Seminar AI for Healthcare: Challenges in Translating Promises into Patient Outcomes



Time and place:

This course will be conducted online. Registration will be enabled via StudOn starting in May. If you are interested in attending the seminar, please send an email to

Fields of study

  • WPF INF-MA from SEM 1
  • WPF AI-MA from SEM 1
  • WPF MT-MA-BDV from SEM 1

Prerequisites / Organizational information

FAU students register for the course in StudOn. Registration will be enabled via StudOn starting in May. If you are interested in attending the seminar, please send an email to
The seminar is offered as a compact course during summer intersession in September (exact dates are tbd).
This course is offered under the prerequisite that the corresponding funding for Prof. Dr.-Ing. Mathias Unberath and Dr. Nishant Ravikumar is granted by FAU.


Artificial Intelligence in general, and machine learning (AI/ML) in particular, have become a major thrust of healthcare research. Concisely, it is now widely accepted that learning-based approaches will be a core building block of personalized and precision medicine. The reasons for this are twofold: First, these methods either automate data analysis tasks that would be intractable otherwise thus paving the way for innovative decision making; and second, they offer recommendations in high-variance decisions based on population-scale evidence used for their development, thus potentially decoupling provider experience and outcomes.
Unfortunately, most of the recent successes on private in house or public grand challenge data have been linked to neither improved outcomes nor clinical impact but are limited to task-based comparisons in sandbox settings. Furthermore, developed techniques that have been validated thoroughly in a research setting often fail/perform poorly in clinical ones, and do not account for inherent biases in the data and/or experimental setup.
In this seminar, we will review recently published research on AI/ML for healthcare that successfully translated into clinical practice to identify key factors in study design, method development, infrastructure, or regulation that enable translation.
The seminar will focus on three distinct areas: digital pathology, medical image computing, and computer-aided interventions. Where possible, guest lectures from academia, clinics, as well as industry will be invited as part of the seminar.

Students will be able to

  • independently identify challenges in translating technical solutions from the bench to the bedside, and assess how close to clinical feasibility a technical solution is

Students will have acquired competences to
  • perform an unstructured literature review on an assigned subject

  • independently research the assigned subject

  • present and introduce the subject to their peers

  • give a scientific presentation in English according to international conference standards

  • summarize their findings in a written report that adheres to good scientific practice

The overall grade consists of two parts: A 30-minute seminar presentation (50% final grade, comprised of content and delivery). The goal of the seminar is to prepare a topic for other students in an accessible way.

After all groups have presented their topics, we will break out into smaller teams to further process the seminar talk contents and synergize them into a paper-style report and report-out (conference-style) presentation (~4 pages IEEE and 10 minutes, respectively; 50% final grade, comprised of content and delivery) that discusses at least one core challenge identified throughout the seminar and proposes community guidelines to improve translation of AI research into clinical practice.

Talks and seminar paper should be in English.
Students will work in groups of two if the number of participants allows.

Recommended Literature

Unberath, M., Ghobadi, K., Levin, S., Hinson, J., & Hager, G. D. (2020). Artificial Intelligence‐Based Clinical Decision Support for COVID-19–Where Art Thou?. Advanced Intelligent Systems, 2(9), 2000104. Christopher J. Kelly, Alan Karthikesalingam, Mustafa Suleyman, Greg Corrado & Dominic King: Key challenges for delivering clinical impact with artificial intelligence, BMC Medicine, Vol. 17, Article number: 195 (2019) Adam Bohr and Kaveh Memarzadeh (eds.): Artificial Intelligence in Healthcare, Academic Press (2020) Herein for example: Sara Gerke, Timo Minssen, Glenn Cohen: Chapter 12 - Ethical and legal challenges of artificial intelligence-driven healthcare, Adam Bohr, Kaveh Memarzadeh, (eds.), Artificial Intelligence in Healthcare, Academic Press, pp. 295-336 (2020)

Additional information

Expected participants: 10