I am a chronic innovator who innovates for the betterment of the society. I look for challenges in the contemporary world and try to solve them using technology. Helping people through technology gives happiness to me. When I see somebody benefiting from my technology, my day is made. During the hackathon, while researching over pressing challenges being faced worldwide, we encountered the issue of breast cancer. "More women die due to breast cancer in India than anywhere in the world" As I have roots from India, this statement shocked me and actually inspired me to solve the issue of lack of affordable and accessible diagnostic tool for breast cancer. When I was researching over the topic and understanding the problem of breast cancer, the stories of the people fighting Breast Cancer and the one who actually lost their loved ones inspired me more and more to come up with a solution.
What it does
Working together, our team has come up with an elegant solution,Brognosis. To aid in the early detection and prevention of breast cancer, the Brognosis system uses an ultrasound glove to assist women in performing breast self-examinations. After the scan, an image of the inner parts of the breast appears on a smart phone and is instantly analyzed for cancerous lumps. Brognosis provides painless, radiation-free and delivers accurate results in less than five minutes for less than $1 per exam.It is estimated to be cheaper than any other diagnostic service available. Its easy to use and its more accurate. This is a market opportunity of a billion dollars in India and 8-9 billion dollars globally, annually and our team is going after it.Brognosis analyzes the images of the inner breast, which are obtained through the use of a custom wearable glove, which with the help of ultrasound sensors, does imaging of the inner breast. The images captured by the glove are then sent to the mobile phone application via Bluetooth. These images feed into an algorithm on the Brognosis mobile phone application which uses a convolutional neural network and machine learning to check for lumps and lesions in the Breast. The algorithm then outputs heat maps to highlight problem areas that clinicians can examine to make more-informed diagnoses, providing a more objective assessment than conventional tests. In addition, a remote diagnostic application facilitates sharing of images across to gain consultation from experts right-away via a cloud based tele-oncology platform.
Brognosis also instructs the user on how to preform the self examination by providing them with video tutorials and provides other relevant information towards treatment and healthy lifestyles against breast cancer.
Currently, there exists no portable, non-invasive and easily accessible screening tool for Breast cancer . However there are various diagnosing methods such as Mammography and MRI which are have limitations such as low specificity and sensitivity , accuracy and radiation respectively. The current system requires a personalized solution for quick and easy screening of theBreast Cancer. Our innovation includes the features of existing products along with its own advantages to revolutionize Breast cancer screening around the world. The following are the key competitive advantages of the Brognosis system: • Simple and easy to use. • Instantaneous results. • No additional bulky electronic equipment required. • Cheap and portable – ideal for rural and remote use. • Doctor remote consultation and feedback. • Secure image capture and Lesion detection algorithm. • Improved Diagnostic Accuracy. • Multiple Operating System support such as Android and web.
How I built it
We applied Human Centred design while developing Brognosis. After Finalizing on developing a better screening tool for Breast cancer, we started researching on the current scenario and the exisiting solutions and spent around 8 hours researching what the current solutions lack how we could develop something which provides better screening and at a cheaper rate. We came up with 6 ideas ranging from a simple selfie to screen breast cancer using image processing , to, making the use of tactile feedback to detect lumps and lesions in breast , and developing a piezoelectric mechanism based device for screening. We finally narrowed down to Brognosis due to its numerous advantages over the other ideas and because of the number of references we got justifying the technology used. The following are the links to the references: Breast cancer early detection methods for low and middle income countries, a review of the evidence: http://www.ncbi.nlm.nih.gov/pubmed/22289154?dopt=Abstract
Breast tumor detection using piezoelectric fingers: first clinical report.
We consulted a number of Doctors both from the United states( Developed nation) as well as India ( Developing nation). The product has been Co-created with the help from experts in the field of oncology. Majority of the time was spent on the research part as we wanted the product to be accurate, precise and user friendly.
Brognosis is a three tier system, the Brognosis glove, the Brognosis mobile phone application and the Brognosis dashboard. For the Brognosis glove, we used ultrasonic sensor running on Arduino nano microcontroller.The glove also makes the use of magnetometer and accelerometer to check for the correct positioning of the glove during the screening time. All the images collected by the glove are sent over to the mobile phone application via bluetooth. This happens with the help of HC-05 bluetooth module. The mobile phone application has been developed for android platform and sends the data to the Microsoft Azure cloud for processing. All the processing takes place on the cloud and the images are fed into our customized algorithm undergo processing and are screened for lumps and lesions. The results obtained are then displayed on the mobile phone application instantly. The mobile phone application has some features like find device, panic button and exercise tutorials. The data collected during the screening time is visualized using the R programming language and automated reports are generated using Azure automation.
Challenges I ran into
Finding sensors precise enough to detect very small lumps and lesions. Applying and building everything within 72 hours. Rapidly designing and prototyping all hardware and electronic parts. Supplying enough current to electronics. Simulating the mechanism involved in the lump detection in real time.
Accomplishments that I'm proud of
Coming up with a project which can really save lives and help deal with one of the biggest pressing challenges Developing something which helps all the sections of the society. Providing a great user experience in the device. Developing the desired product even when we were missing the most crucial hardware components. Taking the first step towards eliminating Breast Cancer.
What I learned
PennApps was full of learning. We learned about so many great APIs which the companies have to offer. Also learnt about the machine learning platform and Azure services provided by Microsoft. In total, I learnt about multiple disciplines involved in hacking ranging from Design to hardware development. Another important thing that I learnt was sticking to the deadline and developing the solution within the given span of time. Product management I would say. We love the fact that as part of PennApps, we got a chance to work on a solution that impacts a million lives.
What's next for Brognosis
We hope to continue working on this to be able to deploy this on a large-scale for hospitals as well as rural health clinics to save all the lives lost due to breast cancer . We will be applying for an FDA approval for this technology before we launch it. In the meantime we will be partnering with organisations dealing with breast cancer and start a pilot program with them after we fully develop the product.