GlucoPy

Python Toolbox for glycaemic signal processing

Table of Contents

• Main Features
• Dependencies
• Installation from source
• Documentation
• Examples
  • Create a Gframe object from a .csv file
  • Create a Gframe object from a .xslx file
  • Create a Gframe object from a pandas DataFrame
  • Create a Gframe object from GlucoPy's datasets
  • Calculate some metrics
  • Plots
    • Ambulatory Glucose Profile (AGP)
    • Frequency Histogram
    • MAGE Plot
• Collection of Python Signal Processing Library Repositories
• Public Datasets

Main Features

• Importing and exporting CGM signals in formats: .csv, .xslx, .xls
• Glycaemic signal in units: mg/dL (default), mmol/L
• Calculation of metrics like Mean, MODD, TIR, MAGE, BGI, Q-Score, etc.
• Automatic summary of the metrics.
• Generate Plots for analysis and visualization of the entire data or separated by day, some examples are Box Plot, Frequency Histogram, Trace Plot, Ambulatory Glucose Profile (AGP), etc.
• Very easy to use all metrics are calculated calling a method of the class Gframe, and the plots are generated using that same class.

Dependencies

• Pandas
• Numpy
• Plotly
• Scipy
• NeuroKit2
• Astropy
• Requests (NeuroKit2 dependency)

Installation from source

In the GlucoPy directory (same one where you found this file after cloning the repository), run:

pip install .

Documentation

The official documentation is hosted on github.io/GlucoPy/

Examples

All the Metrics are implemented as methods in the glucopy.Gframe class

Create a Gframe object from a .csv file

import glucopy as gp

# Read the file
gf = gp.read_csv('path/to/file.csv')

Create a Gframe object from a .xslx file

# Read the file
gf = gp.read_xlsx('path/to/file.xlsx')

Create a Gframe object from a pandas DataFrame

import pandas as pd

# Create a DataFrame
df = pd.read_csv('path/to/file.csv')

gf = gp.Gframe(df)

Create a Gframe object from GlucoPy's datasets

# Import the dataset
gf = gp.data("prueba_1")

Calculate some metrics

Check the documentation of Gframe for a complete list of metrics

# Mean
mean = gf.mean()

# MODD
modd = gf.modd()

# TIR
tir = gf.tir()

# MAGE
mage = gf.mage()

# BGI
lbgi = gf.lbgi()
hbgi = gf.hbgi()

Plots

All the plots use a Gframe object as first parameter. Check the documentation of glucopy.plot for a complete list of plots

Ambulatory Glucose Profile (AGP)

gp.plot.agp(gf)

Frequency Histogram

gp.plot.freq(gf)

MAGE Plot

gp.plot.mage(gf)

Collection of Python Signal Processing Library Repositories

• PyRadiomics
• StatsModels
• BioSPPy - Biosignal Processing in Python
• Type-2-Diabetes-Prediction-Using-Short-PPG-Signals-and-Physiological-Characteristics
• splearn: Python Signal Processing
• spm1d: One-Dimensional Statistical Parametric Mapping in Python and MATLAB
• NeuroKit2: The Python Toolbox for Neurophysiological Signal Processing
• PyGSP: Graph Signal Processing in Python
• rGV — Analysis of Continuous Glucose Monitor Data

Public Datasets

• Awesome-CGM

  This is a collection of links to publicly available continuous glucose monitoring (CGM) data.

• OhioT1DM

  The OhioT1DM dataset is available to researchers interested in improving the health and wellbeing of people with type 1 diabetes. It contains 8 weeks worth of data for each of 12 people with type 1 diabetes. a Data Use Agreement (DUA) is required.

• ShanghaiT1DM and ShanghaiT2DM

  ShanghaiT1DM and ShanghaiT2DM are publicly available Datasets for research purposes, they contain Type 1 (n = 12) and Type 2 (n = 100) diabetic patients in Shanghai, China.

• CG Map

  A reference resource for characterization of CGM data collected from more than 7,000 non-diabetic individuals, aged 40-70 years.