This is the official WhiskeyDxEngine repository. The initial purpose of this application is to proof-of-concept an algorithm in which the author was told repeatedly is impossible to do programmatically. Python, PHP, Visual C#, Java, and, even JavaScript would probably be the ideal release quality version of this application in day-to-day usage (likely ASP - because the US Government tech standard is retarded) Sorry Microsoft -> Much Love Bill.
Also the language/stack choice was in part made as an insult to the Lansing Community College administration who told me I could not do this as it's not possible... so here we will not only do the "impossible" but do so in the most difficult of development pipelines - and yes the app will be written in parts of Inline Assembly. I also chose to forgo commonality of use in this concept because I do not want script kiddies involved -> you must REALLY know how to program to partake in the development - thus the GNU pipeline GDB symbols, G++, Make etc.
Our Veterans Deserve The Best and Brightest Minds.
One Veteran suicide or death due to disease/disablement is one to many. The sad reality is the "22 daily" statistic is reality. Coming from first-hand experience as a prior-service 68 Whiskey aka Combat Medic/Healthcare-Specialist... the Author of this application and algorithmic process - John McDonnell - understands some of the many unnecessary barriers Veterans face when placing disability claims for injuries/disease/ailments incurred during their Service.
As a prior-service Medic whose literally entered data into ALHTA and VistA (weird AltaVista) - John fully understands the medical documentation procedure and diagnosis pipeline. The fact is only medically-licensed professionals aptly named Doctors can legally Diagnose or Dx.
The time it takes to evaluate an existing legal diagnosis for the purpose of VA compensation is far too long. The current process yields results after an estimated 1-5 years or for us compsci nerds @ a Big-O rate of O(N!). A simple reduction in process speed may potentially prevent a Veteran suicide from financial constraints. Getting disability compensation delivered at a process pace of O(1) or near instantaneously yields nothing but positive results.
The faster we can accommodate Veterans needs effectively and efficiently the sooner we may just be able to save a life or prevent a Veteran suicide. The faster we can compensate the 7% of our population classified as Veterans who are in need of disability compensation - the better off we will be overall. The money need not be drawn from taxpayers – a new fund can be created as a way to offset the human debt factor of death from war... i.e. a fund or monetary source to distribute as needed upon a recognized and diagnosed service-related disease/illness/injury etc.
- Create a standardized Veteran/Dependent code – this is needed within the code to be able to reference the payment rating. IE Veteran Alone = DependentStatusCode = 1; Veteran and Spouse Only; DependentStatusCode = 2; (maybe try to continue with the semantics of it as well for calculations derived from the codes).
- Create a FING standardized patient ID format or agree to existing (FHIR) Etc. Agency-wide.
- Practice what the Government Preaches and FOLLOW STANDARDS.
- Keep It Simple Stupid and Try not to Reinvent the Wheel (SINGLE-USE for the Nerds).
- Assess the computing environment.
- Is this system CUDA Capable? If Yes → load CUDA MODULE (In the initial first run stage a CUDA system to run the processing engine on would be ideal for the massive amount of patients the system will need to analyze).
- Is this system multi-cored? (same concept as above – more CPU power faster results) count/threading additional core potential for tackling process heavy-operations.
- Create synthetic patient data source (Synthea) i.e. mock (VistA) EHR to run & test against (which the VA uses! for fsakes).
- Import the most current diagnostic codes from the WHO (ICD-X) i.e. ICD-10 (generate restful curl call)
- Aggregate, Import, and, Generate restful 38 CFR C (Probably already exists somewhere)
- Aggregate, Import, and, Cross-reference 38 CFR C with ICD-10 and benefit rate schedules.
- Ideally obtain real patient data and search for and apply the translated results to newly initiated claims.
-
Clone the repo:
git clone https://github.com/Renderlogic/WhiskeyDxEngine.git -
Compile with debug symbols utilizing the following command :
g++ -o bin/WhiskeyDxEngine main.cpp usr/LinuxSystemUtils.cpp lib/ArgRouter.cpp -ggdb -
Or just download the binary WhiskeyDxEngine and ./ that mofo
-
Create a SQL database and install the ICD table
[url]https://github.com/Renderlogic/WhiskeyDxEngine/raw/master/data/ICD-10-CM-SQL-2020.sql[/url]
- Ideally a Supercomputer such as NVIDIA DGX stations.
- Ideally CUDA capable NVIDIA GPU
- Ideally Multi-cored CPU
- Linux RPM-Variant
- MySQL with appropriate schema and data tables created (see scripts directory for automated bash scripts)
- Run Application via chron job weekly and upon user-invocation.
- Optional General EditorConfig C++ formatting standard to simply diffs and unify team development.
My IDE of choice for nearly every type of programming project is NetBeans. NetBeans EditorConfig isn't native so must be cloned/built and loaded if not already do so and can be cloned here: https://github.com/welovecoding/editorconfig-netbeans.git
- Source Code is obviously managed via git.
- GDB, GCC, or other Linux-compatible C++ development environment required as of now NetBeans with C/C++ activated and configured provides a GUI to GDB for easier debugging - I being a programming super-elitist super coder god of the programming gods do so from the command line. Actually, not doing it out of the said arrogance but merely to help reinforce my cmd skill.
All debugging is done x64_86 via command line GDB. Binary builds are compiled with ggdb symbology available. WDX has a directory in the Renderlogic DebuggingTheFramework Repository as well which features GDB logs vardumps etc - useful files to quickly help see runtime data and structures as needed... etc. For instance the array of processing unit counts from LinuxSystemUtils can be seen here along with a few other object property dumps used within the assignToHardwareStructure method.
[url]https://github.com/Renderlogic/DebuggingTheFramework/tree/master/WhiskeyDxEngine/gdb-output
The WhiskeyDxEngine Project just follows a traditional master/development build for now. Features can be branched and pull requested if additional developers join eventually. The project began with the first 23 commits made on a single master pipeline. At commit 23 a development branch was introduced as the code from here will start to get pretty complex and require much testing etc. As the project is really in its infancy stage expect development to heavily outpace master for some time -- i.e. many many commits ahead until routine pull requests are merged periodically.
- root directory of project
README.md (this file) main.cpp: the actual program entry point and then some. gitignore, editorconfig, Makefile, .dep.in
- usr (unix system resources)
LinuxSystemUlils class for managing system resources on Linux boxes.
- scripts
various automation scripts for generating relational tables, schemas, and, populating data in mysql.
- data
original data sources used for the project and various exported variants of said (sql, xml etc) use as needed.
- bin
binary objects directory
Please by all means help me develop this application especially if you have a medical development background. VistA EHR experienced-developers welcome. Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change. Please make sure to update tests as appropriate.
I'll list the stuff I've created for this project that has immense potential use for other projects.
Currently the only thing is over 70,000 disease codes and descriptions from ICD-10-CM 2020 I have pain stakely placed into an SQL table.
FREE ICD-10-CM SQL TABLE
It's a self-contained table you can add to your projects free of charge. Its in the data directory or can be downloaded directly here: https://github.com/Renderlogic/WhiskeyDxEngine/raw/master/data/ICD-10-CM-SQL-2020.sql
Copyright (C) 2019 GNU General Public License v3.0 Renderlogic, Inc,