Research & Publications

DistillMed offers an efficient solution for automating ICD coding through neural distillation, achieving 98.7% teacher model accuracy without losing accuracy. This model addresses the need for faster, cost-effective coding in healthcare with significant potential to enhance medical coding efficiency.

This survey provides the first comprehensive exploration of LLM integration into Multi-Robot Systems. It systematically categorizes applications across high-level task allocation, mid-level motion planning, low-level action generation, and human intervention, highlighting key applications in diverse domains and examining challenges that limit adapting LLMs in MRS.

This systematic review covers all applications of AI in the clinical use of Diabetic Nephropathy, analyzing 24 peer-reviewed articles. The study revealed that AI frameworks outperformed conventional statistical methods, while identifying limitations such as lack of data availability and external validation.

This study examines how diabetes patients were exposed to subacute risk during the unanticipated lockdown, with most unable to engage in physical activity, leading to increased blood glucose levels. The research proposes initiatives to prevent similar outcomes in future lockdown scenarios.

This review broadly emphasizes different chemical and radiation techniques for industrial wastewater treatment, providing an overview of their advantages, disadvantages, and applications. The article also discusses potential improvements and other advantageous methods in treating wastewater.

This study uses cutting-edge machine learning techniques to predict crime types in Philadelphia. By harnessing the power of deep neural networks, we uncovered that location and time are the most pivotal features for accurate crime prediction in urban areas. Our models, on average, show a 17% increase in accuracy over state of the art approaches.

This study examines female sexual offense cases during the COVID-19 pandemic at a tertiary medical college in Dhaka, Bangladesh, analyzing patterns, contributing factors, and implications for public health and safety measures during crisis periods.