Deep Learning Based Non-Invasive Screening of Autism Spectrum Disorder Using Transfer Learning
Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition characterized by persistent challenges in social communication. Early intervention is paramount; however, traditional diagnostic pathways often take years due to a lack of specialized clinicians. This research proposes an automated screening tool using facial image analysis. By employing the VGG16 architecture via Transfer Learning, we extract high-level spatial features from facial landmarks to identify markers associated with ASD. Our findings indicate that computational models can provide a significant preliminary screening layer, reducing the burden on clinical resources.
Published by: Nisha Sharma, Bablu Jaipal
Author: Nisha Sharma
Paper ID: V12I3-1213
Paper Status: published
Published: June 10, 2026
Lane Detection System using Python OpenCV
Lane detection is an important perception module in advanced driver-assistance and autonomous driving because it helps a vehicle interpret road geometry and remain centered within the lane. This paper presents a compact lane detection pipeline developed in Python with OpenCV using classical image-processing techniques. The method processes each video frame in sequence and applies grayscale conversion, Gaussian smoothing, Canny edge detection, a region-of-interest mask, and Probabilistic Hough Transform line extraction. The detected segments are separated into left- and right-lane candidates using slope-based rules, averaged to reduce noise, and drawn on the original frame to create an annotated road view. The system was tested on real driving video captured from a front-facing camera under normal daylight conditions. The results indicate that the approach performs well on straight roads and moderate curves when lane markings are visible, but its robustness decreases under shadows, glare, faded paint, and partial occlusion. Because the pipeline is lightweight, deterministic, and capable of near real-time execution on standard hardware, it is a useful baseline for educational and prototype intelligent transportation systems. The paper also discusses the problem context, design objectives, implementation steps, results, limitations, and future extensions such as adaptive thresholding, temporal tracking, and learning-based lane recognition.
Published by: Yash Bhardwaj
Author: Yash Bhardwaj
Paper ID: V12I3-1211
Paper Status: published
Published: June 9, 2026
Lean Manufacturing Techniques to Enhance the Productivity of Hard Coat Process in Automotive Lighting System
Kaizen, a Japanese term meaning "change for the better," is a core Lean management philosophy focused on continuous, incremental improvement. Unlike radical overhauls, Kaizen emphasizes small, daily changes made by every employee—from the shop floor to the executive suite—to cumulatively drive significant gains in productivity and quality. As a Lean technique, Kaizen functions as the primary engine for eliminating waste (Muda), reducing variability, and optimizing process flow. It relies on a participatory culture where frontline workers are empowered to identify inefficiencies and propose solutions at the source of work, known as the Gemba. The methodology is typically operationalized through the Plan-Do-Check-Act (PDCA) cycle, ensuring that improvements are data-driven, tested, and standardized to prevent regression. The implementation of Kaizen within a Lean framework results in measurable benefits, including reduced lead times, lower defect rates, and enhanced employee morale. Ultimately, Kaizen transforms an organization into a "learning enterprise" capable of sustaining long-term competitive advantages without requiring massive capital investment.
Published by: Majaharali Shikalgar, Dr. Javed G. Khan
Author: Majaharali Shikalgar
Paper ID: V12I3-1206
Paper Status: published
Published: June 5, 2026
Real-Time Smart Road Accident Detection System Using IoT and Sensors for Faster Emergency Response
This literature review examines studies related to real-time smart road accident detection systems using IoT, sensors, communication technologies, and intelligent emergency response methods. The reviewed studies show that accident detection systems commonly use accelerometers, gyroscopes, GPS, GSM/GPRS, microcontrollers, cloud platforms, dashboards, and real-time alerting mechanisms to reduce delays in emergency notification. Some studies focus on low-cost vehicle-mounted systems, while others explore broader emergency response platforms, V2V/V2I communication, smart traffic lights, driver monitoring, and AI-based accident or incident detection. The reviewed studies also show that machine learning, deep learning, computer vision, YOLO-based models, and multi-agent systems can support accident recognition and traffic monitoring. However, several limitations remain, including limited real-world testing, false alarms, threshold sensitivity, weak emergency coordination, connectivity problems, limited sensor fusion, and security or privacy concerns. Overall, the reviewed studies support the need for a practical hybrid system that combines IoT-based sensing, location tracking, reliable communication, emergency alerting, and, where suitable, AI-based validation for faster and more reliable road accident response.
Published by: Qutaiba Almassri
Author: Qutaiba Almassri
Paper ID: V12I3-1171
Paper Status: published
Published: June 3, 2026
AI-Driven Breath Analysis for Early Lung Cancer Detection Using Optimized Ensemble Learning of VOC Biomarkers
Lung cancer is one of the most prevalent and deadly diseases worldwide, primarily due to late-stage diagnosis and the limitations of conventional detection methods. Early and accurate identification is crucial for improving patient survival rates. This study proposes a novel, non-invasive approach for lung cancer prediction using AI-enhanced breath analysis based on volatile organic compound (VOC) biomarkers. The proposed system utilizes sensor-based breath data to capture VOC patterns associated with lung cancer. Advanced Preprocessing and feature selection techniques are applied to identify the most relevant biomarkers, reducing data complexity and improving model efficiency. An ensemble machine learning framework, combining multiple classifiers such as Random Forest, Support Vector Machine, and Gradient Boosting, is employed to enhance prediction accuracy and robustness. Experimental evaluation demonstrates that the proposed model achieves high accuracy, precision, and recall, outperforming individual classifiers. The system also shows strong potential for real-time implementation due to its computational efficiency. This approach offers a cost-effective, portable, and non-invasive alternative to traditional diagnostic techniques. Overall, the integration of VOC biomarker analysis with feature-selected ensemble learning provides a promising solution for early lung cancer detection, paving the way for improved screening methods and better clinical outcomes.
Published by: Syed Naseemtaj, Syed Nafeesa Thehseen
Author: Syed Naseemtaj
Paper ID: V12I3-1201
Paper Status: published
Published: June 2, 2026
Association of Hypertension and Diabetes Mellitus with Reperfusion Timelines in STEMI Patients Undergoing Primary PCI at a Tertiary Care Cardiac Centre in Jodhpur, Rajasthan: A Prospective Observational Study
Background ST-Elevation Myocardial Infarction (STEMI) remains one of the leading causes of cardiovascular morbidity and mortality worldwide. Timely diagnosis and early reperfusion therapy are the cornerstones of management and significantly influence clinical outcomes. Evaluation of demographic characteristics, cardiovascular risk factors, and reperfusion timelines is essential for improving quality indicators in acute cardiac care. Aim: To evaluate the demographic profile, cardiovascular risk factors, STEMI patterns, culprit vessel distribution, and reperfusion timelines among STEMI patients presenting to a tertiary care cardiac centre. Materials and Methods This prospective observational study was conducted at Trinay Hospital, Jodhpur, Rajasthan. A total of 60 consecutive patients diagnosed with STEMI and undergoing primary percutaneous coronary intervention (PCI) were included. Data regarding demographic profile, cardiovascular risk factors, STEMI type, culprit vessel, door-to-ECG time, door-to-balloon time, and total ischemic time were collected using a structured STEMI data collection tool. Statistical analysis was performed using descriptive statistics. Statistical Data Analysis: Chi-Square Analysis of Hypertension and Diabetes Mellitus Correlation. A chi-square test was performed to determine the association between hypertension and diabetes mellitus among STEMI patients. Table 1: Correlation Between Hypertension and Diabetes Mellitus Variable Diabetes Present Diabetes Absent Total Hypertension Present 28 11 39 Hypertension Absent 9 12 21 Total 37 23 60 Statistical Findings • Chi-square (χ²) value = 5.84 • Degrees of freedom = 1 • p-value = 0.015 Interpretation A statistically significant association was observed between hypertension and diabetes mellitus among STEMI patients (p < 0.05). Patients with hypertension were more likely to have coexisting diabetes mellitus. Graphical Analysis Histogram Analysis of Reperfusion Timelines The histogram demonstrated that total ischemic time remained substantially higher than door-to-ECG and door-to-balloon times. Although in-hospital management timelines were within acceptable international standards, delayed patient presentation contributed significantly to prolonged ischemic duration. Comparative Demographic Analysis Gender-wise Comparison Variable Male Female Mean Door-to-Balloon Time 79.4 minutes 84. Interpretation of Comparative Analysis Patients with diabetes mellitus and hypertension demonstrated relatively prolonged ischemic times and delayed reperfusion compared to patients without these risk factors. Female patients also showed slightly prolonged treatment timelines compared to male patients. Results: The mean age of patients was 56.7 ± 8.7 years. Male patients constituted 68.3% of the study population. Hypertension was present in 65%, diabetes mellitus in 61.7%, smoking history in 53.3%, alcohol consumption in 46.7%, and prior coronary artery disease in 25% of patients. Inferior wall STEMI was the most common presentation (45%), followed by anterior wall STEMI (36.7%). The right coronary artery and left anterior descending artery were equally involved as culprit vessels (43.3% each). Mean door-to-ECG time was 12.48 minutes, while mean door-to-balloon time was 81.18 minutes. Mean total ischemic time was 282.6 minutes. Conclusion: STEMI predominantly affected middle-aged male patients with multiple cardiovascular risk factors. Early reperfusion metrics observed in the present study were within acceptable international standards for primary PCI centres. Continuous monitoring of STEMI timelines and quality indicators can further improve patient outcomes.
Published by: Dr. Dhruva Sharma
Author: Dr. Dhruva Sharma
Paper ID: V12I3-1199
Paper Status: published
Published: May 28, 2026
