Jakob Nikolas Kather, MD MSc

Titles and Affiliations

Professor, Clinical Artificial Intelligence
Else Kröner Fresenius Center for Digital Health
Senior Physician, Medical Oncology
National Center for Tumor Diseases
Heidelberg, Germany

Research Area

Developing a comprehensive groundwork for accurate and reliable integration of artificial intelligence into breast cancer care.

Impact

Artificial intelligence (AI) tools are utlized in virtually all aspects of life, including in oncology. As patients, caregivers, and clinicians all have access to vasts amounts of information through AI, the accuracy and safety of using it remains under scrutiny. Large language models (LLMs) are advanced AI systems that can process and generate text as a person would. Trained on huge amounts of text, it is still not clear how efficient LLMs are at answering breast cancer questions. Dr. Kather and his colleagues are utlizing their considerable expertise in LLMs and AI to solve this pressing problem. By creating and testing a comprehensive benchmark for evaluating LLM performance specifically in breast oncology, Dr. Kather hopes to optimize their use and provide a groundwork for responsible integration of AI into breast cancer care to ultimately benefit patients.

What’s Next

Dr. Kather’s team will build on their extensive experience developing AI-driven biomarker discovery tools, decision-support systems for oncologists, and automated research workflows. In the coming year, they will develop an AI system that can automatically gather and organize breast cancer knowledge from scientific literature, clinical guidelines, and conference abstracts. The data in the resulting structured database will be verified by independent experts. Next, the validated information will be used to generate questions of varying difficulty, including “safety-critical” questions where incorrect answers could potentially harm patients. A group of experts—including oncologists, pathologists, and patient advocates—will work to refine these questions and ensure they are clinically relevant and properly categorized.

Once the series of questions at varying levels of difficulty are created, the Kather group will test ten different LLMs to answer them, analyzing the models’ accuracy, consistency, and confidence levels. The goal of their work is to identify the types of questions the LLMs struggle to answer and potentially to reveal any patterns in their errors. By creating transparent benchmarks and assessment tools, they hope to leverage the potential benefits of AI—such as personalized patient education and faster creation of medical guidelines—while minimizing risks. This project lays the groundwork for responsible AI integration into breast cancer care, ultimately helping to deliver better care more consistently and safely to patients.

Biography

Jakob Nikolas Kather, MD MSc holds dual appointments in medicine and computer science at the Technical University (TU) Dresden, Germany, serves as a senior physician in medical oncology at the University Hospital Dresden and holds an additional affiliation with the National Center for Tumor Diseases (NCT) in Heidelberg. His research is focused on applying artificial intelligence in precision oncology. Prof. Kather’s research team at TU Dresden is using deep learning techniques to analyze a spectrum of clinical data, including histopathology, radiology images, textual records, and multimodal datasets. Guided by the belief that medical and tech expertise needs to be combined, medical researchers in his team learn computer programming and data analysis, while computer scientists are immersed in cancer biology and oncology. Prof. Kather chairs the “Working group on Artificial Intelligence” at the German Society of Hematology and Oncology (DGHO) and is a member of the pathology task force of the American Association for Cancer Research (AACR). His work is supported by numerous European and national grants, which enable the team to develop new deep learning methods for medical data analysis techniques and to apply them in precision oncology.