Home Articles Big Data and machine learning as a force for good

Big Data and machine learning as a force for good

reasoning_machines

Before deciding to write this blog I have thought long and hard about whether I should even publish here on Compare the Cloud. As an organisation we promote and debate the benefits of cloud computing and the technology that supports the marketplace, but I have always believed that there is room for larger issues. Often our industry affects the way we live and so should be discussed openly.

Recently I was hospitalised with a serious condition with long-term side effects. I am not fishing for sympathy nor is it the focus of this blog, still it had an impact and that is what I want to share with you.

Lying in the hospital (perhaps having taken too much Morphine) I studied intensely the environment around me. I considered how computing advances are taking too long to reach the area that needs it most, our humanity.

My personal view of the world changed during my stay in hospital, as did my personal definition of what is humanity. I never truly understood helplessness or suffering until I witnessed a bowel cancer patient in incredible pain. Suffering he still bravely smiled as his family sat around him.

The guilt and helplessness I felt in not being able to assist this fellow patient has stayed with me. I have never experienced before and doubt I ever will again that emotive feeling. The hospital staff must find it difficult to cope.

Good luck in your recovery. Mate, you know who you are and even though I am a West Ham football fan when Southampton beat us at Upton Park I smiled inside at the thought of you cheering.

So where does my illness, the hospital stay and Big Data and machine learning fit together? They do, with a little explanation. Allow me to provide an explanation of what Big Data and machine learning is and how they can change our world.

Traditional data and computing systems
Computers were never really smart. They just seem so. A computer, from Alan Turing’s computational machine to the laptop in front of you, works on the basis that it is ‘under instruction’. Data is fed into the system in a structured format and commands are executed according to that instruction.

A computer never learns in the sense that we do. Its capability is restricted by the instructional process; by the data that is fed into the machine. Take away the instructions and we have a useless box decoration with a slice of silicon in the middle. Computers follow orders and never think outside the box.

Machine reasoning
Modern computers, the big amazing thinking machines like IBM’s Watson can show us something new. Yes, at the most basic level the computer is still simply fed its instructions, but moving a level up the next software layer is helping computers reach a greater understanding.

Computing systems today are able to review and compare unstructured data and extrapolate values. These learning machines are taking Big Data, the volume of disparate information, and predict relationships. This combination is producing computer systems with the ability to rationalise, in an almost human manner.

This ability has the power to advance humanity and redefine how we are able to diagnose, treat and manage patients. A thinking machine with rational insight based on Big Data comparisons, absorbing millions of patient notes, measuring clinical trials and aggregating statistical data, could integrate via sensors with the Internet of Everything (IoE).

Data sensors
Sensors are already part of our daily lives. They surround us and bind computers together. From simple vehicle diagnostic tools to warn us our tyres are running low or that we need to fill up to heart-rate monitors for fitness addicts sensors are all around.

With improving connectivity sensors can join a larger, central system to provide a constant stream of data to our big thinking machines. As anyone who has read Freakonomics might have already realised, comparing data from a wider pool of inputs can fundamentally change the understanding of a given issue. The application of sensor data in this way could allow us to diagnose, predict or even prevent ailments. The better we are measured and informed the better we might be able to help those now suffering in hospitals.

Technology practitioner
In patient care technology will never replace a real person. Let’s get this fact out early. Technology should not be used to replace those kind smiles or reassuring words, the genuine humanity, that comes from a nurse – or in my case the wonderful hospital cleaners making me cups of tea.

We should not use machine learning or Big Data as a means for achieving austerity. A cost-cutting human resources tool would be wrong and an abuse of technology.

Our reasoning machines, inter-connected by the IoE, could augment our current medical capabilities. As a tool it could be as important as the introduction of antibiotics or non-invasive surgical techniques.

For example Big Data and machine learning could personalise patient care. Hospital bed sensors could be used to compare data, gathered from motion-behaviour, heart rate and thermal imaging, to identify whether this patients movements indicate pain or potential suffering. The diagnosis could predict outcomes based on treatment and suggest the means to reduce discomfort.

Medical equipment could also share its data for comparison by even bigger machines. This would aggregate common complaints, national patterns and enable us to predict outcomes; and eventually prevent them.

We’re not talking about a utopian goal. The perfect system is never perfect. What I am suggesting is using the technology we already have available to us to improve the human condition. The same computing power that pushes terabytes of information around the globe could help us medically. As humanitarian efforts go these Big Data medical marvels offer more to boast about that aggregating visitor statistics on a website.

I’d like to see my Southampton supporting friend never need to suffer again in pain. Imagine being able to have sensors feeding data to learning machines – we might even be able to prescribe medication without the long wait for the doctor to make his way to the ward, finally. In the future we might even be able to simulate drugs trials, create new chemical combinations that could more effectively combat the spread of diseases such as Ebola in Africa.

The future is available today
My idea here is not all fantasy and it certainly is not science fiction. We are making great things happen in the cloud marketplace. Many of those technical leaps could easily be used in a medical context.

Dr John E. Kelly, Senior Vice President and Director of IBM Research, suggests that as “genomic research progresses and information becomes more available, we aim to make the process of analysis much more practical and accessible through cloud-based, cognitive systems like Watson. With this knowledge, doctors will be able to attack cancer and other devastating diseases with treatments that are tailored to the patient’s and disease’s own DNA profiles. This is a major transformation that can help improve the lives of millions of patients around the world.”

This is not a plug for IBM but Dr Kelly is absolutely right. Technology can bridge the huge amounts of medical and clinical data we are creating everyday. We don’t call it Big Data for nothing. Connecting all the pieces of our human condition, fighting the forces that shorten our mortal time on this earth, is an endeavour not so different as when we first connected our computers to a young Internet.

Billions are poured into research to bring make our connections faster, more efficient and more powerful (if not sometimes greener in our use of electricity). The pharmaceutical industry is investing billions in the pursuit of miracle drugs. These industries need to reap the rewards of good product, monetisation is what keeps the money flowing back into development, but both could equally contribute to the human effort.

It often said that pharma-companies should open source their developments, maybe in our drive to beat cancer, to cure and to heal we should open source our technology where possible.

Success often brings the cost of the cure, or the provision of services down. I believe it is possible for businesses to profitable and ethical. A debate of the commercial reality is probably bigger than the scope of this blog. Nevertheless as we move toward a new world of computational and technical advancement, remember to look inside yourself. Use these amazing technologies for the greater good.

I will finish this blog post with an apt quote from Mahatma Gandhi:

You must not lose faith in humanity. Humanity is an ocean; if a few drops of the ocean are dirty, the ocean does not become dirty.

Dedication
I would like to dedicate this post to the staff of The Royal Hampshire County Hospital, Winchester. Thank you for nursing me back to health and allowing me to look at my two young children through new eyes. I promise not to waste the opportunity you have given me.