API Endpoints

Diagnosis Prediction

POST /api/v1/diagnosis/predict

Predict disease diagnosis based on patient anamnesis.

Request Body:


{
  "age": 45,
  "gender": "male",
  "bodyLocation": "trunk",
  "symptoms": [
    {
      "type": "redness",
      "severity": 7,
      "duration": 14
    }
  ],
  "images": [
    {
      "url": "https://storage.blob.core.windows.net/...",
      "quality": 0.95
    }
  ]
}

Response:


{
  "diagnoses": [
    {
      "disease": "Psoriasis",
      "probability": 0.85,
      "confidence": "high"
    },
    {
      "disease": "Eczema",
      "probability": 0.12,
      "confidence": "medium"
    }
  ],
  "metadata": {
    "model_version": "v2.1.0",
    "inference_time_ms": 245
  }
}

Red Flags Detection

POST /api/v1/diagnosis/red_flags

Detect urgent medical conditions requiring immediate attention.

Request Body: Same as diagnosis prediction

Response:


{
  "hasRedFlags": true,
  "flags": [
    {
      "type": "rapid_growth",
      "severity": "high",
      "description": "Lesion shows rapid growth pattern"
    },
    {
      "type": "bleeding",
      "severity": "medium",
      "description": "Spontaneous bleeding reported"
    }
  ],
  "urgency": "high",
  "recommendation": "Immediate physician consultation required"
}

Image Quality Validation

POST /api/v1/image_quality/validate

Validate image quality and suitability for diagnosis.

Request Body:


{
  "imageUrl": "https://storage.blob.core.windows.net/...",
  "imageData": "base64_encoded_image_optional"
}

Response:


{
  "isValid": true,
  "score": 0.92,
  "issues": [],
  "feedback": {
    "blur": 0.05,
    "lighting": 0.95,
    "framing": 0.90
  }
}

Example with Issues:


{
  "isValid": false,
  "score": 0.45,
  "issues": [
    "Image is too blurry",
    "Poor lighting detected",
    "Lesion not properly framed"
  ],
  "feedback": {
    "blur": 0.65,
    "lighting": 0.30,
    "framing": 0.40
  }
}

Patient Triage

POST /api/v1/diagnosis/triage

Assess patient urgency and assign priority level.

Request Body: Same as diagnosis prediction

Response:


{
  "priority": "high",
  "urgencyScore": 8.5,
  "estimatedWaitTime": "24 hours",
  "recommendedAction": "physician_review_within_24h",
  "reasoning": [
    "Red flags detected",
    "Symptoms indicate potential urgent condition"
  ]
}

Adaptive Questionnaire

POST /api/v1/questionnaire/next

Get next question in adaptive questionnaire flow.

Request Body:


{
  "previousAnswers": [
    {
      "questionId": "chief_complaint",
      "answer": "rash"
    },
    {
      "questionId": "duration",
      "answer": "2_weeks"
    }
  ],
  "bodyLocation": "trunk"
}

Response:


{
  "questionId": "symptom_itching",
  "questionText": "Is the rash itchy?",
  "questionType": "yes_no",
  "options": [
    {
      "value": "yes",
      "label": "Yes"
    },
    {
      "value": "no",
      "label": "No"
    }
  ],
  "importance": "high",
  "isRequired": true
}

Questionnaire Complete:


{
  "questionId": null,
  "questionText": null,
  "isComplete": true,
  "completeness": 0.95,
  "readyForDiagnosis": true
}

Error Responses

All endpoints return standard error format:

400 Bad Request


{
  "error": "Validation error",
  "details": [
    {
      "field": "age",
      "message": "must be between 0 and 120"
    }
  ]
}

500 Internal Server Error


{
  "error": "Model inference failed",
  "message": "TensorFlow model error: ..."
}

503 Service Unavailable


{
  "error": "Service temporarily unavailable",
  "message": "Model loading in progress"
}