Contact

Program case study

AIMed

AI for healthcare and Vietnamese traditional medicine.

How a public healthcare AI platform turns a medicinal-herb reference into lookup, interaction checking, and plant recognition — and what the live platform does and does not establish.

Public platform — live

  • Healthcare AI
Visit aimedvn.net All projects
Microscopic view of botanical leaf cells from medicinal plant research
Program artwork, not a screenshot of the software.

Research context

AIMed sits in the Healthcare AI area, where the lab studies how medical knowledge and intelligent assistance reach real use without losing the evidence behind them.

Its domain is Vietnamese traditional medicine: a large body of reference material where a wrong statement carries a real cost, so the platform keeps the assistant's job narrow and the reference visible.

Approach

  1. Keep the knowledge surface concrete: a searchable reference of 5,190+ medicinal herbs, with the herb entry as the result the visitor reads.
  2. Treat interaction checking as its own step, so a visitor sees a check rather than an assurance.
  3. Use AI plant recognition as an input aid that leads back to a herb entry rather than to an open-ended answer.

System architecture

Diagram — how a visitor reaches the AIMed reference. Read top to bottom: the platform is the entry point, the herb reference is the knowledge it serves, and the two AI-assisted checks sit on top of that reference.

  1. Public entry point

    AIMed

    The live public platform. This case study links to it rather than reproducing it.

  2. Reference

    Medicinal-herb knowledge base

    The reference the platform is built on.

    • 5,190+ medicinal herbs

  3. Check

    Interaction checking

    Reports whether selected herbs interact, as a distinct step the visitor can read.

    • Interaction checking

  4. Recognition

    AI plant recognition

    An AI-assisted identification step that leads from a plant to a herb entry.

    • AI plant recognition

Current implementation

Published status
Public platform — live

The status above is the published one, and the platform is reachable at its own address, linked from this page.

What this does not claim

  • No accuracy, coverage, or usage figures for the platform are published on this site, and this case study adds none.
  • The interaction check is described as a feature, not as clinical guidance.

Evidence

What the record establishes

The published record is the live public platform itself: herb lookup, interaction checking, and plant recognition are reachable at aimedvn.net.

What it does not establish

No clinical validation, outcome study, or benchmark result is published here, so none is claimed.

Roadmap

Every item below is planned, proposed, or under investigation. None of it is shipped, and this list is not a release schedule.

  1. Planned

    Keep the herb reference current

    Maintaining the reference is planned work. No release date, scope, or version is published for it.

  2. Proposed

    Publish an evaluation of plant recognition

    Proposed: a documented evaluation of the recognition step, so its limits can be read rather than assumed.

Collaboration

AIMed grows through clinical and domain partners: clinicians reviewing how interaction information is worded, and herb-knowledge contributors keeping the reference accurate.

Partner with us How the lab works

Work on a program with us.

Each program takes on partners where the research question meets real practice — clinical review for AIMed, agent runtime governance for ActantOS and Agent Platform, academic workflows for ScienzaOS, and document-heavy collections for AgenticOCR.

Discuss a collaboration
  • Clinical pilots
  • Research collaborations
  • Dataset partnerships
  • Student theses
  • Open-source contributions