Microsoft vows to 'solve' cancer within 10 years

25 September 2016
News
Microsofts new AI platform Hannover, published earlier in an article on the ICT&health website, is only part of the new program the company is rolling out in the battle against cancer. In the coming ten years Microsoft vows to use its considerable knowledge and technological prowess to eradicate cancer. How? To treat it like a computer virus.

The company has recruited some of the world’s best biologists, programmers and engineers who will approach cancer as a bug in a computer system. This summer Microsoft opened its first laboratory to test out the findings of its scientists who are creating huge maps of the internal workings of cell networks. They’re also creating a computer made from DNA which could live inside cells and look for faults in bodily networks, like cancer. If it spotted cancerous chances it would reboot the system and clear out the diseased cells.

Chalk and Cheese

“I think it’s a very natural thing for Microsoft to be looking at because we have tremendous expertise in computer science and what is going on in cancer is a computational problem. It’s not just an analogy, it’s a deep mathematical insight. Biology and computing are disciplines which seem like chalk and cheese but which have very deep connections on the most fundamental level, ” says Chris Bishop, laboratory director at Microsoft Research.

Microsoft is developing molecular computers built from DNA which you can compare with a doctor. It’ll destroy the cancer cells it finds. Head of the biological computation group at Microsoft, Andrew Philips, said: “It’s long term, but… I think it will be technically possible in five to 10 years time to put in a smart molecular system that can detect disease.”

They’ve already created software that imitates the healthy behavior of a cell, so they can compare it with a diseased one, to find the problem and work on fixing it. This “Bio Model Analyser software” is already being used to help researchers understand how to treat leukemia more effectively.

A century free of cancer

Dr Jasmin Fisher, senior researcher and an associate professor at Cambridge University, said: “If we are able to control and regulate cancer then it becomes like any chronic disease and then the problem is solved. I think for some of the cancers five years, but definitely within a decade. Then we will probably have a century free of cancer."

In the future smart devices will monitor health continuously. It will compare it to how the body should function and doing so, instantly detect problems. “My own personal vision is that in the morning you wake up, you check your email and at the same time all of our genetic data, our pulse, our sleep patterns, how much we exercised, will be fed into a computer which will check your state of well-being and tell you how prone you are to getting flu, or some other horrible thing,” Dr Fisher adds.

“In order to get there we need these kind of computer models which mimic and model the fundamental processes that are happening in our bodies. Under normal circumstances cells divide and they die and there is a certain balance, the problems start when that balance is broken and that’s how we get uncontrolled proliferation and tumours. If we could have all of that sitting on your personal computer and monitoring your health state then it will alert us when something is coming.”  

Scanning made easier

The process for radiologists nowadays to scan a tumour and create a 3D map on dozens of sections by hand takes up to four ours. Improved scanning to quickly map the size of tumours will decrease that time into minutes. Microsoft engineers have developed a programme which can delineate a tumour within minutes, meaning treatment can happen immediately. The programme can also show doctors how effective each treatment has been, so the dose can be altered depending on how much the tumour has been shrunk.

“Eyeballing works very well for diagnosing,” said Antonio Criminisi, a machine learning and computer vision expert who heads radiomics research in Microsoft’s Cambridge, UK, lab. Expert radiologists can look at an image – say a scan of someone’s brain – and be able to say in two seconds, ‘Yes, there’s a tumour. No, there isn’t a tumour. But delineating a tumour by hand is not very accurate.”

Microsofts system will be able to evaluate 3D scans pixel by pixel so the radiologist can exactly see how much the tumour has grown, shrunk or changed shape since the last scan. It will also provide information about things like tissue density, so it becomes more easy to determine whether we’re dealing with a cyst or a tumour. “Doing all of that by eye is pretty much impossible,” added Dr Criminisi.