JEEVAN SETU

Team Name: CodeSlayers

​

Team Leader : Gowthaman M – gowthamanm.cse2024@citchennai.net

Member 2 : Arun Kumar – arunkumara.cse2024@citchennai.net�Member 3 : Senthanamuthan S – senthanamuthans.cse2024@citchennai.net

Member 4 : Hari Venkatanarayanan V – harivenkatanarayananv.cse2024@citchennai.net

Member 5 : Harish Madhavan S – harishmadhavans.cse2024@citchennai.net

​

Contact(Team leader) : 7200650539

PROBLEM STATEMENT:

“Thalassemia care is currently reactive and fragmented, leading to slow blood transfusions, ineffective donor and genetic matching, and elevated patient risk due to the lack of a cohesive, predictive, AI-driven care ecosystem.”

• Thalassemia care today is largely reactive, with treatments triggered only by symptoms or scheduled visits.
• This leads to treatment delays and increased risk of iron overload complications.
• Communication between patients, donors, and specialized centers remains fragmented.
• The absence of a unified, predictive, AI-driven ecosystem slows donor identification.
• Complex matches (e.g., HLA stem cell donors) become inefficient, leaving patients without proactive, continuous care.

​

PROPOSED SOLUTION:

“JeevanSetu: AI-powered platform for timely blood access and Thalassemia care.”

• The system uses predictive analytics to forecast blood transfusion needs
• enhances stem cell transplant feature through Smart HLA Matching using OCR and NLP to accurately process medical records
• incorporates near-geography based suggestion for timely access
• uses GenAI-driven Donor Nurture Pipeline improves donor engagement and retention

DIAGRAM:

DIAGRAM EXPLAINER TEXT:

Objective: A unified ecosystem turning fragmented medical data into proactive, life-saving interventions.

• Intelligent Data Intake: Uses OCR and NLP to instantly convert unstructured lab reports and EHRs into structured data for tracking HLA markers and blood trends.
• Predictive Analytics: Implements Time-Series AI to forecast hemoglobin drops, allowing for scheduled transfusions before a health crisis occurs.
• Smart HLA Matching: Automates high-resolution 10-point genetic matching for stem cell transplants, reducing coordination time from weeks to seconds.
• Geo-Intelligence & Retention: Pairs real-time local mapping for instant donor sourcing with GenAI-driven "Impact Narratives" to ensure long-term donor engagement.
• Secure Integration: A privacy-first architecture that orchestrates data flow between hospital APIs, blood banks, and patient dashboards.

FEATURES:

• Predictive Transfusion Alerts: Uses time-series AI to forecast hemoglobin drops and pre-schedule blood needs.
• Smart HLA Matching: Automates high-resolution 10-point genetic pairing for stem cell transplants using OCR and NLP.
• Intelligent Data Intake: Extracts structured medical data directly from unstructured lab reports and EHRs.
• Geo-Intelligence: Provides real-time, location-based matching to connect patients with the nearest available donors and centers.
• GenAI Donor Nurture: Increases donor retention by generating personalized impact narratives and engagement stories.
• Unified Care Dashboard: A centralized interface for real-time coordination between patients, donors, and clinicians.

NOVELTY:

• Predictive Crisis Prevention: Replaces reactive panic with Time-Series AI to forecast hemoglobin drops, enabling patients to schedule transfusions weeks in advance before symptoms occur.
• Cognitive HLA Decoding: Utilizes OCR and NLP to instantly translate complex lab reports into structured 10-point genetic matches, automating the search for curative stem cell donors.
• GenAI Impact Storytelling: Directly addresses donor fatigue by generating personalized narratives that emotionally connect donors to the specific biological impact of their contribution.
• Unified Geo-Coordination: A "Google Maps for Life" ecosystem that synchronizes patients, donors, and hospitals in a single real-time, privacy-first dashboard.

TECH STACK:

1. AI & Core Intelligence

Languages: Python (Primary for ML/Analytics).

Medical NLP & OCR: Tesseract or Hugging Face for automated HLA report parsing.

Predictive Engine: Scikit-learn or TensorFlow for time-series transfusion forecasting.

2. Frontend & User Experience

Web/Admin: React.js for clinical and administrative dashboards.

Mobile App: Flutter or React Native for patient/donor cross-platform accessibility.

Mapping: Google Maps API for real-time geo-spatial donor tracking.

3. Backend & Data Management

API & Logic: Node.js (Express) or FastAPI for high-speed AI service orchestration.

Databases: PostgreSQL (Relational medical records) and MongoDB (Unstructured donor logs).

Security: HIPAA-compliant data encryption and ABDM-standard interoperability gateways.

​

THANK YOU