Imperial College crowdsources smartphone network to fight cancer

Imperial College crowdsources smartphone network to fight cancer

Harnessing the processing power of smartphones could be a useful way to crunch clinical data, as healthcare continues to work towards the goal of precision medicine. Jessica Twentyman reports. 

Computer processing power is a vital ingredient in the journey towards precision medicine – the development of tailored treatments for patients. But what if the computing resources needed to crunch vast volumes of clinical data were found not in sophisticated server clusters or the public cloud, but in our pockets?

That’s the thinking behind a project announced this week by researchers at Imperial College London and the Vodafone Foundation, the arm of the mobile operator that supports projects focused on delivering public benefit. They’re planning to recruit smartphone users and their devices as volunteers in the fight against cancer.

It’s an intriguing use of crowdsourcing: the partners propose to get members of the public to download an app, developed by the Vodafone Foundation, called DreamLab. This enables smartphones plugged in overnight to join forces and work together on data analysis – similar to the distributed-resource approach taken by SETI and other scientific research programmes.

The app is free to download and use for Vodafone customers, so the processing won’t eat into their monthly allowances. It can also be used across rival networks, if users choose to donate some of their monthly data allowance, or agree to connect via Wi-Fi.

Tailored treatments

The aim of the DRUGS (Drug Repositioning Using Grids of Smartphones) project, led by Dr Kirill Veselkov of the Department of Surgery & Cancer at Imperial College, London, is to identify more effective combinations of existing medicines. That’s important, because while traditional treatments are typically dictated by the type of cancer a patient has, this research aims to use genetic profiles to find the best options for individual patients.

“We are currently generating huge volumes of health data around the world every day, but just a fraction of this is being put to use,” said Dr Veselkov. “By harnessing the processing power of thousands of smartphones, we can tap into this invaluable resource and look for clues in the datasets. Ultimately, this could help us make better use of existing drugs and find more effective combinations of drugs tailored to patients, thereby improving treatments.”

According to Vodafone, the combined might of smartphones is not to be dismissed. The company reckons that a desktop computer with an eight-core processor running 24 hours a day would take 300 years to process the data. But a network of 100,000 smartphones running six hours a night could finish the job in just three months. For example, a new device could solve up to 24,000 problems in six hours if fully charged and plugged in.

Smartphone grids

This isn’t the first use of DreamLab for this kind of work. The app has already been successfully trialled at the Garvan Institute of Medical Research in Sydney, Australia.

In February, the Institute announced that it had completed the first of two cancer research projects running on DreamLab, Project Decode. Like the project at Imperial, this aims to understand cancer based on an individual patient’s DNA profile. 

More than 122,000 DreamLab users completed Project Decode in half the time it would have taken the computing infrastructure at the Garvan Institute. The DreamLab community crunched 25 million research calculations – in triplicate to validate the results.

“Project Decode reaching 100 percent completion on DreamLab is an exciting milestone,” said Dr Warren Kaplan, chief of informatics at the Garvan Institute. “We are excited to already be seeing some interesting patterns that we will explore further. In June, we plan to report our findings in a research publication.”

Internet of Business says

Smartphone grids as a source of distributed computing could be put to all sorts of uses. Today’s devices, after all, are often as powerful as supercomputers once were, according to a team of researchers at the University of Helsinki.

In a project called Ubispark, they’re taking the idea further still, and exploring the use not just of smartphones, but also of other connected devices, such as smart TVs, to form local computing clusters for the purposes of large-scale data processing.

Of course, for all of this to work, they need volunteers who are willing to have their smart devices used by the project. With the projects at Imperial College and the Garvan Institute, however, the focus on cancer research will be a powerful incentive for people to get involved – either those who have experienced cancer themselves or have seen it affect friends or family.