Upload a lab report, prescription, or clinical note. MedDrop reads it, removes patient identity, extracts medical data, and stores it in a standards-based format your systems can use — all in about a minute.
Drop a PDF, Word document, or photo of a lab report. MedDrop reads text from the document — including scanned or photographed pages in 8 languages.
~5 secondsBefore any analysis happens, all personally identifiable information is automatically detected and masked — names, dates of birth, addresses, phone numbers. The AI never sees who the patient is.
~5 secondsA medical AI model analyzes the anonymized text and extracts structured data — lab values with reference ranges, diagnoses, medications — and maps them to international medical codes (LOINC, ICD-10, ATC).
~50 secondsBased on the document type, specialized AI models add an extra layer of analysis — detecting diseases, medications, anatomical structures, cancer markers, or genetic information with high precision.
~5–15 secondsYou see the extracted data side-by-side with the original document. Review, edit if needed, and confirm. The data is then stored as standardized FHIR resources that any compliant system can use.
You decideThe most common use case. Upload a photo or PDF of lab results from any laboratory.
You upload a lab report (PDF, photo, scan)
MedDrop reads every test result, value, unit, and reference range
Values outside normal ranges are automatically flagged
Each test is mapped to its international LOINC code
A structured table of all lab values with flags for abnormal results. Stored as FHIR Observations — queryable, comparable across time, and ready for clinical dashboards.
Extract medication lists from prescriptions, discharge letters, or clinical notes.
You upload a prescription or medication list
AI identifies every medication, dosage, frequency, and route
Drug names are matched to ATC codes for standardization
A specialized pharma detection model double-checks the findings
A clean medication list with international drug codes, dosages, and frequencies. Stored as FHIR MedicationStatements for medication reconciliation and interaction checking.
Dense clinical documents with diagnoses, procedures, lab results, and medication changes — all in one go.
You upload a discharge letter or summary
AI extracts diagnoses (with ICD-10), lab values (with LOINC), and medications (with ATC)
Specialist models for diseases, drugs, and anatomy enrich the extraction
You review everything in a structured view before storing
Observations, Conditions, and MedicationStatements — the complete clinical picture from one document, coded and ready for your EHR system.
Histopathology and biopsy reports require specialized understanding of tissue types, morphology, and staging.
You upload a pathology or biopsy report
AI extracts findings, staging, and diagnoses
Specialized pathology, oncology, and anatomy models identify tissue types, cancer markers, and anatomical structures
Pathological findings linked to anatomical locations and cancer staging. Conditions coded with ICD-10 for oncology workflows.
Results from genetic panels, BRCA testing, pharmacogenomics, or whole-genome sequencing reports.
You upload a genetic test report
AI extracts mutations, variants, and genetic findings
Genomics, DNA, and protein detection models identify genes, mutations, and their clinical significance
Structured genetic findings with gene names, variant descriptions, and clinical implications. Ready for precision medicine workflows.
Culture results, sensitivity testing, and infectious disease reports.
You upload a microbiology or culture report
AI extracts organisms found, sensitivity patterns, and diagnoses
A species detection model identifies pathogens and organisms with high accuracy
Identified organisms, antibiotic sensitivities, and related conditions. Supports infection control and antibiotic stewardship workflows.
Drop a document — PDF, Word, or photo
Check the AI's work before storing
View all stored data in the FHIR viewer
Strips patient identity before AI analysis
Reads documents and extracts structured data
12 models for diseases, drugs, anatomy, oncology, genetics...
Securely stored, always accessible
International standard format, interoperable with any system
Every action logged for compliance
Every AI model and database runs on your own infrastructure. No cloud services, no external API calls, no data leakage.
Names, dates of birth, addresses, and phone numbers are automatically detected and removed before any AI processing.
Every upload, extraction, review, and confirmation is logged with timestamps and user details for regulatory compliance.
AI extracts and suggests — but nothing is stored until a clinician reviews and confirms. Human-in-the-loop by design.