Ongoing Collaboration

Personalized Recommender System for Virus Research and Diagnosis Laboratory Network: Advancing Diagnostic Decision-Making through Artificial Intelligence.  The project aims to use artificial intelligence methods for Personalized Diagnostic Decision-Making & provide Optimized recommendations based on Patient's health information to identify viruses and prioritize pathological test recommendations

Developing AI Based Risk Prediction Model to Gestational Diabetes Mellitus Using Anthropometric, Biochemical, and Genetic Markers. The identified primary objective of this Collaboration is to develop an ML-based predictive model for GDM using social, clinical, anthropometric, biochemical, and genetic parameters and to evaluate the predictive performance of the model in terms of sensitivity, specificity, and overall accuracy. Also to identify key predictors and their interactions that contribute most significantly to GDM risk.

This joint work is to develop a deep learning model, designed to predict ambient noise levels in urban environments. Traditional noise prediction methods face challenges in accurately forecasting noise due to the complex and non-linear nature of noise data. The proposed Neural hierarchical interpolation-based model overcomes these challenges by effectively extracting deep features and learning long-term dependencies in time-series noise data. The model was tested across 30 sites from major Delhi NCR city location (Data of 3 years shared by Delhi Pollution Control Board and NPL), demonstrating superior accuracy and lowest error rates compared to existing models.

To develop an AI incorporated limited sampling strategy (LSS) to predict specific Pharmacokinetic parameters of first line Anti TB drugs. Hypothesis: AI integrated limited sampling strategy (LSS) will be able to predict plasma concentrations at non-sampled time frames (Ppredicted) of intensive sampling and derive the ‘required’ PK values to predict PK failure and possible therapeutic failure.  Will start a project funded by ICMR Titled “Development of an Artificial Intelligence-driven Pharmacokinetics-based Algorithm as an Aid for Better Management of Drug Resistance in Tuberculosis"

Working Jointly with this research Group from Europe in various research problems. That includes Artificial Intelligence in Healthcare, AI based Assistive Devices, Mental Health, Cancer detection, Ovarian cancer, breast cancer etc. Working in a diverse set of research Problems.

Working in a Joint Research Problem to make a Deep Learning Model to classify pulmonary Sounds to detect pulmonary disorders.  1D DeepRespNet and 2D DeepRespNet are developed in this work that were trained and evaluated with normalised 1-D time series and 2-D spectrograms of six types of lung sounds.   

The work explores the use of machine learning models to predict the electrochemical properties of supercapacitor electrodes made from two different set of data acquiring approaches that are cyclic voltammetry and galvanostatic charge discharge method. Various models, including machine learning regression models and neural networks, were developed to forecast specific capacitance, electrical conductivity, and sheet resistance. The study demonstrates that proposed integration enhances the electrode's electrochemical performance, and machine learning can efficiently optimize the composition of these materials for improved energy storage capabilities.

Design and Development of Deep Neural Network for Tracking Foliar Disease Progression in Maize Across Developmental Stages, Cropping Patterns, and Agroclimatic Zones using Mobile Application. This study proposes using advanced AI techniques, including Generative AI, deep CNNs, and Transformer-based models, for early detection of foliar diseases in maize. The findings will be used to develop an AI-enabled mobile application that allows farmers to detect diseases in real-time and offline, reducing the need for expert intervention and enabling timely management.

"Exploring the Potential of Machine Learning, big data analytics and Remote Sensing for Soil Health Monitoring in Indian Agroecosystems". This project aims to develop and validate machine learning models for predicting soil health parameters using remote sensing and big data analytics. By integrating data from remote sensing, soil sensors, and historical agricultural records, the project will provide a comprehensive analysis of soil health. Additionally, it seeks to create a scalable framework for continuous soil health monitoring across various agroecosystems in India.

"Detection and Mapping of Foliar Diseases in Wheat Fields Using UAV-Based Hyperspectral Remote Sensing and Deep Learning for Precision Agriculture." The project focuses on detecting and mapping foliar diseases in wheat fields using UAV-based hyperspectral remote sensing combined with deep learning techniques. This approach enables precise disease identification and mapping, enhancing crop management and yield. The integration of advanced remote sensing and AI facilitates timely interventions and supports precision agriculture practices.