Skip to content
A new study has found that AI can significantly improve the accuracy and consistency of pathologists’ assessments in diagnosing melanoma.
The research, led by Sweden’s Karolinska Institutet in collaboration with Yale University, focused on tumour-infiltrating lymphocytes (TILs) — immune cells found near tumours that are important for diagnosis and prognosis in cancers like melanoma.
In the study, nearly 100 participants assessed tissue samples for TILs. One group worked unaided, while another group used an AI tool trained to quantify TILs.
Results showed the AI-assisted group produced more accurate and consistent evaluations. Because the study was retrospective, researchers could compare their assessments with actual patient outcomes — and those using AI were notably closer to the mark.
Balazs Acs, study author and Associate Professor at Karolinska Institutet, said:
“We now have an AI-based tool that can quantify this important biomarker. While more research is needed before it can be used in clinics, the results are very promising.”
The study highlights AI’s growing potential in clinical pathology, especially in supporting complex cancer diagnoses and treatment decisions.
Read the full story here.
