So guys here comes one of the most advanced data science medical projects that I have done till now in my whole Data Science journey. It is called HealthCure which is an all-in-one medical solution. We know the future is all about AI so here is my idea of bringing 7 disease detections under one platform using the power of AI. Although these results are perfect yet an experiment we did, which can prove to be a revolution in the coming years.
The main advantage of this project is that we can get the test results immediately at our home with a just few clicks.
7 disease detections:
- Covid-19 detection
- Brain Tumour detection
- Breast Cancer detection
- Alzheimer detection
- Diabetes detection
- Pneumonia detection
- Heart disease detection
Convolutional Neural Networks…
- A Convolutional Neural Network (ConvNet/CNN) is a Deep Learning algorithm that can take in an input image, assign importance (learnable weights and biases) to various aspects/objects in the image, and be able to differentiate one from the other.
- The pre-processing required in a ConvNet is much lower as compared to other classification algorithms. While in primitive methods filters are hand-engineered, with enough training, ConvNets have the ability to learn these filters/characteristics.
- The architecture of a ConvNet is analogous to that of the connectivity pattern of Neurons in the Human Brain and was inspired by the organization of the Visual Cortex.
- A ConvNet is able to successfully capture the Spatial and Temporal dependencies in an image through the application of relevant filters.
- The role of ConvNet is to reduce the images into a form that is easier to process, without losing features that are critical for getting a good prediction.
All Disease Detections…
- Used custom-made CNN architecture for this detection.
- The accuracy achieved was around 93%.
Brain Tumour detection
- Used VGG-16 for feature extraction.
- Used custom-made CNN ahead of CNN.
- The accuracy achieved was around 100% ?(just tested on 10 images).
Breast Cancer Detection
- Used Random Forest for this use case.
- The accuracy achieved was around 91.81%.
- Trained CNN architecture for this use case.
- The accuracy achieved was around 73.54%.
- Used Random Forest for this use case.
- The accuracy achieved was around 66.8%.
- Used custom CNN architecture for this use case.
- The accuracy achieved was around 83.17%.
Heart Disease Detection
- Used XGBoost for this use case.
- The accuracy achieved was around 86.96%.
- As time passes, we will be available with more and more data and we will try to make our models even more accurate by training on much more data.
- Also, we will be adding more disease detections that can be detected using X-ray scans or just by inputting simple numbers.
- We are also planning to add more features like if a person is found positive then our app will show him what precautions he needs to take and how he can cure himself.
- We will also be storing the detection records.
- So these are some future improvements/additions we plan to add.
How to run the project
Create a conda environment and install required libraries
conda create -n health python=3.9 conda activate health pip install opencv-python numpy tensorflow sklearn imutils flask xgboost
When you have successfully created the environment, installed the required libraries, and activated it, simply run the following command in the terminal.
NOTE – This was my college major project which I did with the help of my 1 more teammate Yash Kelkar.
About Training Code, PPT, and Project Report…
Also, I am getting a very high number of requests for Training Files, PPT and Project reports. So I have finally decided to share those with the people who are in real need. But as you know quality doesn’t come free, so you have to pay a minimalistic fee for all those resources.
Although I believe in providing the content for free(this whole website is an example), but if I provide that code and documents for free, everyone out there will be using that to resell that, and also that will be no justice to my and my team’s efforts.
If you are interested and want all the resources which include Training Code + App + PPT + Project Report, pay me a minimalistic fee of ₹500 on my UPI ID: abhi70420@paytm and send me a screenshot on my mail email@example.com for the files access.
Do let me know if there’s any query regarding data science projects or medical projects by contacting me via email or LinkedIn.
So this is all for this blog folks, thanks for reading it and I hope you are taking something with you after reading this and till the next time …
Read my previous post: OBJECT DETECTION USING SSD