HBN Diagnostic Data Preprocessing

A Python preprocessing package for working with the Healthy Brain Network clinician consensus diagnostic data.

Features

The raw HBN dataset includes final clinician diagnostic data given across ten numbered diagnosis columns. The order of these diagnoses is not indicative of severity, chronology, or importance, and this format requires data manipulation to be useful for analysis. This package transforms the data to a wider format, organized by specific diagnoses or categories rather than diagnosis numbers. It also includes the option to filter by diagnostic certainty or time of diagnosis and creates a visualization of the diagnostic data. Option to either run interactively in the command line (recommended if not familiar with the dataset) or to install as a python package.

DX_01	DX_01_Cat	DX_01_Sub	DX_01_Time	DX_01_Confirmed	DX_01_Presum	DX_01_RC	DX_01_RuleOut	DX_02	DX_02_Cat	DX_02_Sub	DX_02_Time	DX_02_Confirmed	DX_02_Presum	DX_02_RC	DX_02_RuleOut	...
ADHD - Hyperactive Type	Neurodevelopmental Disorders	ADHD	1	1	0	0	0									...
Selective Mutism	Anxiety Disorders	Anxiety Disorders	1	1	0	0	0	Autism Spectrum Disorder	Autism Spectrum Disorder	Neurodevelopmental Disorders	1	0	0	0	1	...

↓

Neurodevelopmental_Disorders_CategoryPresent	ADHD_Hyperactive_Type_DiagnosisPresent	ADHD_Hyperactive_Type_Time	ADHD_Hyperactive_Type_Certainty	Autism_Spectrum_DiagnosisPresent	Autism_Spectrum_Time	Autism_Spectrum_Certainty	Anxiety_Disorders_CategoryPresent	Selective_Mutism_DiagnosisPresent	Selective_Mutism_Time	Selective_Mutism_Certainty
1	1	Current	Confirmed	0			0	0
1	0			1	Current	Rule-Out	1	1	Past	Confirmed

For more information on the HBN data, please see the HBN Data Portal

Installation

Install this package via :

pip install git+https://github.com/childmindresearch/hbn-ddp.git

Run interactively in the command line

hbnddp

Quick start

from hbnddp import HBNData

data = HBNData.create(input_path="path/to/data.csv")
processed_data = data.process(
    # change to output path if you want to save the results
    output_path="path/to/output.csv",
    # pivot on diagnoses, subcategories, categories or all
    by="diagnoses",
    # add certainty filter if desired
    certainty_filter=None,
    # set True to visualize results
    viz=True,
)

Notebook Example

Links or References