AI Handwritten Prescription Reader
Advanced optical character recognition (OCR) parses complex handwritten doctor prescriptions, extracting medication names, dosage frequencies (1-0-1), administration timing, and treatment durations into structured schedules.
How to Photograph a Prescription for Best OCR Results
- Lay the prescription flat on a flat, evenly lit surface — avoid folding or wrinkling.
- Use direct overhead lighting. Natural window light or a bright overhead lamp gives best contrast. Avoid shadows.
- Hold the camera directly above at 90° to the paper — angled shots distort text.
- Include the entire prescription — doctor's name, date, and all medication lines.
- Ensure the handwriting is in focus — tap to focus on the text area on smartphones.
Example Output
| Medicine | Dose | Frequency | Timing | Duration |
|---|---|---|---|---|
| Tab. Azithromycin 500mg | 500mg | OD (Once Daily) | After food (PC) | 3 days |
| Tab. Pantoprazole 40mg | 40mg | BD (Twice Daily) | Before food (AC) | 5 days |
| Tab. Paracetamol 650mg | 650mg | SOS (When needed) | After food (PC) | As required |
Frequently Asked Questions
Can the AI read illegible handwriting?
Trained on thousands of real Indian doctor handwriting samples. Extremely illegible text will be flagged with a low-confidence warning for manual verification.
What abbreviations does it understand?
All standard Latin abbreviations: OD, BD, TDS, QID, SOS, AC, PC, HS, and numeric formats like 1-0-1. See the full abbreviation guide
What if my prescription has multiple medicines?
Fully supported. Each medicine is extracted as a separate row with its own dosage, frequency, timing, and duration — complete structured medication schedule.
Is my prescription image stored?
No. Processed in-memory only and immediately discarded. Zero data retention — never stored, logged, or shared.