Dimzang Labs Become a Partner

EDDS · How it works

How EDDS turns results into understanding.

Follow one sample profile from a new lab result to a finding a clinician can review.

In developmentNot clinically validated
Walk through an example

Scenario

Ama, 42. Five steps.

Ama is a sample profile, not a real person. Every value below is illustrative.

  1. 01

    A new lab result arrives.

    Ama's latest full blood count is uploaded. EDDS reads the document and turns it into structured values.

    Ingestion

    Lab report · 14 Oct

    PLT  256 ×10⁹/L
    MPV  10.8 fL
    RBC  4.6 ×10¹²/L

    Text recognised · 14 values structured

  2. 02

    It joins her history.

    The new result is placed alongside her six earlier results, so it is read in sequence rather than alone.

    History · Platelets
    • Jan 301
    • Mar 295
    • May 306
    • Jul 290
    • Aug 274
    • Sep 263
    • Oct 256
  3. 03

    Her baseline is compared.

    Her stable results from January to July define what is usual for her: a personal band, much narrower than the population reference range.

    Personal baseline

    298 ×10⁹/L baseline

    Personal band 280–320 · from her January–July results

  4. 04

    Meaningful change is flagged.

    The last three results sit below her usual band and keep moving the same way. Each is still inside the population range.

    Deviation

    −14% vs personal baseline

    Platelets below baseline over 3 results

    Status: emerging signal

  5. 05

    A clinician reviews the finding.

    The AI layer writes a plain-language summary of the statistics. A medical expert reviews it and decides what, if anything, to do next.

    Finding

    Platelets drifting below Ama's usual range

    Three consecutive results below her personal baseline, still within the population range.

    Reviewed This is a signal to discuss, not a diagnosis.

The pipeline

From data to review.

Six steps from raw results to reviewed insight.

  1. DataResults enter from relevant sources.
  2. StructureSnapshots become one coherent record.
  3. BaselineWhat's normal for this person is established.
  4. ChangeMeaningful deviation is measured.
  5. InterpretationPatterns are explained in context.
  6. ReviewFindings are reviewed by medical experts.

What you'd see

Three views. One record.

A simplified, interactive model of the app. Choose a view to see it on the tablet.

Latest upload

Full blood count · 14 Oct

14 values · 1 needs attention

Platelets

256 ×10⁹/L

Below personal band · within population range

Red blood cells

4.6 ×10¹²/L

Within personal band

Sample report

A longitudinal report, page by page.

What a report built from Ama's sample results could look like. Use the buttons or your arrow keys.

EDDS · Longitudinal report

Sample profile

Ama, 42

Haematology · 7 results · January to October

Not a diagnosis. Findings require review by a medical expert.
Summary

One emerging signal

  • Platelets Below personal baseline
  • Mean platelet volume Within band
  • Red blood cell count Within band
Not a diagnosis. Findings require review by a medical expert.
Marker trend · Platelet indices

Platelets −14%, MPV +12% against personal baseline

PERSONAL BASELINEPLT · PlateletsMPV · Mean platelet volumeRBC · Red blood cellsCHANGE DETECTEDJanMarMayJulAugSepOct+10%0%−10%
Not a diagnosis. Findings require review by a medical expert.
Finding

Plain-language summary

Platelet counts have drifted below Ama's usual range across the last three results. All values remain within the population reference range. This pattern may be worth discussing with a clinician.

Reviewed by a medical expert

Not a diagnosis. Findings require review by a medical expert.
Methodology & limitations

How to read this report

  • Baselines are statistical, built from this person's earlier results.
  • A flagged change is a signal for review, not a diagnosis.
  • EDDS is a prototype and is not clinically validated.
  • Results depend on the quality and number of earlier results.
Not a diagnosis. Findings require review by a medical expert.
Page 1 of 5

Safeguards

Built to be careful.

Clinician in the loop

Findings are reviewed by medical experts.

Not a diagnostic

A flagged change is a signal to discuss, never a diagnosis.

Consent & privacy

Designed around informed consent and collecting only what is needed.

Explainable statistics

Statistics come first, so every flag can be traced to the numbers behind it.

EDDS questions

Before you ask.

All questions
Is EDDS available to use?

Not yet. EDDS is a working prototype in development. It is not clinically validated and is not available to patients or the public. We are working toward structured clinical evaluation with partners.

Does EDDS diagnose disease?

No. EDDS is being designed to organise results over time and flag meaningful change for review by medical experts. A flagged change is a signal to discuss, not a diagnosis.

What kind of data does EDDS work with?

Today the prototype works with laboratory results. Portable diagnostics and wearables are future directions, and smart sanitary systems are at the concept stage.

Who reviews EDDS findings?

Findings are designed to be reviewed by medical experts. The AI layer helps summarise the statistical analysis; it does not replace clinical judgement.

How can my institution run a pilot or collaborate?

Tell us about your organisation and what you'd like to explore through the partnership enquiry form. Our team reviews every submission and follows up to start the conversation.

Help take EDDS to clinical evaluation.

We're looking for clinical institutions to help move EDDS from prototype toward structured evaluation.

Partner on EDDS Back to EDDS overview