07. September 2021

Event Series of the AI Research Group Event Series of the AI Research Group – Summer Program 2022

Summer Semester Program 2022

The guiding aim of the CST AI research group is to provide a forum in which researchers can gain new perspectives on their own work and regular input from across the university on AI-related topics.

We are now meeting on Mondays 12:15-1:45 pm (CET) in the IZPH building (Poppelsdorfer Allee 28). All the events will be organized in a hybrid format.

If you or someone from your team wishes to join, please send an email to Dr. Charlotte Gauvry (cgauvry@uni-bonn.de), and she will add you to the mailing list. Should you wish to present or discuss a paper, a work in progress, or any question regarding AI, please let her know.

Here is the full program, which will be updated regularly.

Next session

July 4 – AI & MENTAL HEALTH: Dr. Theodor Rüber (Epilepsy Department Bonn) “Multi-modal evaluation of epilepsy patients using computational methods”

Abstract: One-third of all focal epilepsy patients are therapy-refractory (Kwan et al. 2010) and may, thus, qualify for epilepsy surgery (Schulze-Bonhage and Zentner 2014). Presurgical evaluation, however, yields negative results in up to one-half of all cases (Bien et al. 2013). The hope that these patients may still cherish relies on scientific advances in developing anti-epileptic drugs and those in presurgical evaluation. Besides qualitative guidelines (Rosenow et al. 2014), no strict protocol or quantitative metrics exist for deciding whether or not a patient is admissible to surgery. It requires careful evaluation of risk and reward, where the interpretation and weighing of all relevant factors rely on experts' experience and knowledge (Zijlmans, Zweiphenning, and Klink 2019, Jehi, Chagin, et al. 2015). To this day, no clinically established framework exists, which explicitly models the uncertainties of single- and multi-modality diagnoses, integrates them with other relevant clinical factors and derives explicit measures for the presurgical evaluation. With this research project, we aim to formalise the decision-making process into a computational framework. In the first part of the project, we will transform existing data from the past treatment of patients into a dataset suitable for computer-aided analysis. The dataset shall combine patient attributes, multi-modal data and the clinical evaluation resulting from experts. We will use the medical records of all patients diagnosed with epilepsy and admitted to surgery as a foundation to build the dataset. We will also include post-operative examinations to validate derived models. With more than three thousand subjects, this dataset will be the single largest multi-modal dataset about epilepsy. We will develop a data-acquisition tool and distribute it to cooperating institutions to expand the dataset and allow external validation. In the second part of the project, we will model the complex associations in the decision process by utilising modern statistical techniques, especially artificial intelligence (AI) models. A strong focus shall be put on attention-based networks as classifiers since they offer good interpretability and potentially allow the combination of data from imaging and time-dependent domains into one system (Vaswani et al. 2017, Dosovitskiy et al. 2020). Applied methods should generally fall into the category of explainable AI (XAI), which has been argued as essential for the ethical evaluation of such systems and to facilitate clinical consensus (Zhang et al. 2019).

Previous sessions

June 27– Dr Julia Mönig and Dr Sergio Genovesi "Certified AI — An ethical analysis of Use Cases"

June 20, 4:15-5:45 pm – AI & MENTAL HEALTH: Dr. Niclas Braun (Clinic for psychiatry and psychotherapy Bonn) "Virtual reality induces symptoms of depersonalization and derealization"

June 13 – Dr Apolline Taillandier "Transhumanist visions of world order"

May 23 – Dr Johannes Lierfeld "Legal issues of advanced humanoid robotics"

May 16 – Dr Oliver Braganza "Proxy divergence: Goodhart’s law as an emergent feature of complex goal-oriented systems"

May 2 – Dr Uwe Peters "Linguistic bias: Clarifying the concept and evaluating the evidence (invoked byphilosophers)"

April 25 – Dr Charlotte Gauvry "Can we experience the passage of time?"

April 5 – Dr Uwe Peters “The impact of mindshaping: AI systems and human cognition are not equally opaque”

Wird geladen