Research & Development

AI Research & Model Development

Our research division is developing foundational AI models for symptom analysis, medical imaging, and diagnostic support. Each project follows rigorous research methodology with transparent progress tracking and measurable objectives.

Anjali Aggarwal

Chief Research Officer

Symptom-to-Department Classification Model

Intelligent symptom analysis for accurate department routing in healthcare facilities

Developing a natural language processing model that analyzes patient symptom descriptions in multiple Indian languages and predicts the appropriate medical department for consultation. The model uses a custom-trained transformer architecture with domain-specific fine-tuning on Indian healthcare data, achieving context-aware understanding of vernacular symptom reporting.

Overall Development Progress 70%

Last updated: 29 March 2026

✓

Data Collection

Completed

✓

Model Training

Completed

✓

Initial Build

Completed

Testing Phase

In Progress

Fine-Tuning

Pending

Validation

Pending

NLP

Transformers

Multi-Lingual

Classification

Aashna Tasawwur

Research Officer (Lead)

Brain Tumor Detection Model

Deep learning model for automated MRI-based brain tumor identification and classification

Building a convolutional neural network for detecting and classifying brain tumors from MRI scans. The model employs advanced image processing techniques including data augmentation, attention mechanisms, and ensemble learning to achieve clinical-grade accuracy. Designed for deployment in resource-constrained settings with minimal computational overhead.

Overall Development Progress 55%

Last updated: 27 March 2026

✓

Dataset Prep

Completed

✓

Architecture Design

Completed

✓

Model Training

Completed

Performance Testing

In Progress

Fine-Tuning

Pending

Clinical Validation

Pending

Computer Vision

CNN

MRI Analysis

Medical Imaging

Research Team

Devansh Diwakar

Ujjwal Sharma

Vikash Yadav

Research Officer

X-Ray Multi-Disease Detection with Explainable AI

Comprehensive chest X-ray analysis with visual explanations and heat mapping

Developing a sophisticated multi-label classification model for detecting multiple respiratory and thoracic conditions from chest X-rays. The system integrates explainable AI (X-AI) techniques including Grad-CAM heat maps, attention visualization, and anatomical region highlighting to provide clinically interpretable results. Pain point mapping identifies specific areas of concern with localization accuracy, enabling radiologists to validate findings efficiently.

Overall Development Progress 48%

Last updated: 28 March 2026

✓

Literature Review

Completed

✓

Dataset Curation

Completed

Model Development

In Progress

X-AI Integration

In Progress

Testing

Pending

Clinical Trials

Pending

Explainable AI

Grad-CAM

Multi-Label

Radiology

Heat Mapping

Platform Engineering

Web Application & Integration Development

Building the unified healthcare platform that connects AI models, patient records, and care networks into a seamless, ABHA-integrated experience accessible across India.

Prabhat Yadav

Web Application Head

Unified Healthcare Platform Development

Full-stack web application integrating AI models, patient records, and care coordination

Engineering the core web application that serves as the central interface for patients, healthcare providers, and administrators. This platform integrates all AI diagnostic models (OejaSymp, OejaScan), manages ABHA-linked health records with blockchain-verified data integrity, and orchestrates the OejaRekha volunteer network. Built with scalability and offline-first architecture to serve rural India with intermittent connectivity.

Overall Development Progress 65%

Last updated: 29 March 2026

✓

Architecture Design

Completed

✓

Frontend Framework

Completed

✓

Backend APIs

Completed

AI Model Integration

In Progress

ABHA Integration

In Progress

Security Audit

Pending

React.js

Node.js

PostgreSQL

REST APIs

PWA

ABHA