Homepage of Breast cancer Tumor type Diagnoser.
Providing inputs to make tumor type predication.
Data Visualization of Inputs vs Output for Doctor interpretation.
Result by 9 Classification Algorithms and Majority voting.
Training set of mammogram scans of affected ones.
Training set of mammogram scans of healthy ones.
Result of CNN on making prediction using trained CNN.
Breast cancer Diagnoser based on Mammogram Scans

Every Life Matters

Today, technology developed a lot so that we can prevent and treat any diseases easily. But sometimes, like in case of giving treatment for breast cancer, there is the failure rate is higher than the success rate. This is only because doctors are able to diagnose the disease only after the breast cancer-causing tumour growth has crossed the threshold. Doctors find difficulty in treating victims due to this. In many cases, doctors needed even more time to know which type of breast cancer tumour the patient is affected so that they could treat them accordingly. Hence, there is a need to come up with a system that can diagnose breast cancer and the two types of breast cancer tumours: Benign (non-cancerous) and Malignant (cancerous).

To solve these problems and save the lives of women who suffer from this dreadful disease, we come up with Machine Learning and Deep Learning model (CNN) algorithms which can predict the breast cancer tumour type from input values and breast cancer from a mammogram scan respectively with high accuracy and precision. The Machine Learning module is trained using 9 Classification algorithms that learns differently using supervised learning and makes a final prediction using the Majority vote of the individual model results. The Convolutional neural network is trained using test and train image set of affected and healthy mammogram scans.

Unlike traditional methods to diagnose it, the models can analyse thousands of patient details in a few minutes and can speed up the process of finding affected people, treat them in an early stage, silence the tumour and save them. The end-users of this project are the doctors who input patient details on the website and yields its results. Our team strongly believe that this idea would help the early detection of breast cancer in women and save lives.

Individual Contribution

All 3 of us have explored Web development frontend and Django framework, Machine Learning, Deep Learning separately as different domains. We have contributed equally in our expert stack to come up with this platform.

Hariharan -

I have contributed to developing the backend of the website with Django Framework and integrated frontend with backend code and ML algorithms. It was really challenging for me, as its the first time I am using Machine learning algorithms in Web Development.

Ashwath -

I have contributed to developing complete frontend part of this website as per our requirements using HTML, CSS, Javascript and Bootstrap. I also contributed to developing the CNN model for diagnosing breast cancer from mammogram scans.

Nikita Anand -

I have contributed to developing the nine different Machine Learning classification models for breast cancer tumour type prediction. It was challenging for it as I don't have prior knowledge in this domain and also implemented the majority voting for output to increase model accuracy.

Challenges we came across

All 3 of us have explored Web development frontend and Django framework, Machine Learning, Deep Learning separately as different domains. This is our first project implementing ideas from multiple domains in this one platform which would help in early detection of breast cancer and its type and help doctors to treat patients effectively at the earlier stage itself. Learning the requirements and implementing them in the given time was the main challenge we faced along with the errors during development. Sincere thanks to my Teammates: Nikita Anand and Ashwath for developing the idea along with me with their fullest effort. At the end of this hack, we are so glad that we came up with the best final AI expert to save lives. Thanks to The Virtual Hackathon 2020 team for organizing this hack which served as a motivation to come up with our ideas.