Fimlab to enhance tissue sample examination with AI technology in Finland

Fimlab Laboratories will enhance the examination of pathological tissue samples using artificial intelligence (AI) technology. AI will analyze samples based on a large dataset taught by humans, interpreting images to achieve diagnostic accuracy comparable to skilled pathologists, or even slightly better, as stated by Mika Tirkkonen, the head of pathology at Fimlab. The AI is trained by analyzing a significant number of digitized tissue samples, a process that has been underway since 2021. It flags suspicious areas, allowing doctors to confirm diagnoses, ensuring patients are not solely reliant on AI assessments.

In Finland, where there is a shortage of pathologists—currently between 160 and 200—the integration of AI is not intended to replace specialists but to manage growing sample volumes and workforce shortages. By 2030, it is projected that cancer diagnoses may reach 43,000 annually, according to Tirkkonen. AI will focus primarily on common cancer types such as prostate, breast, and gastrointestinal cancers, and will also automate labor-intensive manual tasks.

Fimlab’s six-million-euro project is unfolding in phases, with initial efforts on breast and prostate cancer samples. Preliminary AI use in routine diagnostics is expected by the end of the year. Collaboration with physicians will improve reporting accuracy, although there are no immediate plans to radically overhaul diagnostics.

Looking into the future, AI could develop predictive models that operate independently of human input. It has the potential to identify histological patterns that predict cancer behavior, thereby refining treatment decisions. However, the complete replacement of human analysts in tissue sample assessment remains a distant goal. Fimlab stands as Finland’s largest laboratory company, supporting various health and wellness service areas.