Vish1811/SEPSIS_DETECTION

Repository contains solution to problem statement on which our team is going to work in SIH2020

SEPSIS DTECTION USING MACHINE LEARNING

This repository contains the Sepsis Detection project(ABHIGYATAA) source code. Abhigayataa is web solution made for the early dection of SEPSIS.It is the project TEAM - HEISENBUGS have done for SIH 2019(Smart India Hackathon, 2019).

# LIVE Visit Live Site

Overview

1. The goal of this Challenge is the early detection of sepsis using physiological data.
2. Develop a digital solution that uses artificial intelligence to detect patient deterioration before it’s too late and trigger notifications to clinicians and care teams.
3. Train healthcare providers and front-line staff to quickly recognize symptoms to identify sepsis and treat patients at the earliest.
4. Educate preventing infections in health care settings and in the community so that infections that can lead to sepsis can be stopped before they happen.
5. Provide guidelines to follow infection control requirements (e.g., Hand hygiene) and ensure one to receive recommended vaccines

Table of contents

• Introduction
• Symptoms
• Causes
• Effects
• TaskToDo
  • Awareness
  • Detection
  • Notification
  • Documentation

Detail

• Awareness
  • gif
  • why awareness
  • TechnologyUsed
• Detection
  • Approach
  • flowchart
  • graphs
  • TechnologyUsed
• Notification
  • flowchart
  • TechnologyUsed
• Documentation
• FutureScope
• Contributing
• References
• License
• Special thanks

Introduction

According to World Health Organisation (WHO) , “Sepsis is a potentially life-threatening organ dysfunction caused by a dysregulated host response to infection. Any type of infectious pathogen can potentially cause sepsis.”

Symptoms

General Symptoms

1. A fever above 101ºF (38ºC) or a temperature below 96.8ºF (36ºC)
2. Heart rate higher than 90 beats per minute
3. Breathing rate higher than 20 breaths per minute
4. Probable or confirmed infection

Causes

1. Kidney infection
2. Abdominal infection
3. Meningitis
4. Cellulitis
5. Pneumonia

Effects

1. Adults 65 and older
2. Infants and children
3. Pregnant women or women who have just given birth, had a miscarriage or had an abortion
4. People suffering from a severe burn or physical trauma.
5. Premature prolonged rupture of the fluid-filled membranes that surround the fetus
6. Infection in the mother (such as chorioamnionitis)
7. Malnutrition due to social isolation, functional limitations, poor or restricted diets, chronic disease, dementia, depression, poor dentition, polypharmacy, and alcohol or substance abuse.
8. Endocrine deficiency like hypoadrenalism, hypothyroidism and hypogonadism.

Awareness

As Awareness is a very important aspect for preventing the spread of any disease.We aim to spread awareness in two ways:

1. Through a Website
2. Through a Telegram Bot

WEBSITE

---

To get more details about SEPSIS visit our website ABHIGYATAA

---

Features of website

1. Covers all the important aspects regarding Prevention and Symptoms
2. ML prediction Model is integrated with it that will do the task of early Detection of SEPSIS using the SOFA Score.
3. Training Module for the employee of the Health care institutions.

Technology USED

HTML5 , CSS , BOOTSRAP4 , JAVASCRIPT , JQUERY , PYTHON in combination with FLASK

TELEGRAM BOT

---

To get more details about SEPSIS visit our Telegram Bot @sepsissupport

---

Why Telegram BOT?

1. A Telegram Chatbot will be included as tool to spread Awareness.
2. Real-Time Assistance along with Better Interaction will be provided once the chatbot gets trained enough.
3. More people can be handled 24/7 service will be provided and only necessary queries will be forwarded to a real person.

Technology USED

Python , PHP , Telegram API

Detection

This is how our System Works.

flowcharts

Notification

1. Our website provide a notification section to express alerts if the prediction for sepsis label is found positive.
2. a Separate user-interface will be designed using python library for triggering Alerts to the Hospital Staff.

Documentation

1. We created a well docomented code with stepwise pathway for easy understanding.
2. PPT for the Presentation of the Idea.

Future Scope

1. More efficient Algorithms can be used for prediction of SEPSIS.

Contributing

Vishnu Kumar Sonu Verma Aditi Shreya Piyush Agarwal Namrata Gupta Aanchal Singh

Disclaimer: All arrays mentioned in this section must exist for the examples to work.

References

1. https://www.physionet.org/content/challenge-2019/1.0.0/
2. https://www.datacamp.com/community/tutorials/decision-tree-classification-python
3. https://towardsdatascience.com/using-bagging-and-boosting-to-improve-classification-tree-accuracy-6d3bb6c95e5b
4. https://towardsdatascience.com/early-detection-of-sepsis-using-physiological-data-78d5f31fab9d
5. https://iopscience.iop.org/article/10.1088/1757-899X/428/1/012004

License

Special Thanks

To our Seniors who provided us guidance at necesscary point of time.