Skip to content

Develop an ML algorithm for early detection of the risk of survival of the Preterm newborn babies through Pain-Scale Assessment.

License

Notifications You must be signed in to change notification settings

Tej-git-212/AI-ML-for-Newborn-Babies-in-Healthcare

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

32 Commits
 
 
 
 
 
 

Repository files navigation

AI-ML-for-Newborn-Babies-in-Healthcare

Shouting out to all of you who are interested in open-source contribution in the field of Healthcare using modern technologies like Artificial Intelligence(AI), Machine Learning(ML), Deep Learning(DL) and Computer Vision(OpenCV) and by using the most famous and easy programming language - Python, this is the place where you can contribute, learn and get a wonderful experience to showcase your skills in the future !!

Hey guys👋 myself Samarth, I'm the Program Admin at DevIncept for the 30-day open-source contribution program(DCP2021).

I am super excited to share with you all about this project which helps to save lives of many newborn babies which are at a very high-risk, especially the ones that are born before 9 months of pregnancy (also called as Preterm babies).

Don't worry. This is just a kind of basic to intermediate level project and you don't have to think much about it as whether it could be a lot tedious or you would not be able to do it in 30 days. You can definitely contribute a lot, learn a lot during these 30-days and even if you are interested in contributing more for this project after the next 30 days, feel free to reach out to me anytime. If you are really interested in Healthcare, then selecting this project for contributing is your best choice.

Saving a life of a baby just by writing some codes and implementing ML algorithms?! Is it ever possible🤔??

Yes, it is possible😍.

This project is capable of saving a newborn baby's life which is at high-risk of survival.

But how??

You will be accessing the condition of the baby at regular intervals using Pain Scale Assessment technique. To build the pain-scale model, we use Machine Learning algorithms to train the raw image/video dataset, and for obtaining frames of images from our model i.e., the baby condition's video dataset, we use OpenCV technique. Hence by implementing these techniques, we can help doctors by providing the continous monitoring report on the baby's health condition and help them to save the lives of those babies which are at a very high-risk of survival.

The main idea of this project is to develop a Machine Learning algorithm for classifying the facial expressions of Newborn babies in the Hospital as per the Pain Scale Assessment on a scale of 10. It involves 5 expressions mainly, they are:

  1. Neutral
  2. Pain due to Hunger
  3. Pain due to Discomfort
  4. Pain due to physical body pains/internal organ pains
  5. Pain due to Attention-seeking

These 5 pains can be classified under pain scale into 3 different categories. They are:

  1. Mild/No Pain (Scale 0 to 2)
  2. Moderate Pain (Scale 3 to 6)
  3. Severe Pain (Scale 7 to 10)

Pain Scale

Based on this classification algorithm, and based on our Data Analysis obtained through the babies Facial expressions, the doctors can get to know the Low-risk and High-risk behaviour from the babies and focus on giving required attention and proper treatment to the babies that are under high-risk severity and can help their life from disasters.

Tech Stack required:

  • Python software installed --> You can install it from here if not installed previously - https://www.python.org/
  • Any python supporting IDE(Jupyter Notebook/Pycharm)
  • Tensorflow --> pip install tensorflow
  • OpenCV --> pip install opencv-python
  • Numpy --> pip install numpy

For any queries, you can always contact me via:

I wish to see you all soon.

About

Develop an ML algorithm for early detection of the risk of survival of the Preterm newborn babies through Pain-Scale Assessment.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published