Glycemic-Patterns
Glycemic Patterns is an application that recognizes patterns in glucose data collected by blood sugar monitoring devices. These patterns are composed by a set of rules that express in which situations the patient may be in risk of suffering a condition (Hyperglycemia, severe hyperglycemia and hypoglycemia)
Description
The main purpose of Glycemic Patterns application is to help physicians and diabetic patient to identify risk situations or disorders in the amount of glucose in the blood. Each day is divided in blocks determined by the meals indicated by the patient in the device. These blocks are time periods that start 2 hours before the meal and finish 4 hours after and may be overlapped with other blocks. The risk situation are identified in the next block with data of the current block, the previous block and the previous day.
The application generate a decision tree for each condition (Hyperglycemia, severe hyperglycemia and hypoglycemia) and recognize the patterns that are human-readable and presented in a report. This report contains the decision tree, the patterns and useful information that may be helpful to physicians and patients to identify potential risk situations.
Features
- Compatibility with FreeStyle devices
- Report generation in PDF and HTML format
- Manual feature selection to use in decision trees: Mean, Std, Max, Min, MAGE (Glycemic variability)
- Multilanguage support (English and spanish)
Installation
To install Glycemic-Patterns, both in Windows and Linux, it is necessary to use Anaconda. After installing, open the Anaconda Prompt (Windows) or console (Linux) and create an environment using environment.yml. Then, install the package using setup.py file:
$ conda-env create -f environment.yml --force $ source activate glycemic-patterns (glycemic-patterns)$ python setup.py install
Usage
To generate a report using the command line:
report.py [-h] [--version] [-o [output path]] [-f [FORMAT]] [-v] [-vv] FILEPATH [FILEPATH ...]
To generate a report using Python prompt:
from glycemic_patterns.model import Model
trees = Model('<file_path>')
trees.fit()
trees.generate_report(output_path='<output_path'>, format='pdf')