Google has been continually improving its artificial intelligence (AI) portfolio, and considering “search” still makes up 80% of Google’s revenues it would be logical to assume that the firm is improving its AI portfolio to improve its search capabilities. However when you consider the development of AI you realise that Google is actually using search to make its AI better.
This was always the plan for the Google engine. Every search and subsequent click on a given result is training Google AI to know the “right” result for any given query. Then when you consider the 3.5 billion searches taking place on Google everyday, you begin to get an idea of the potential speed of their AI development. So if your AI is a search engine you need to feed it with more real searches to make it better. But what if your AI is a doctor?
The healthcare industry is widely touted to be the next proving-ground for AI. In fact, AI has already begun demonstrating some highly advanced clinical competence in diagnosing skin cancer and diabetic retinopathy, for example. In 2017 we reported on the potential of smart hospitals in supporting healthcare professionals and improving patient care, but in 2018 AI is set to take the smart hospital a huge step forward.
Like a doctor, an AI system can look at, listen to, or even smell the patient, it can also run tests and analyze historical information on the patient, or scan public records for matching symptoms. However, despite the same information, an AI system can identify things a doctor never could. Researchers have developed an AI system that can diagnose Parkinson’s Disease by analysing patient voice recordings. Another can effectively smell cancer by identifying certain molecules in the patient’s breath.
The potential early diagnosis offered by AI forms the basis of our future healthcare system. From there we have already positioned ourselves to increase the effectiveness of treatment, and AI would continue to develop better ways to support recovery throughout the healthcare process. However, in order to reach this goal we need more data, or more specifically the companies with the AI need more access to the data held by healthcare companies – this data is highly legislated for privacy and often of poor quality in digital terms.
If AI is to revolutionize healthcare in these ways, the IT sector will need to get more involved in the healthcare process. In recent years we have begun to see increasing healthcare investments by some of the world’s biggest AI focused companies, such as Google, Apple, and Amazon, all vying to dominate healthcare in the same way Google dominates search – by having the best data and starting to crunch it before the competition.
“There is a scramble for data. This scramble is made even more frantic by first-mover advantage. Google maintains its leadership in search due to its user volume and consequently the amount of data it has to improve and perfect its algorithm,” says John Loder, Head of Strategy at Nesta’s Health Lab. “Similarly, successful clinical AI will be fed data to analyse, which will also help it improve. The company that first establishes a strong market position will be hard to dislodge,” he added.
IT is now rife with healthcare partnerships. IBM has been training its Watson AI system alongside a team of physicians and analysts at Memorial Sloan Kettering Cancer Center. Google owned, British artificial intelligence company Deepmind has partnered with Moorfields Hospital for better diagnosis of eye conditions and diseases. While Amazon has partnered with Cerner, a $22 billion healthcare company.
It has not been plain sailing for IT companies in the healthcare space however. This legislatively complex sector places patient confidentiality as a core value, and as such cannot provide access to the data riches that AI systems depend on. Just last year, the UK Information Commissioner’s Office ruled that Deepmind’s deal with London’s Royal Free Hospital Trust for information on 1.6m patients – to create a diagnosis and detection system for acute kidney injury – breached UK data law and the trial was stopped.
“Our investigation found a number of shortcomings in the way patient records were shared for this trial. Patients would not have reasonably expected their information to have been used in this way, and the Trust could and should have been far more transparent with patients as to what was happening,” said Elizabeth Denham, the information commissioner. “The Data Protection Act is not a barrier to innovation, but it does need to be considered wherever people’s data is being used,” she added.
Furthermore, data quality is a problem as healthcare partnerships have not given IT firms the ability to influence data collection and recording in a way that suits digital and AI inputs. For that the IT firms need more power over the process, making the next logical step for the likes of Google, IBM, Apple and Amazon, the acquisition of healthcare providers.
By acquiring healthcare providers, IT firms will have direct access to patients, making it much easier for them to encourage patient consent, as well as re-design procedures for better data quality. Patient data would become available immediately and in real-time, then by taking advantage of wearable and smart home technology IT firms would likely extend patient monitoring beyond healthcare facilities – thereby spawning a new era in data driven patient care.
“AI is looking increasingly like the transformative technology of our times,” states Nesta’s John Loder in a recent article. “Cutting edge learning machines are being trained to drive a car, recognise a face in a crowd, and diagnose diseases; capabilities that were out of reach even a few years ago. In an age when vast rewards and adulation can go to those who solve relatively trivial problems such as ordering a taxi, or sharing a photo, AI could be a genuinely radical technological leap.”