International Journal of Engineering Science Technologies

ENGINEERING APPROACHES IN THE DIAGNOSIS OF SLEEP APNEA

Hatice Bilgili — 2026-03-09

Sleep apnea is a sleep disorder that significantly affects human life and occurs as a result of repeated obstructions in the respiratory system lasting at least 10 seconds during sleep. The most common type, Obstructive Sleep Apnea (OSA), affects the upper respiratory tract, whereas Central Sleep Apnea (CSA) occurs due to dysfunction in the respiratory control center in the brain. Sleep apnea manifests with symptoms such as fatigue upon awakening, snoring, and daytime sleepiness. If left untreated, it may lead to serious health complications including stroke, cardiovascular diseases, and hypertension. Polysomnography (PSG) is the most widely used diagnostic method for sleep apnea. However, this test involves several limitations in terms of time consumption, patient comfort, and financial cost. Therefore, there is an increasing need for alternative engineering-based diagnostic support methods to complement polysomnography. Recent advancements in Biomedical Engineering, Electrical and Electronics Engineering, and Software Engineering have enabled the development of portable, cost-effective, and highly compatible systems for sleep apnea detection. Sleep apnea can be identified through the processing of physiological signals such as electroencephalography (EEG), electrocardiography (ECG), electrooculography (EOG), and oxygen saturation levels. The acquired data are analyzed using artificial intelligence techniques and machine learning algorithms, which have become prominent tools in biomedical signal analysis. Furthermore, the integration of wearable devices and Internet of Things (IoT)-based technologies allows continuous monitoring of patients in home environments. This study discusses the significance of engineering-based solutions in sleep apnea diagnosis and highlights their contributions to modern healthcare technologies.

MAGNETIC DRUG TARGETING USING MAGNETIC NANOPARTICLES FOR CANCER THERAPY

Hatice Bilgili — 2026-03-12

Globally, cancer is still among the most frequent causes of mortality.Traditional cancer treatment modalities, including surgery, radiotherapy, and chemotherapy, present limited selectivity and severe side effects. Chemotherapeutic agents target the growth and survival of rapidly dividing cancer cells, but they can also kill healthy cells through systemic circulation. Hence, the quest for specific drug-targeting systems for tumor-targeting agents in the therapeutic arm has attracted considerable attention in cancer therapy. Magnetic nanoparticles (MNPs) have attracted significant attention in biomedical research owing to their specific physical and features such as their small particle size and high surface-to-volume, and superparamagnetic behavior. Magnetite (Fe₃O₄) and maghemite (γ-Fe₂O₃), which belong to the iron oxide nanoparticle family, are extensively investigated for biomedical because they are biocompatible and exhibit controllable magnetic properties. These nanoparticles can be functionalized with biocompatible coatings and anticancer agents to develop magnetic drug delivery systems. Magnetic Drug Targeting (MDT) is reliesed on the principle of magnetic nanoparticles conjugated with therapeutic agents directed to the cancerous tissue region by a magnetic field applied externall Such an approach makes increased drug accumulation within the targeted tissue possible, and a substantial impact has been achieved with systemic toxicity and side effects minimized. Furthermore, magnetic nanoparticles have been applied widely across biomedical fields, including magnetic resonance imaging (MRI), hyperthermia therapy, biosensors, and tissue engineering. The basic properties of magnetic nanoparticles, their biocompatibility, their application in cancer anticancer drug targeting using magnetic drug, and their properties in the context of the basic characteristics are investigated in this review as well. In addition, discussion has been provided on magnetic targeting dynamics, drug applications, and anti-cancer drugs with magnetic nanoparticles, as well as targeted drug system issues, based on the available literature. Drug delivery using magnetic nanoparticle systems is a promising strategy for optimizing cancer treatment while minimizing side effects.

EFFECT OF POYNTING-ROBERTSON FORCE ON THE RESONANT MOTION OF GEOCENTRIC SATELLITE

Md Sabir Ahamad — 2026-03-30

This paper is to discuss the effects of Poynting-Robertson force on the resonant motion of geocentric satellite. In presence of Poynting-Robertson force the resonances and are occurred. Also discuss the amplitude and time period of the geocentric satellite at all these resonant points.

AI-BASED SOIL HEALTH ANALYSIS AND CROP RECOMMENDATION SYSTEM FOR SMART FERTILIZER MANAGEMENT IN PRECISION AGRICULTURE

Madhuri Deepak Mulje — 2026-04-09

Precision agriculture is changing the modern agriculture system by adopting the artificial intelligence (AI) technology that enhances the performance of agricultural systems and environmental conservation. The study introduces an artificial intelligence solution that assesses the state of soil and suggests farming methods to attain the most optimal use of fertilizers and enhanced crop production outcomes. The system employs machine learning algorithms to handle the key soil parameters that comprise pH, moisture levels, nutrient content and temperature readings. The system takes these inputs to decide the level of soil fertility as it also recommends the kind of crops and their particular requirements of fertilizer. The given model uses empirical data to pursue three goals that encompass the reduction of fertilizer use and minimization of environmental damage and the progress of more environmentally friendly practices in agriculture. The system will help farmers make prompt decisions as it will give them a smart system to communicate with. As demonstrated by the experiment, our system is superior when compared to the traditional methods, not only in terms of selecting crops precisely, but also with efficient nutrient management. The strategy will allow farmers to adopt clever farming practices that will offer them inexpensive and environmental-friendly practices that yield high agricultural yields.

A COMPREHENSIVE REVIEW OF WIDE-AREA MONITORING AND CONTROL OF POWER SYSTEMS USING PHASOR MEASUREMENT UNITS PMUS

Sreedhara Babu Korukonda — 2026-04-14

The modern power systems are becoming increasingly complex due to the large-scale interconnections, the integration of renewable energy sources, and the dynamic changes in loads, which require the sophisticated monitoring and control systems. Phasor Measurement Units (PMUs) have led to the creation of Wide-Area Monitoring, Protection, and Control (WAMPAC) systems, capable of transforming the grid, boosting its reliability, stability, and operational efficiency. PMUs are also used to offer time-synchronized measurements (synchro phasors) that can offer real-time insight into the dynamics of a system, which was not possible with traditional SCADA systems. This paper provides an extensive overview of the PMU technology and its application in Wide-Area Monitoring Systems (WAMS), their structure, essential elements, and communication structure. It also discusses key applications like real-time monitoring, state estimation, oscillation detection, fault location, voltage stability assessment, and wide-area control. The study also addresses some of the critical challenges, such as high implementation costs, data management complexities, cybersecurity threats, and communication latency, as well as interoperability issues. Further recent developments, including the combination of PMUs with smart grids, artificial intelligence, cloud computing, and Wide-Area Control Systems (WACS) are discussed, as well as future research directions, including optimization of costs, improved security, and autonomous operation of grids. In general, PMU-based WAMS is found as a foundational technology in terms of designing intelligent, resilient, and sustainable power system.

A COMPREHENSIVE REVIEW OF FINITE ELEMENT MODELLING TECHNIQUES FOR REINFORCED CONCRETE BEAMS UNDER IMPACT AND BLAST LOADING CONDITIONS

Gavhane Santosh Mohanrao — 2026-04-14

RC beams are important structural components that are becoming vulnerable to extreme dynamic loads like impact and blast events whereby they experience complex nonlinear responses where strain rates are high, energy transfer is rapid and localized damage. It is vital to understand how they behave under these conditions in order to achieve structural safety and resilience. The Finite Element Modelling (FEM) has proved to be a strong computational model that can effectively model material nonlinearity, stress wave propagation, cracking and failure mechanisms that cannot be effectively modeled by experimental techniques. This paper is a review of FEM methods used in the study of RC beam under impact and blast loading, with the main issues addressed in the paper being modelling methods, constitutive models of materials, methods of loading, and failure, and validation strategies. It also looks at recent developments, such as hybrid modelling methods, incorporation of high-performance materials, and also explains the current problems, such as the computational complexity and uncertainty in the parameter. The purpose of the review is to offer a synthesized view of the existing practices, and to determine the future research directions towards improving the face validity and efficacy of numerical simulations in structural engineering.

A COMPREHENSIVE REVIEW ON GENETIC ALGORITHM-BASED OPTIMIZATION OF STEEL TRUSS STRUCTURES FOR WEIGHT REDUCTION AND STRUCTURAL EFFICIENCY

Santosh Haridas Surtar — 2026-04-17

The steel truss structure is a typical engineering structure since it is extremely robust relative to its weight and its structure. Nevertheless, there is still a considerable challenge in trying to optimize these structures to ensure minimum weight, safety and performance. Conventional optimization methods may be ineffective in dealing with nonlinear and multiple constraints design spaces. The artificial algorithms based on the concept of natural selection and evolution that have become a powerful tool to resolve such optimization problems are called Genetic Algorithms (GAs). This review paper will discuss in detail the GA-based optimization solution to steel truss structures. It talks about techniques, encoding, constraint management, mixed methods and newer techniques. Challenges, comparative studies, future research directions are also pointed out in the paper.

CUSTOMER CHURN PREDICTION IN TELECOM USING MACHINE LEARNING AND DATA MINING

Smita Pandey — 2026-04-20

Customer churn prediction is crucial in reducing the loss of customers and enhancing the retention strategies by telecom companies. This paper suggests a machine learning-based system in the case of the IBM Telco Customer Churn data to identify the probability of a customer switching the service. The methods of data preprocessing, including dealing with missing values, coding of categorical variables, log transformation, and feature scaling are used to improve the quality of data. Exploratory Data Analysis (EDA) will be performed to find some trends and factors that are important in churn. There are several supervised learning models, such as Decision Tree, Random Forest, and X GBoost that are implemented and evaluated. Random Oversampling is used to deal with the issue of class imbalance in order to enhance the model performance on minority class examples. Training and testing accuracy is used to evaluate the models with the ensemble models (Random Forest and XG Boost) performing well and generalizing better than the Decision Tree model. The findings show that the type of contract, technical support and payment method are important factors influencing the customer churn, which means that machine learning techniques are quite helpful in the customer retention strategies of the telecom industry.

EMOTION-AWARE ADAPTIVE MUSIC RECOMMENDATION SYSTEM USING REAL-TIME AFFECTIVE STATE ANALYSIS

Harish Barapatre — 2026-04-30

Emotion plays a critical role in human–music interaction, influencing listening behavior, mood regulation, and cognitive engagement. Existing music recommendation systems, such as those used in Spotify and Apple Music, primarily rely on historical user preferences, collaborative filtering, or genre-based classification, which fail to capture the dynamic and real-time emotional states of users. This limitation results in suboptimal personalization and reduced user satisfaction.

This paper proposes an Emotion-Aware Adaptive Music Recommendation System that integrates real-time affective state detection with intelligent music mapping. The framework utilizes multimodal inputs such as facial expressions, textual sentiment, or physiological cues to infer user emotions and dynamically adjust music recommendations. A structured pipeline is designed to process emotional signals, compute emotion intensity scores, and map them to suitable music features such as tempo, genre, and energy levels.

Unlike traditional systems, the proposed approach emphasizes context-aware personalization, enabling continuous adaptation to changing user emotions. The system is conceptualized with a mathematically grounded scoring mechanism and an interpretable decision layer to ensure transparency and robustness. The proposed framework contributes to the advancement of affective computing in entertainment systems and provides a foundation for next-generation intelligent media platforms.

AN INTELLIGENT DEEP LEARNING-BASED FRAMEWORK FOR REAL-TIME SIGN LANGUAGE RECOGNITION USING VISION-BASED GESTURE ANALYSIS

Harish Barapatre — 2026-04-30

Sign language recognition has emerged as a critical research area aimed at reducing the communication barrier between hearing-impaired individuals and the general population. Traditional communication methods often rely on human interpreters, which are not always accessible, scalable, or cost-effective. Recent advancements in computer vision and deep learning have enabled the development of automated systems capable of interpreting hand gestures and translating them into meaningful text or speech. However, existing systems often suffer from limitations such as sensitivity to background noise, lack of real-time performance, and insufficient generalization across different signers 1, 2.
This paper proposes an intelligent vision-based sign language recognition framework that leverages deep learning techniques for accurate and real-time gesture interpretation. The system captures hand gestures through a camera interface, performs preprocessing to extract relevant spatial features, and utilizes convolutional neural networks (CNNs) for feature learning and classification. Additionally, temporal dependencies in dynamic gestures can be modeled using sequence-based architectures, enhancing recognition capability [3]. The proposed framework is designed to be scalable, robust to environmental variations, and deployable in real-world assistive applications.
The primary contribution of this work lies in designing a structured, end-to-end pipeline that integrates gesture acquisition, feature extraction, classification, and output generation into a unified system. The framework aims to improve accessibility, enable real-time communication support, and serve as a foundation for future multimodal interaction systems.

DEEP LEARNING AND BLOCKCHAIN-ENABLED FRAMEWORK FOR BITCOIN PRICE PREDICTION AND SECURE TRANSACTION INTELLIGENCE

Harish Barapatre — 2026-04-30

Bitcoin price prediction has become a critical research problem due to its extreme volatility and increasing adoption in financial systems. Traditional statistical and machine learning models often fail to capture the complex nonlinear dependencies and temporal dynamics present in cryptocurrency markets. In recent years, deep learning techniques such as Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU) have demonstrated strong capability in modeling sequential financial data and extracting hidden temporal patterns Nakamoto (2008), LeCun et al. (2015). However, most existing approaches rely solely on historical price data and ignore the rich transactional and structural information available in blockchain networks.
This paper proposes a hybrid conceptual framework that integrates deep learning-based time-series prediction with blockchain-based transaction intelligence. The proposed system utilizes historical Bitcoin price data, trading volume, and blockchain-derived features such as transaction count, hash rate, and wallet activity to enhance prediction accuracy. Additionally, blockchain technology ensures data integrity, transparency, and resistance to tampering, thereby improving trustworthiness in financial prediction systems Hochreiter and Schmidhuber (1997), Cho et al. (2014).
The framework combines feature engineering, deep neural architectures, and secure blockchain data validation into a unified pipeline. This approach not only improves predictive capability but also introduces a secure and verifiable mechanism for financial data processing. The proposed model is expected to provide more robust and reliable Bitcoin price forecasts compared to conventional methods.