AI & MLFeatured Case Study
AI-Powered Blockchain Healthcare System
Undergraduate Final Research: Tamper-Proof Prescription Verification
96.4%
OCR Accuracy
100%
Tamper Prevention
<1.2s
Verification Time
The Problem Statement
Illegible handwriting and forged medical prescriptions contribute to over $30B in global pharmaceutical counterfeits and adverse drug interactions annually.
Architected Solution
Engineered an end-to-end pipeline using TrOCR (Transformer OCR) to transcribe handwritten doctor prescriptions, cross-checked with doctor cryptographic signatures stored immutably on a private blockchain ledger.
Architectural Implementation Details
- Fine-tuned Transformer-based OCR model for doctor handwriting character sets
- Smart contracts verifying prescription authorization, dispense limits, and expiration
- Zero-knowledge proof validation of patient medical confidentiality
- FastAPI inference gateway integrated with medical dispensary web dashboards
Technologies Deployed
PythonPyTorchTransformersSolidityFastAPIOpenCVWeb3.js