Features

Kenyan invents app that could beat breast cancer

John Mwenda’s groundbreaking innovation, can predict a person’s chances of developing cancer using the environment and genetics

John Mwenda’s intellectual capabilities were first realised when he was in primary school. He was registered for the Kenya Certificate of Primary Education exam when he was in Class Seven. When he sat for the exam he scored 390 out of the possible 500.

“My teachers insisted that I needed a greater challenge because I was consistently a top performer. But I owe it to my parents; they played a great role in moulding us this way by exposing us to books at a young age,” he says.

He sat for his Form Four examination at Chuka Boys High School in 2008 and joined campus at Kenya Methodist University to pursue a degree in Mathematics and Computer Science in 2011.

And when he graduated in 2016, he decided to venture into entrepreneurship rather than look for a job.

Two years down the line, Mwenda has been working in the Artificial Intelligence and Data Science field coming up with prediction models for presidential elections in the US, Germany, France, Rwanda and the last concluded elections in Kenya.  But the most phenomenal of all is developing Cleo, a mobile application that promises to extend life of breast cancer patients by providing more accurate and timely information on breast cancer risk and control than any other medical information available. The idea was borne out of a personal experience. He had seen a friend’s aunty, Pauline suffer from the disease and hospitals not helping much even after spending millions in top-notch healthcare. She had stage three breast cancer and it was spreading fast.  He decided to come up with an app to predict where the cancer would go, when and how it would get there.

“We were successful in predicting the areas the cancer would move to. We applied Artificial Intelligence technology on data sets on breast cancer collected over a 30-year period to determine patterns and use the results to predict cancer spread based on a person’s characteristics. Unfortunately, time had run out for her. She passed on a few weeks after we came up with the solution,” he says.

But Mwenda’s journey to give cancer patients a second chance at life had just begun. “It was a great thing that we were able to accurately predict the metastasising of breast cancer; what if though, we could predict the chances of a person getting cancer in the first place over the course of their lifetime, thereby having them take preventive measures when they matter the most? This way, we would, hopefully, save the lives of many more men and women,” he says.

Mwenda, and his team got into research. They found a way to use technology to understand the genetics and the environment of a person and use that information to predict their chance of getting breast cancer.  The breast cancer detector app is more accurate than a mammogram or a medical expert and is able predict chances of a person developing breast cancer and helps you prevent it.  It updates itself automatically as the lifestyle and environmental factors change, thereby always remaining accurate.

So how does it work? The app is divided into three parts; genetics, environmental and lifestyle. Only the first part requires user input. “For us to figure out your genetic composition, we ask you a couple of questions and use complex genetic and mathematical models to figure out your genetic predisposition to cancer. This process is 82 per cent accurate.  Next, they analyse data from your mobile phone from GPS and accelerometer to figure out environments you are mostly exposed to and if those environments expose to breast cancer.

They analyse your personality based on your mobile phone data to try and understand how your lifestyle influences your chance of getting breast cancer. This is akin to the way mobile money lending apps analyse your data to determine whether to lend you money. This process is 90 per cent accurate.

Show More

Related Articles