Combinostics: Artificial intelligence assists doctors in diagnostics

Combinostics Oy specialises in clinical diagnostics and data analysis, in order to develop solutions for more reliable diagnostics of neurological diseases. Tekes has been involved in Combinostics’ development work since its earliest stages of research.

When doctors examine neurological diseases, they make various types of diagnostic measurements which produce a great deal of data. In many cases, some results point in one direction, and others in another, which complicates decision-making.

A service developed by the Tampere-based Combinostics assists doctors in their work.

"Today, radiologists interpret most images visually. We have developed a software that measures various types of structures from magnetic resonance images. When the numerical measurements taken from the images are combined with other data, we get a holistic picture of the patient's status. Doctors have said that this is exactly the kind of tool they need," says Jyrki Lötjönen, Chief Scientific Officer at Combinostics.

The company name, Combinostics, comes from the words 'combined diagnostics', which describe the company's mission – combining various types of data for diagnostic purposes. The company's first software product, which is based on the measurement of diagnostic images, was launched in October 2016. Another product, to be launched in spring 2017, compares the patient's measurements to large databases in order to assess the degree of similarities between the patient and previous cases.

Research companies require business expertise

The technology applied by Combinostics was invented more than ten years ago, when the company's four co-founders researched data-based tools at VTT Technical Research Centre of Finland Ltd. After the first project, more projects around the theme followed, until it was time for the team to establish a business of its own.

"Doctors encouraged us to bring our innovation to the market. Since we were confident that artificial intelligence-based applications will become a daily tool at some point, we decided to set up a company around our research," explains Lötjönen.

Combinostics strengthened its researcher staff by recruiting its CEO from a GE Healthcare unit in the UK. His business competencies have been essential.

"There is a lot of good research going on in Finland, but the research results are often not put into practice. For many researchers, the transition to commercialisation and the business world is a frightening leap into the unknown."

Lötjönen urges researchers to partner with business specialists from the very beginning. He admits that the co-founders also need motivation to set up a business that combines research and entrepreneurship.

"We were passionate about seeing the results our years of research work appear on doctors' desks. A couple of years ago, we made the decision to establish a business of our own, and it still feels right," he said.

Peer support from colleagues in Silicon Valley

Combinostics has made use of the Team Finland network by making two visits to the Mecca of growth entrepreneurs – Silicon Valley in California, USA. A key takeaway from these visits was the sense of community.

"Other companies in our sector are struggling with the same problems," Lötjönen points out.

He felt that it was also extremely important to meet colleagues from other sectors. The training and pitching events, conferences and company visits in San Francisco were also very rewarding.

Several Combinostics projects have been funded by Tekes. For example, some of the company's image processing software algorithms were developed in projects that were completed as long as 10–15 years ago.

"In the field of medicine, inventions often have a much longer time-span than in many other sectors. When something is under development, it is difficult to say when the innovation will actually end up in stores."

For further information, please contact

Jyrki Lötjönen
Combinostics Oy
Tel. +358 (0)50 346 3250


Text: Kaiku Helsinki
Photo: Combinostics

Pia Mörk