Note : I had uploaded a wrong video in devpost due to last minute hurry burry. So here goes the link.
Link to the demo : https://youtu.be/1dRGT1ogyF0
This is the web app which can be a good mate for you to allow early recognitions to risk-survival diseases.
Inspired by the health problems I had witnessed across the under-represented groups who resides in regions which has no basic infrastructure.
This project is a web app which which can be a good mate for you to allow early recognitions to risk-survival diseases. Health can be studied on own through the user inputs with the predictions determined from your vitals data. This would provide you with the feasibility of being self independent when there are no provisions to doctors around in the remote places.
- DIABETES DATASET (https://www.kaggle.com/uciml/pima-indians-diabetes-database)
- BREAST CANCER DATASET (https://www.kaggle.com/uciml/breast-cancer-wisconsin-data)
- HEART DISEASE DATASET (https://www.kaggle.com/ronitf/heart-disease-uci)
- CHRONIC KIDNEY DISEASE DATASET (https://www.kaggle.com/mansoordaku/ckdisease)
- LIVER DATASET (https://www.kaggle.com/uciml/indian-liver-patient-records)
- MALARIA DATASET (https://www.kaggle.com/iarunava/cell-images-for-detecting-malaria)
- PNEUMONIA DATASET (https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia)
I had built prediction models for diseases that are most prevalent in MENA regions and used html and css for front end development. Used the datasets from kaggles, worked on model and extracted .pkl files, deployed .pkl files on flask and created user interfaces using html.
- DIABETES (98.25%)
- BREAST CANCER (98.25%)
- HEART DISEASE (85.25%)
- KIDNEY DISEASE (99%)
- LIVER DISEASE (78%)
- MALARIA (96%)
- PNEUMONIA (95%)
The issue which needs deserved attention is the health section. Especially in MENA regions, people had gone a toll to various diseases due to lack of proper facilities and no one to care for. Strong solution is needed to come out of this crisis and stay fit.
This App would be a best fit solution to protect people who are in despair and has no access to health infrastructure especially in remote regions. It would also help in critical situations when they are confined from movements. Having this pocket doctor in hand can benefit them in knowing their one's health. People would gain confidence in knowing their next step and get relaxed to overcome regular doctor checkups.
There are quite many challenges to list out. First being a sole member with no team, then working on each and every prediction and provide efficiency out of it using hypertuning. Then deploying the pickle files using flask. It's my first time to work on full stack developement and that too from mere scratch of the surface. It was totally a new journey with many struggles and hiccups along the way. At moment, I felt it's never ending. After all this, again as a beginner in html who just started attending online class a day before. I also didn't know about the existence of proc file. I had given my best to implement what I had learnt just in the span of 24hours.
I'm extremely proud of what I had achieved. I was a one woman army here. Getting my hands on experience with all the new terminologies and step by step procedures, it was really enrinching. Wish I had more time to make this even more better. I could have finessed my learning journey as well
I had learnt about flask deployment, proc file, html integration with ML models, writing html scripts, integrating them with backend and the overall workflow of the project.
Due to less span of time, I couldn't explore more deeper into designing and front end part. I wanted to add a good design to front end and make it attractive. For now, it gets job done but would like to give good user designing experience to this app in future and make it more functional by adding more predictions that showcases versatility.
Prathima Kadari