Darkfield microscopy · AI analysis

Stop interpreting.
Start quantifying.

The first AI-powered darkfield blood analysis built for naturopathic professionals. Detect 58+ cell morphologies, quantify findings, and deliver client-ready reports in under 90 seconds.

No credit card · Results in 90 seconds · English & Español

58+
Cell morphologies detected
94%
Classification accuracy
90s
Average analysis time
EN · ES
Bilingual reports
The problem

Manual interpretation is costing you clients, hours, and credibility.

Darkfield analysis is a powerful diagnostic lens — but the way most practitioners do it today is slow, inconsistent, and impossible to defend.

01 — Time

Each case takes 40+ minutes of squinting at a screen.

Between capture, interpretation, and writing up the report, a single consultation eats half your afternoon. You can’t scale a practice one microscope at a time.

02 — Subjectivity

Two practitioners see two different slides.

Without quantification, findings are opinions. Clients notice. Peer review notices. And repeat consultations become impossible to compare over time.

03 — Reports

Clients forget 80% of what you said in the room.

Hand-typed notes and verbal summaries don’t stick. Without a visual, bilingual, evidence-backed report, the value of your work evaporates the moment they leave.

How it works

Upload a sample. BloodWorx does the rest.

Three steps. One pipeline. Built on a detection-tracking-classification model trained on thousands of annotated darkfield fields.

Live darkfield microscopy view from a real practitioner sample — densely packed red blood cells with rouleaux formation visible
STEP 01

Capture from any darkfield scope.

Stream live video or upload still images straight from your microscope camera. No proprietary hardware, no sample shipping, no waiting on a lab.

Drag, drop, done — any resolution, any setup. Your existing microscope just got an AI co-pilot.

RAW VIDEO JPG / PNG UP TO 4K
BloodWorx AI detection panel showing 13 anomaly types and 52 cells flagged with bounding boxes and confidence scores
STEP 02

AI quantifies every cell in the field.

Our RT-DETR detection pipeline counts, classifies, and confidence-scores every cell — across 58+ morphologies including rouleaux, parasitised RBCs, fibrin networks, thrombocyte aggregations, and more.

Every detection is auditable. Every confidence score is shown. No black box.

RT-DETR 58+ CLASSES 94% ACCURACY
Health Implications Overview chart from a real BloodWorx report, showing 17 detected categories ranked by prevalence
Medical Considerations Overview chart from a real BloodWorx report, showing 17 medical categories with inflammation as the dominant finding
STEP 03

Deliver a quantified, bilingual report.

Findings are aggregated into Health Implications and Medical Considerations charts — visual, defensible, and ready to discuss in your consultation.

Branded PDFs in English or Spanish. Track changes across visits. Turn a one-off consult into a long-term care program.

PDF EXPORT EN · ES YOUR BRANDING
What you get

Built for practitioners who charge what they’re worth.

Quantified findings, not vibes.

Every cell class is counted, measured, and scored with a confidence interval. Defensible to skeptical clients, peers, and regulators.

Bilingual reports, auto-generated.

One click delivers a branded PDF in English or Spanish — with your logo, your disclaimers, and your recommendations baked in.

Your data stays your data.

Private cloud, EU-hosted, zero training on your client samples. Export anything, delete everything, any time.

Track cases over time.

Compare the same client’s morphology across visits — visually and numerically. Turn one-off consults into long-term programs.

I was spending three hours on every case. BloodWorx cut that to twelve minutes — and my clients finally keep their reports.
DR
Dr. Rosa Martín Naturopath · Barcelona · Beta cohort 02
LIMITED BETA · COHORT 03 · CLOSING SOON

Your next three analyses are on us.

Join the closed beta and get three full darkfield analyses — delivered, reported, quantified — before you pay a cent. Only 50 practitioners per cohort.

37 / 50 spots claimed in cohort 03