bhargavaganti/medico

AI-powered medical terms detection tool.

Medico

Medico:- Medico is a medical terms detection system via your Voice.

Current stable versions

Installation

Run these commands on Terminal/CMD/bash dpending on your OS

• run git clone https://github.com/pranayjoshi/Medico. This will clone the Repo to your local system.
• than run cd Medico .This will set Medico as your present directory.

Installing Dependencies

1. Windows Users

• BASH Installed
  • Run the following command to install the dependencies required:- ./install.sh
• BASH not Installed
  • Just run the setup.py by pip install -r requirements.txt or pip3 install -r requirements.txt depending on the pip version.
  • Install pyaudio by pip install install/PyAudio-0.2.11-cp37-cp37m-win_amd64.whl.

2. Mac OSX USers

• Run the following command to install the dependencies required:- ./mac_install.sh

3. Debian based Linux Users( Ubuntu, Mint etc..)

• Run the following command to install the dependencies required:- ./debian_install.sh

4. For NIX Users

• Run the following command to install the dependencies required:- ./install.sh

Dependencies for Medico has succesfully installed on your system.

Running Files

1. Python Scripts/Commands

• To run the software use:- python3 run.py or python run.py based on your python version.
• To run the tests use:- python3 RunTests.py or python RunTests.py based on your python version.
• Start Conversing after you hear I am ready for your command.

2. Bash Scripts/Commands

• To run the software use:- ./start.sh.
• To run the software use:- ./runtest.sh.
• Start Conversing after you hear I am ready for your command.

Description

• A software that recognize medical terms
• Dataset used:- Snomed international(Sample)

Working.

• Takes the Medical Conversation(Mainly between Doctor and Patient) as the input.
• Use that Voice Conversation and Convert it to text using Speech To Text.
• Than the fetch_recent.py takes the file containg all the conversation and returns the latest conversation.
• After that the Punctuator Model takes the latest conversation and does the Magic(adds the punctuations) to the conversation.
• Than we use the Punctuated Conversation and the whole conversation/document gets divided into particular sentences, by the Sentence Tokenizer Model.
• After that we use the Tokenized Sentence, and check them One by One wheter they are Medical Sentences(Contains Medical Terms) or not.
• If they are considered Medical Statements than:-
  • The Medical Term Detection Model starts. It Further divides those Medical Sentences into 100000+ Categories.
  • After that a Printer Function prints all the necessary details.
• Otherwise the sentence is Skipped.
• At Last a Final Report is printed, Displaying all the Medical Terms found in the Whole Conversation.

Thank you