Why so much fuss around data?
Artificial Intelligence (AI) is the new buzzword in the technology world. It is becoming impossible to imagine the future of the healthcare sector without AI. Numerous startups and tech firms have already embarked on the AI hype wagon, promising a host of sophisticated healthcare solutions. In fact, AI is already significantly changing the healthcare landscape.
AI has a symbiotic relationship with data, as well as learning from it. In 2016 alone, the amount of data generated in the world exceeded the total data ever generated by mankind. As per a McKinsey estimate, in not so distant future big data and machine learning in pharma and medicine could generate a value of up to $100 Bn annually.
The healthcare sector armed with AI is sitting on a data goldmine. For the healthcare industry seeking more intelligent solutions, AI will usher in a disruptive transformation beyond anybody’s imagination.
The data availability conundrum
Despite the excitement surrounding AI, its path to becoming the “Next Big Thing” in medicine is not free from obstacles. The recent reports about erroneous cancer diagnosis by IBM Watson system revealed a lot about the challenge that AI faces.
IT behemoth IBM was the one of the first company to test the water with the much talked about ‘IBM Watson Health Cloud’. In their own words, it is a global platform built with the aim of helping healthcare providers make timely, evidence-based decisions about health-related issues.
A few months ago, it was reported that Watson supercomputer often provides erroneous cancer treatment advice and that company medical specialists and customers identified “multiple examples of unsafe and incorrect treatment recommendations”. As per the reports, internal documents indicated that the training and effectiveness of the Watson for Oncology system was flawed due to the small number of cases and the inclusion of artificial cases.
With the exponential growth of computing power, we already can handle and interpret big data meaningfully. But AI requires personal data on a huge scale, at least until it becomes intelligent enough to teach itself. Thus, AI is dependent on the quality of training data to be effective. In many cases, such data is in very short supply or difficult to access. It’s a catch 22 facing the entire field of machine learning for health care.
Data security and privacy concerns
In the aftermath of Cambridge Analytica fiasco, setting precedents for the nefarious and unethical use of the technology, there is a renewed debate about data security and privacy.
Most healthcare data are heavily protected by law, and medical records are typically owned by the treatment providers. Thus, any breaches or failure to maintain data integrity can have serious legal and financial ramifications.
At least in data usage, you can’t have your cake and eat it too. How the AI service providers are going to balance this delicate act will be crucial to the progress of this sector.
The Israeli way
Any nation with a robust healthcare database coupled with a technological know-how will rule the roost in the AI world. The nation that fits this bill perfectly is the tiny startup nation of Israel.
Israel started digitalization of healthcare data 20 years ago much ahead of many other OECD nations. Israel has medical records of close to 9 million people collected over the past 20 years.
With its robust healthcare system and strong research and development infrastructure in place, Israel has “huge” global competitive advantage.
The Israeli government on March 2018 approved a National Digital Health plan, which, despite mounting privacy concerns, plans to create a digital database of the medical files of some 9 million residents and make them available to researchers and enterprises.
The five-year program will have a budget of nearly $280 million. The funds will go toward digitizing and sharing patient data among the country’s health funds, relying on AI tools to more accurately detect abnormalities and find correct diagnoses.
The key features
The digital-health initiative unveils an anonymous Israeli database, named the Mosaic Project, which can show long-term disease and illness trends of its citizens for the past 20 years, offering a huge data set for researchers.
In addition, regulators will work together to make sure information can be accessed anonymously, maintaining privacy and securing information and access permissions. Participation in all the projects is exclusively on a voluntary basis.
The project will create a community of volunteers to contribute clinical, genomic, and other information developing customized medical solutions and in-depth analysis of big data.
The project will also establish a centre for encouraging innovation in healthcare – connecting Israeli start-ups to MNCs – and the promotion of international projects with other countries’ health systems.
The fund plans to collaborate with the government-funded program and continue its investments in digital health and in other sub-sectors, such as biopharmaceuticals and medical devices.
The program will also help establish a national centre for genetic sequencing in Israel, for the research and analysis of samples.
The political will to leverage technology
With a technology as disruptive as AI whose long-term effect is beyond anybody’s wildest imagination, there should be a comprehensive political framework to see the developments through.
The Israeli model is driven by the understanding that “Nothing is more powerful than an idea whose time has come”.
While there is a global buzz around AI and its application among the business community. Nowhere in the world except Israel, you can see the political intent to utilize AI to its fullest potential.
Israel exhibited a similar political will in the propagation of their cybersecurity space last decade. And the result is a cybersecurity superpower nation punching way above its weight. In 2017, Israel garnered some 16% of all cybersecurity investment globally an astonishing sum compared to Israel’s small size.
Israel’s prime minister Benjamin Netanyahu is upbeat about Israel’s possibility of tapping into the $6 trillion global digital health sector. By his conservative estimates, Israel might be able to snag about 10% of this market potential, worth nearly $600 billion.
European protectionism vs Israeli chutzpah
Earlier this year on 25th May the EU legislation, General Data Protection Regulation (GDPR) came into effect. This has been hailed as a protective tool that offers citizens more control over, and the assurance of greater security for, their personal data.
GDPR contains rules that both, directly and indirectly, limit the development and use of AI. But the data science experts believe that this will hamper development AI based therapeutics.
The GDPR covers even offshore usage of European data. Noncompliance with the GDPR’s extremely complex set of rules results in stiff penalties that make advanced data processing a legally and financially risky endeavour.
Trade-offs between data protection and innovation will have a negative impact on the development and use of artificial intelligence (AI) in Europe, putting EU firms at a competitive disadvantage compared with their competitors in North America and Asia.
On the contrary, Israel through the new legislation is seeking to encourage the export of digital health solutions and encourage foreign enterprises to invest in Israel in these fields.
Therefore, it won’t be surprising if Israel became an alternative heaven for AI companies disenchanted by the complexities of GDPR.
Willingly or unwillingly we live in an interconnected world where we can log out of the internet anytime we like, but we can never leave. Commercial tracking and targeting are becoming the new normal.
Data privacy undoubtedly poses several challenges to AI applications in medicine. Data protection awareness is becoming increasingly relevant and organizations need specific governance guidelines when dealing with AI.
In terms of data usage for AI, there are three approaches largely –
- The ultra-conservative GDPR like approach.
- The absolutely no respect for data privacy approach.
- The Israeli approach of calculated risk-taking allowing commercial exploitation of public data albeit with restrictions like anonymity and consent in place.
If the Israeli approach is the foolproof one is not beyond debate. However, to realize the optimal potential of AI in medicine or related fields we need to find a “no trade-offs” middle ground between data protection and innovation.
Israel is trying to hit the sweet spot by making data accessible for AI applications yet making it anonymous to protect privacy. This will prevent data security regulations from becoming the Achilles’ heel of a burgeoning data-driven economy. And larger economies can take a leaf out of Israel’s example to set the ball rolling in the AI front.
Do comment and share if you find this article useful.