- Fotolia

IBM’s Watson supercomputer brought to Finnish healthcare

Authorities in Finland are working with IBM to improve healthcare through cognitive computing

Digitisation of healthcare is expected to bring benefits to patients, professionals and hospitals alike in the form of increased efficiency, flexibility and cost savings.

In Finland, for example, a deal between IBM and the Finnish Funding Agency for Innovation (Tekes) aims to tap into health data and IBM’s cognitive computing platform Watson Health.

Announced in September 2016, the agreement will drive innovation and the development of a digital healthcare ecosystem in Finland, as well as open doors for new business opportunities globally. Finnish companies and healthcare professionals get access to the cognitive and analysis powers of Watson, while IBM gains data and partnerships to further develop its platform.

“We wanted to bring this kind of major global player to Finland in a more significant way, so its artificial intelligence and cognitive solutions can be used to enhance our healthcare economy,” Anssi Pulkkinen, strategic head of health and well-being at Tekes, told Computer Weekly.

IBM is investing in Finland. It is planning to set up Watson Health centres of expertise and innovation in Finland, as well as the tech giant’s first national imaging centre outside the US. These centres are expected to employ 150 people in the next few years.

The Watson Health centre of excellence will be the first to open its doors in early 2017, at IBM Finland’s headquarters in Helsinki, while the imaging centre is scheduled to launch later in the year.

“Watson and Watson Health are very new units for IBM globally. It has not had representation in Finland before, so we have been dependent on global expertise. These centres will have experts trained here,” said Mirva Antila, deputy country manager at IBM Finland. “The Watson Health centre of excellence will be the first one in the Nordics and also serve our customers in Europe.”

The call of health data

The one key ingredient all artificial intelligence (AI) and cognitive computing systems, such as Watson, need to learn and evolve with is data. And it is data that caught IBM’s interest in Finland.

“Finland has many characteristics that make it unique,” said Antila. “For example, all healthcare and patient data has been digitised in Finland for the past 30 years. Although it is currently dispersed and not in efficient use, it is already in digital format.”

The Nordic country has national electronic health records, biobanks and a cancer registry database of all cases in the country dating back to 1953. There is also the national health data archive, Kanta, which covers electronic prescriptions, patient data repository and the citizens’ health data service for the country’s 5.5 million inhabitants.

The presence of data does not mean it is easily available, and Finland is working to make its health data more accessible. The country is planning new, more permissive legislation for the secondary use of data on well-being and health. Its upcoming genome centre is tasked with creating a national genome database and facilitating its use in clinical care and research.

Predicting illness

These advances in the foundations for digital healthcare are important, as Finnish health professionals, researchers and companies seek to develop data-driven healthcare applications and services. IBM is already working with several partners, from startups to universities, to test the use of cognitive computing in improving healthcare.

Key among them is the Hospital District of Helsinki and Uusimaa (HUS), a joint authority covering 24 municipalities, which is planning three pilot projects to trial Watson Health’s cognitive and data analysis skills in predicting health issues.

These include the early identification of serious bacterial infections in prematurely born babies and using imaging data to detect a specific type of cerebral hemorrhage. HUS is also employing cognitive computing in oncology to help provide patients with more personalised cancer care.

Read more about IT in healthcare

“We have chosen these because bacterial infections in prematurely born babies and cerebral hemorrhages are the type of conditions where there is a lot of data, and if that could help us to diagnose these conditions even a few hours earlier, it could have a significant effect on the patient’s prognosis,” Visa Honkanen, director of strategic development at HUS, told Computer Weekly.

If the pilot results are encouraging, HUS aims to expand the projects to combine cognitive computing with genetic information and Finnish health data to further improve the diagnostics of specific illnesses.

“When machine intelligence is taught, you first have to create a model it can use to predict certain situations. When that model is then applied in real situations with new data, we really see if it is beneficial and what it can offer us,” added Honkanen. “If the model works, AI should learn to improve its predictions.”

Hospitals without a hospital

IBM, Tekes and HUS all believe artificial intelligence will play a major role in the future of healthcare. It is anticipated to support the detection and diagnosis of diseases, as well as enable more personalised healthcare as the amount of health, patient and research data rapidly grows.

“Machine learning will be very beneficial, particularly if you have a large number of sensors collecting high amounts of data that no human can analyse in a reasonable way, said Honkanen. “A machine can do the pre-filtering and analysis of this data, and give red flags before a human brain gets involved. However, the idea of AI becoming a diagnostic detective that solves problems no one else can is still far away compared to the human brain.”

One future vision Tekes advocates is a hospital without a hospital, currently given the working title of “IoT hospital”. This refers to the compliant and efficient use of health data combined with digital technologies and monitoring, enabling treatment at home while keeping hospital visits to a bare minimum.

But making this vision a reality requires more innovation in digital healthcare, and this is what Tekes believes its partnership with IBM will help Finnish companies achieve.

“When Finnish companies start to solve healthcare problems in collaboration with companies such as IBM, it leads to solutions that have never been seen before,” said Tekes’ Pulkkinen. “We want to achieve systemic benefits for patients and the national economy.”

Read more on Healthcare and NHS IT