AI Sharpens Pathologists' Interpretation of Tissue Samples

Pathologists' examinations of tissue samples from skin cancer tumours improved when they were assisted by an AI tool. The assessments became more consistent and patients' prognoses were described more accurately. This is shown by a study led by Karolinska Institutet, conducted in collaboration with researchers from Yale University.

It is already known that tumour-infiltrating lymphocytes (TILs) are an important biomarker in several cancers, including malignant melanoma (skin cancer). TILs are immune cells found in or near the tumour, where they influence the body's response to the cancer. In malignant melanoma, the presence of TILs plays a role in both diagnosis and prognosis, with a high presence being favourable. An important part of pathologists' work in malignant melanoma is to estimate the amount of TILs. Researchers at Karolinska Institutet have now investigated how pathological assessments were affected by an AI tool trained to quantify TILs.

The study included 98 pathologists and researchers from other professions divided into two groups. One group consisted solely of experienced pathologists. They worked ‘as usual’, i.e. they looked at digital images of stained tissue sections and estimated the amount of TILs according to current guidelines. The second group included pathologists, but also researchers from other professions - all of whom had some experience in assessing pathological images. They also looked at the images ‘as usual,’ but were assisted by AI support that quantified the amount of TILs. Everyone assessed 60 tissue sections, all from patients with malignant melanoma. The study was retrospective, so the images showed tissue samples from patients whose diagnosis and treatment had already been determined.

The assessments made with AI support were superior to the others in several ways. Among other things, reproducibility was very high – the results were very similar regardless of who performed the review. This is important because assessments of TILs can currently vary depending on who performs them, which can compromise medical safety. The AI-supported assessments also provided a more accurate picture of the patients' disease prognoses – since the study was retrospective, there was a ‘correct answer’ to compare with. However, this outcome was unknown to those who assessed the images.

"Understanding the severity of a patient's disease based on tissue samples is important, among other things, for determining how aggressively it should be treated. We now have an AI-based tool that can quantify the TIL biomarker, which could help with treatment decisions in the future. However, more studies are needed before this AI tool can be used in clinical practice, but the results so far are promising and suggest that it could be a very useful tool in clinical pathology," says the study's last author, Balazs Acs, associate professor at the Department of Oncology-Pathology at Karolinska Institutet and a clinically active pathologist.

Aung TN, Liu M, Su D, et al.
Pathologist-Read vs AI-Driven Assessment of Tumor-Infiltrating Lymphocytes in Melanoma.
JAMA Netw Open. 2025;8(7):e2518906. doi: 10.1001/jamanetworkopen.2025.18906

Most Popular Now

AI Catches One-Third of Interval Breast …

An AI algorithm for breast cancer screening has potential to enhance the performance of digital breast tomosynthesis (DBT), reducing interval cancers by up to one-third, according to a study published...

Researchers Create 'Virtual Scienti…

There may be a new artificial intelligence-driven tool to turbocharge scientific discovery: virtual labs. Modeled after a well-established Stanford School of Medicine research group, the virtual lab is complete with an...

From WebMD to AI Chatbots: How Innovatio…

A new research article published in the Journal of Participatory Medicine unveils how successive waves of digital technology innovation have empowered patients, fostering a more collaborative and responsive health care...

AI also Assesses Dutch Mammograms Better…

AI is detecting tumors more often and earlier in the Dutch breast cancer screening program. Those tumors can then be treated at an earlier stage. This has been demonstrated by...

RSNA AI Challenge Models can Independent…

Algorithms submitted for an AI Challenge hosted by the Radiological Society of North America (RSNA) have shown excellent performance for detecting breast cancers on mammography images, increasing screening sensitivity while...

AI could Help Emergency Rooms Predict Ad…

Artificial intelligence (AI) can help emergency department (ED) teams better anticipate which patients will need hospital admission, hours earlier than is currently possible, according to a multi-hospital study by the...

Head-to-Head Against AI, Pharmacy Studen…

Students pursuing a Doctor of Pharmacy degree routinely take - and pass - rigorous exams to prove competency in several areas. Can ChatGPT accurately answer the same questions? A new...

NHS Active 10 Walking Tracker Users are …

Users of the NHS Active 10 app, designed to encourage people to become more active, immediately increased their amount of brisk and non-brisk walking upon using the app, according to...

New AI Tool Illuminates "Dark Side…

Proteins sustain life as we know it, serving many important structural and functional roles throughout the body. But these large molecules have cast a long shadow over a smaller subclass...

The Human Touch of Doctors will Still be…

AI-based medicine will revolutionise care including for Alzheimer’s and diabetes, predicts a technology expert, but it must be accessible to all patients. Healing with Artificial Intelligence, written by technology expert Daniele...

Deep Learning-Based Model Enables Fast a…

Stroke is the second leading cause of death globally. Ischemic stroke, strongly linked to atherosclerotic plaques, requires accurate plaque and vessel wall segmentation and quantification for definitive diagnosis. However, conventional...

Brain Imaging may Identify Patients Like…

By understanding differences in how people’s brains are wired, clinicians may be able to predict who would benefit from a self-guided anxiety care app, according to a new analysis from...