Dementia Prediction API

Using MLP model trained on balanced data

Use the /predict endpoint with a POST request to get dementia predictions for a single patient.

Use the /predict_batch endpoint with a POST request to get dementia predictions for multiple patients.

Use the /predict_time_series endpoint with a POST request to get a consolidated prediction from multiple observations of a single patient over time.

Use the /predict_time_series_aggregated endpoint with a POST request to get a prediction based on aggregated features from multiple observations of a single patient.

Single Patient Example:

    {
        "age": 70,
        "gender": "male",
        "educationyears": 12,
        "EF": 0.75,
        "PS": 0.6,
        "Global": 0.65,
        "diabetes": 0,
        "smoking": "never-smoker",
        "hypertension": "Yes",
        "hypercholesterolemia": "No"
    }

Batch Example:

    [
        {
            "age": 70,
            "gender": "male",
            "educationyears": 12,
            "EF": 0.75,
            "PS": 0.6,
            "Global": 0.65,
            "diabetes": 0,
            "smoking": "never-smoker",
            "hypertension": "Yes",
            "hypercholesterolemia": "No"
        },
        {
            "age": 65,
            "gender": "female",
            "educationyears": 16,
            "EF": 0.8,
            "PS": 0.7,
            "Global": 0.75,
            "diabetes": 1,
            "smoking": "ex-smoker",
            "hypertension": "No",
            "hypercholesterolemia": "Yes"
        }
    ]

Time Series Example:

    {
        "patient_id": "12345",
        "data_points": [
            {
                "timestamp": "2025-04-01T10:30:00",
                "age": 70,
                "gender": "male",
                "educationyears": 12,
                "EF": 0.75,
                "PS": 0.6,
                "Global": 0.65,
                "diabetes": 0,
                "smoking": "never-smoker",
                "hypertension": "Yes",
                "hypercholesterolemia": "No"
            },
            {
                "timestamp": "2025-04-08T11:15:00",
                "age": 70,
                "gender": "male",
                "educationyears": 12,
                "EF": 0.72,
                "PS": 0.58,
                "Global": 0.63,
                "diabetes": 0,
                "smoking": "never-smoker",
                "hypertension": "Yes",
                "hypercholesterolemia": "No"
            }
        ]
    }

Categorical Values: