ระบบนี้เป็นเครื่องมือช่วยตัดสินใจทางคลินิก ไม่ใช้แทนดุลยพินิจของแพทย์·This system is a clinical decision support tool and does not replace clinician judgment. Research prototype — not a medical device, not for autonomous diagnosis.

Thesis Research Prototype

Thai Traditional Medicine —
Psoriasis Clinical Decision Support

An explainable, rule-based decision support prototype combining a curated TTM knowledge base with safety governance and clinician oversight. Designed for transparent reasoning, expert validation, and research demonstration.

ระบบนี้เป็นเครื่องมือช่วยตัดสินใจทางคลินิก ไม่ใช้แทนดุลยพินิจของแพทย์

Disclaimer: This system is a clinical decision support tool and does not replace clinician judgment. Research prototype only — not a medical device, not for autonomous diagnosis.

Workflow overview

From symptom intake to clinician-validated case report.

  1. 1Step 1
    Symptom intake
    Clinician enters standardized findings (erythema, scaling, pustules, BSA, etc.).
  2. 2Step 2
    Rule-based prediction
    Deterministic syndrome prediction with explicit rationale.
  3. 3Step 3
    Knowledge-base lookup
    Evidence panel pulls explainability traces, safety rules, therapy direction, consensus.
  4. 4Step 4
    Clinician review
    Clinician confirms, modifies, or overrides — override is highlighted and stored.
  5. 5Step 5
    Case report & validation
    Exportable case report; expert reviewers rate the case for thesis evaluation.

Key features

Structured symptom intake

Standardized inputs aligned to the project's ML feature schema.

Rule-based syndrome prediction

Transparent, deterministic logic — no black-box model in the prototype.

Knowledge-base evidence

Live lookup of explainability traces, therapeutic directions, safety rules, and consensus.

Clinical safety rules

Triggers urgent-referral flags before any therapeutic suggestion.

Clinician override intelligence

Final decision always rests with the clinician; overrides are logged and compared.

Expert validation mode

Structured 1–5 ratings across usefulness, clarity, safety, evidence, and usability.