Analysis

Overview

Based on subtype defined by Dr. Michael Snyder, we have developed a set of programs and work process to identify what is your diabetic type 2 subtype and then offer what type of food you should eat.

The figure below shows the overall work process

  • Program A: Extract Data from LibreView
  • Program B1: Convert the extracted LibreView data to Daily CSV files
  • Program B2: Convert the extracted LibreView data to Yearly Report files
  • Program C1: Convert the Daily CSV files (from B1) to Subtype Report
  • Program D1: Generate AGP Reports using Daily CSV files (B1)
  • Program E1: Generates a comprehensive report (B1,B2,C1,D1) as html and PDF



Figure 1: Overall work-process


Program A: Extract Data from LibraView

The user per our instruction will download the LibreView data in CSV format and upload it into our server



Figure 2: Part A of the work-process - User Extract Data


The yearly data is shown below needs to be organized for further analysis

Program B: Convert the LibreView data to Daily CSV files

Our program will convert the user CSV into daily data format to be analyzed by the next program.



Figure 3: Part B of the work-process - Convert the data to daily data


Program C: Generate Subtype Reports

The program C functionality is divided as:

  • Use the daily average data scanned each 5 minutes
  • Predict zone that there is a peak or a meal is consumed
  • Identify the subtype based on the zone behavior

Based on your data, program will generate the following chart



Figure 3: Program C Generate Subtypes for each day


  • The top figure represents the CGM data for the 24 hours
  • The middle figure represents same data as figure 1 but the program detect zones that you had meals
  • The bottom figure uses the AI technology to assign Diabetes-Type 2 Subtypes based on zone based CGMProgram C - Overview of data

In the following figure the subtypes for your data is presented.

The histogram with the normal distribution superimposed is presented as well.



Figure 4: Program C Generate Subtypes and histogram


Program C - Overall Results

The original data is presented in the top figure



Figure 5: Program C Summarize the data


The bottom figure represent every plot of every day of the selected data


Program C - Subtype to Food


Figure 6: Program C Generate Overall subtypes


Based on your Diabetic Type 2 Subtype (DT2ST) the following provide the best meal to mange your diabetic.



Figure 7: Program C Generate Overall subtypes


Program D: Generate AGP reports

The following work process converts LibreView data to daily csv files.

The program analyze each daily data (per folder) and create the following results



Figure 8: Program D generate AGP


What is an AGP?

The Ambulatory Glucose Profile (AGP) condenses many days of continuous glucose monitoring (CGM) into a single 24-hour “modal day.”

It overlays the median (50th percentile) glucose and percentile bands—typically 25–75% (dark band) and 5–95% (light band)—so you can spot time-of-day patterns (overnight lows, post-meal spikes, dawn phenomenon) at a glance.