Skip to main content

​
Decision Suite

All seven endpoints follow the same async pattern: submit a POST request, receive 202 Accepted with a job_id, then poll GET /api/v1/jobs/:id for the completed result.
These endpoints are for research, workflow assistance, and decision support only. They are not a substitute for licensed clinical judgment, radiology reads, or pharmacy verification. Validate outputs against clinical guidelines before acting on any result.
Enable step-level diagnostics by setting output.include_diagnostics: true. The completed job payload will include per-step timing and routing information. Use this for monitoring and support debugging, not for end-user UI.
Accepts patient vitals, history, and labs as JSON. Returns a risk level, time-sensitivity classification, and ranked recommended actions.

​
Request

curl -sS -X POST "$PI_BASE_URL/api/v1/health/risk-priority" \
  -H "Authorization: Bearer $PI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "type": "patient_data",
      "data": "{\"vitals\":{\"hr\":110,\"bp\":\"140/90\"},\"complaint\":\"chest pain\",\"age\":58}",
      "image_data": null
    },
    "context": { "locale": "en" },
    "output": { "format": "json", "include_diagnostics": false }
  }'

​
Body parameters

​
input.type
string
required
Must be "patient_data".
​
input.data
string
required
JSON-stringified patient data: vitals, complaint, history, labs. Max 120,000 characters.
​
input.image_data
string
Optional base64-encoded or data:-prefixed image for vision-assisted analysis (HuatuoGPT when configured). Max 8 MB.
​
context
object
Arbitrary JSON metadata (locale, facility, etc.). Max 16,000 characters serialized.
​
output.format
string
"report" or "json". Defaults to "json".
​
output.include_diagnostics
boolean
Include step-level timing in the completed job payload. Defaults to false.

​
Response fields (completed job)

​
risk_level
string
One of "critical", "high", "moderate", or "low".
​
time_sensitivity
string
One of "immediate", "hours", "days", or "weeks".
​
next_action
string
Primary recommended next action for the care team.
​
recommended_actions
array
Ranked list of specific action strings.
​
escalation_triggers
array
Conditions that should prompt immediate escalation.
​
priority_rank_rationale
string
Explanation of the risk classification.
​
risk_flags
array
Specific risk signals extracted from the patient data.
​
disclaimer
string
Mandatory disclaimer text.
Accepts a patient care timeline as JSON. Returns an adherence risk rating, missed follow-up items, predicted dropout window, and intervention recommendations.

​
Request

curl -sS -X POST "$PI_BASE_URL/api/v1/health/adherence" \
  -H "Authorization: Bearer $PI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "type": "patient_timeline",
      "data": "{\"appointments\":[{\"date\":\"2025-01-10\",\"attended\":false},{\"date\":\"2025-02-14\",\"attended\":true}],\"medications\":[{\"name\":\"metformin\",\"last_fill\":\"2024-12-01\"}]}",
      "notes": "Patient reports difficulty affording transport to clinic."
    },
    "output": { "format": "json" }
  }'

​
Body parameters

​
input.type
string
required
Must be "patient_timeline".
​
input.data
string
required
JSON-stringified care timeline (appointments, medications, lab visits). Must be valid JSON. Max 200,000 characters.
​
input.notes
string
Optional free-text clinical notes to supplement the timeline. Max 64,000 characters.

​
Response fields (completed job)

​
adherence_risk
string
One of "high", "moderate", or "low".
​
next_action
string
Recommended immediate intervention action.
​
missed_items
array
Array of missed appointments or medications. Each item includes type, description, optional date, and severity.
​
predicted_dropoff_window
string
Estimated timeframe when the patient is at highest dropout risk.
​
intervention_recommendations
array
Specific outreach or support actions to improve adherence.
​
engagement_score
number
Engagement proxy between 0 (disengaged) and 1 (fully engaged).
​
barriers_detected
array
Identified barriers to care (e.g. transport, cost, literacy).
​
escalation_triggers
array
Conditions that warrant urgent care team follow-up.
​
risk_flags
array
Specific adherence risk signals.
Accepts free-text clinical notes and returns structured entities: symptoms, conditions, medications, risk factors, and a chronological timeline. Useful for EHR pre-population and audit workflows.

​
Request

curl -sS -X POST "$PI_BASE_URL/api/v1/health/notes-structure" \
  -H "Authorization: Bearer $PI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "type": "clinical_notes",
      "data": "Patient is a 45yo female with a 3-week history of productive cough, low-grade fever, and unintentional weight loss of 4kg. Started on amoxicillin 500mg TID 1 week ago with partial improvement. PMHx: Type 2 DM (metformin 1g BD). No known allergies.",
      "format_hint": "soap"
    },
    "output": { "format": "json" }
  }'

​
Body parameters

​
input.type
string
required
Must be "clinical_notes".
​
input.data
string
required
The free-text clinical notes to structure. Max 64,000 characters.
​
input.format_hint
string
Optional hint about the note format, e.g. "soap", "discharge_summary". Max 64 characters.

​
Response fields (completed job)

​
summary
string
Concise narrative summary of the structured note.
​
symptoms
array
Extracted symptoms. Each item includes name, severity, onset, duration, and status.
​
conditions
array
Identified conditions. Each item includes name, icd_hint, status, and confidence.
​
medications
array
Medications mentioned. Each item includes name, dose, frequency, route, status, and start_date.
​
risk_factors
array
Identified risk factors. Each item includes factor and optional detail.
​
timeline
array
Chronological event list. Each item includes date, event, and optional detail.
Accepts a clinical situation as JSON and returns a reasoning-backed recommended action, alternatives, contraindications, and evidence references.

​
Request

curl -sS -X POST "$PI_BASE_URL/api/v1/health/decision-support" \
  -H "Authorization: Bearer $PI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "type": "clinical_query",
      "data": "{\"patient\":{\"age\":72,\"weight_kg\":68},\"presentation\":\"new-onset AF with rapid ventricular rate\",\"current_meds\":[\"warfarin\",\"digoxin\"],\"eGFR\":38}",
      "guidelines": "AHA/ACC 2023 AF guidelines"
    },
    "output": { "format": "json" }
  }'

​
Body parameters

​
input.type
string
required
Must be "clinical_query".
​
input.data
string
required
JSON-stringified clinical situation (patient demographics, presentation, current medications, relevant labs). Must be valid JSON. Max 120,000 characters.
​
input.structured_data
object
Optional additional structured fields to supplement data.
​
input.guidelines
string
Optional reference guidelines to ground the recommendation, e.g. "AHA/ACC 2023". Max 64,000 characters.

​
Response fields (completed job)

​
recommended_action
string
Primary recommended clinical action.
​
reasoning
string
Detailed reasoning behind the recommendation.
​
confidence
number
Confidence score between 0 and 1.
​
alternatives
array
Alternative actions. Each item includes action, rationale, and when_to_prefer.
​
contraindications
array
Specific contraindications relevant to the clinical situation.
​
evidence_references
array
Supporting evidence. Each item includes guideline, recommendation, and strength.
​
red_flags
array
Warning signs requiring immediate attention.
Accepts a medication list as JSON and returns interaction warnings, contraindications, duplicate therapy flags, dosing notes, and renal/hepatic adjustment guidance.

​
Request

curl -sS -X POST "$PI_BASE_URL/api/v1/health/medication-check" \
  -H "Authorization: Bearer $PI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "type": "medication_review",
      "data": "{\"medications\":[{\"name\":\"warfarin\",\"dose\":\"5mg\",\"frequency\":\"daily\"},{\"name\":\"aspirin\",\"dose\":\"100mg\",\"frequency\":\"daily\"},{\"name\":\"ibuprofen\",\"dose\":\"400mg\",\"frequency\":\"TID\"}],\"conditions\":[\"AF\",\"CKD stage 3\"],\"age\":72}"
    },
    "output": { "format": "json" }
  }'

​
Body parameters

​
input.type
string
required
Must be "medication_review".
​
input.data
string
required
JSON-stringified medication list including names, doses, frequencies, and relevant patient conditions. Must be valid JSON. Max 200,000 characters.

​
Response fields (completed job)

​
interactions
array
Drug-drug interactions. Each item includes drug_a, drug_b, severity, mechanism, and recommendation.
​
contraindications
array
Medication-condition contraindications. Each item includes medication, condition, severity, and recommendation.
​
duplicates
array
Duplicate therapy flags (same class or overlapping mechanism).
​
monitoring_recommendations
array
Specific lab or clinical monitoring items to order.
​
dosing_notes
array
Dose-specific guidance notes.
​
renal_adjustments
array
Renal dose adjustment recommendations.
​
hepatic_adjustments
array
Hepatic dose adjustment recommendations.
​
pregnancy_lactation_notes
array
Safety notes for pregnancy or lactation.
​
top_concerns
array
Highest-priority safety concerns for immediate attention.
​
summary
string
Narrative summary of the full medication review.
Accepts an imaging scan plus a modality label. Returns structured findings, a radiologist-style impression, a plain-language explanation, urgency classification, and optional measurements. Uses the MONAI sidecar for segmentation when configured.

​
Request

curl -sS -X POST "$PI_BASE_URL/api/v1/health/scan-analysis" \
  -H "Authorization: Bearer $PI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "type": "medical_scan",
      "data": "data:image/jpeg;base64,<...>",
      "modality": "chest_xray",
      "clinical_question": "Rule out pneumonia in immunocompromised patient"
    },
    "output": { "format": "json" }
  }'

​
Body parameters

​
input.type
string
required
Must be "medical_scan".
​
input.data
string
required
Base64-encoded scan image or data:-prefixed URL. Max 8 MB.
​
input.modality
string
required
Imaging modality, e.g. "chest_xray", "abdominal_ct", "brain_mri". Max 64 characters.
​
input.clinical_question
string
Optional clinical question to focus the analysis, e.g. "Rule out pulmonary embolism". Max 4,000 characters.

​
Response fields (completed job)

​
findings
array
Structured findings. Each item includes region, description, severity, and confidence.
​
impression
string
Radiologist-style overall impression.
​
plain_language_explanation
string
Patient-friendly explanation of the scan findings.
​
anomalies_detected
boolean
Whether any anomalies were identified.
​
urgency
string
One of "stat", "routine", or "non_urgent".
​
recommended_followup
string
Recommended next diagnostic or clinical step.
​
differential
array
Differential diagnoses. Each item includes condition and likelihood.
​
measurements
array
Quantitative measurements. Each item includes structure, value, unit, and normal_range.
Accepts a research query and optional dataset attachment. Returns analysis, statistical insights, methodology review, literature connections, and recommended next steps.

​
Request

curl -sS -X POST "$PI_BASE_URL/api/v1/health/research-assist" \
  -H "Authorization: Bearer $PI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "type": "research_query",
      "data": "{\"question\":\"Does early mobility in ICU patients reduce ventilator days?\",\"study_type\":\"RCT\",\"population\":\"mechanically ventilated adults\",\"outcome\":\"ventilator-free days at 28 days\"}",
      "dataset": null
    },
    "output": { "format": "json" }
  }'

​
Body parameters

​
input.type
string
required
Must be "research_query".
​
input.data
string
required
JSON-stringified research query. Include question, study_type, population, and outcome fields where applicable. Must be valid JSON. Max 100,000 characters.
​
input.dataset
string
Optional base64-encoded dataset attachment for analysis. Max 8 MB.

​
Response fields (completed job)

​
analysis_summary
string
Narrative summary of the research analysis.
​
statistical_insights
array
Statistical findings. Each item includes test, result, interpretation, and p_value.
​
methodology_review
object
Includes strengths, weaknesses, and suggestions arrays.
​
literature_connections
array
Related literature. Each item includes topic, summary, and citations.
​
recommended_next_steps
array
Suggested follow-on research or validation steps.
​
limitations
array
Identified limitations of the current analysis.