconomic indicators. This paper explores the working mechanisms of GAI, highlighting its capacity to produce synthetic data, model a wide range of scenarios, and identify complex patterns and relationships in financial data. GAI differs from conventional forecasting techniques due to its ability to learn unsupervised and its flexibility in dealing with chaotic and non-linear market conditions. Several case studies and real-world applications that show the concrete effects of GAI on financial forecasting are also highlighted in the abstract. Notable examples include more accurate revenue projections, better risk assessment, and better stock price forecasts. These real-world examples highlight the significance of GAI in streamlining decision-making procedures for traders, investors, and financial professionals. The ethical use of AI, transparency, and fairness are emphasized in the discussion of ethical issues in GAI-driven financial forecasting. The abstract discusses the necessity of regulatory frameworks that protect financial data security and privacy while guaranteeing responsible AI implementation. This paper emphasizes how GAI can revolutionize the field of financial forecasting. Through integrating artificial intelligence and sophisticated generative modeling, GAI enables financial institutions and businesses to make data-driven decisions with previously unheard-of precision and assurance. With GAI at the forefront of increasingly accurate and trustworthy forecasts that will ultimately result in better financial stability and strategic decision-making, the field of financial forecasting is about to undergo a paradigm change.
Enhanced Triple Layered Approach for Mitigating Security Risks in Cloud Tajinder Kumar, Purushottam Sharma, Xiaochun Cheng, Sachin Lalar, Shubham Kumar, Sandhya Bansal Computers Materials and Continua, 2025 : With cloud computing, large chunks of data can be handled at a small cost. However, there are some reservations regarding the security and privacy of cloud data stored. For solving these issues and enhancing cloud computing security, this research provides a Three-Layered Security Access model (TLSA) aligned to an intrusion detection mechanism, access control mechanism, and data encryption system. The TLSA underlines the need for the protection of sensitive data. This proposed approach starts with Layer 1 data encryption using the Advanced Encryption Standard (AES). For data transfer and storage, this encryption guarantees the data’s authenticity and secrecy. Surprisingly, the solution employs the AES encryption algorithm to secure essential data before storing them in the Cloud to minimize unauthorized access. Role-based access control (RBAC) implements the second strategic level, which ensures specific personnel access certain data and resources. In RBAC, each user is allowed a specific role and Permission. This implies that permitted users can access some data stored in the Cloud. This layer assists in filtering granular access to data, reducing the risk that undesired data will be discovered during the process. Layer 3 deals with intrusion detection systems (IDS), which detect and quickly deal with malicious actions and intrusion attempts. The proposed TLSA security model of e-commerce includes conventional levels of security, such as encryption and access control, and encloses an insight intrusion detection system. This method offers integrated solutions for most typical security issues of cloud computing, including data secrecy, method of access, and threats. An extensive performance test was carried out to confirm the efficiency of the proposed three-tier security method. Comparisons have been made with state-of-art techniques, including DES, RSA, and DUAL-RSA, keeping into account Accuracy, QILV, F-Measure, Sensitivity, MSE, PSNR, SSIM, and computation time, encryption time, and decryption time. The proposed TLSA method provides an accuracy of 89.23%, F-Measure of 0.876, and SSIM of 0.8564 at a computation time of 5.7 s. A comparison with existing methods shows the better performance of the proposed method, thus confirming the enhanced ability to address security issues in cloud computing.
Business modeling using agile Lalita Chaudhary, Vikas Deep, Vishakha Puniyani, Vikram Verma, Tajinder Kumar Advances in Intelligent Systems and Computing, 2016
