Automated Cancer Diagnostics with Artificial Intelligence

The aim of Kartasalo's research is to develop an artificial intelligence method for identifying biomarkers of prostate cancer inexpensively from tissue biopsy samples. The work is based on digitized diagnostic tissue samples, some of which are also profiled genomically and transcriptomically. The algorithm will be integrated into a cost-effective portable digital pathology scanner developed in Tampere, enabling the method to be utilized in developing countries and small clinics.

"The use of artificial intelligence is currently increasing in clinical work, with experiments gradually becoming more routine. For now, artificial intelligence is primarily a technology that assists doctors. A comparison could be made to features like lane-keeping assistance or automatic braking systems in cars. They support the driver, but the human is still in control. However, I believe that within five years, the everyday work of doctors will also change due to the impact of artificial intelligence. With artificial intelligence, we have an tireless pair of eyes examining research samples, while simultaneously increasing our ability to handle even the largest datasets," Kartasalo explains regarding the future possibilities of artificial intelligence.

In 2024, the Instrumentarium Foundation granted Kartasalo a €50,000 post-doctoral grant for the research project "Automated Prostate Cancer Diagnostics and Molecular Profiling Using Artificial Intelligence and Cost-Effective Digital Pathology Scanners."

Various samples challenge artificial intelligence

Kartasalo, with a background as a civil engineer, earned his doctorate in 2021 from the University of Tampere in the field of biomedical engineering. In his doctoral dissertation, Kartasalo delved into digital pathology and developed a prototype artificial intelligence algorithm for diagnosing prostate cancer.

"We receive images of cancer tissues from Finland and Europe, taken with various equipment in different environments. This makes analysis challenging because artificial intelligence must be able to analyze various types and levels of images. Even the best scanners degrade over time, affecting the quality of images. Artificial intelligence must be able to analyze all these different samples accurately for its results to be trusted," Kartasalo points out.

Images and visual aspects have always interested Kartasalo. He has been interested in photography, and in his research, images are more engaging than mere values and numbers. Kartasalo, who works at the Karolinska Institute in Stockholm, became acquainted with cancer research during his time at the University of Tampere. He wants to be involved in developing a method for detecting cancer more quickly, thus aiding in overcoming this dangerous disease.

Artificial intelligence does not make doctors unnecessary

The current government program in Finland aims at an additional 1000 doctoral degrees being completed over the next three years, in addition to the previous 1600. More medical doctors are needed in Finland because the population is aging, leading to an increase in the number of patients requiring treatment. However, will the use of artificial intelligence also affect the number of medical doctors and physicians needed in the future?

"The use and utilization of artificial intelligence will probably not reduce the need for medical doctors in Finland, at least in the short term. The total workload of doctors will not decrease, but analysis methods and treatment modalities will improve. Hopefully, artificial intelligence will reduce the workload of doctors in certain tasks, such as making routine pathological assessments. This saved time can hopefully be used to cope with the increasing number of patients and the increasingly complex and precise diagnostics, where, for example, pathologists must be familiar with an increasingly wide range of analytical techniques," Kartasalo reassures.

Smooth grant application

Kartasalo's €50,000 grant will be available to him between June 2024 and June 2025. With it, he can focus full-time on his research. Otherwise, a significant portion of the researcher's time would be spent searching for alternative funding channels and preparing grant applications.

"The grant allows me to have a roof over my head in Stockholm and food in the fridge. Some of the grant also goes towards necessary conference trips and participation fees," Kartasalo says.

Kartasalo describes the grant application process of the Instrumentarium Foundation as smooth overall.

"For researchers, it would be easier if the application services of different funders were more alike. With less time and effort, one could participate in more applications. This would leave more time for actual research," he suggests.

Text: Kai Tarkka
Image: Gunilla Sonnebring