EBAIC 2023 : The 1st International Workshop on Ethics and Bias of Artificial Intelligence in Clinical Applications (EBAIC 2023)
Call For Papers
Call for Papers / Participation
The volume of Electronic Health Record (EHR) data has grown dramatically in the past decade due to the wide adoption of EHR systems in healthcare systems. The availability of large amounts of multimodal clinical data has fostered the application of Artificial Intelligence (AI) in clinical care including in clinical decision support, patient management, as well as in clinical and translational research, such as digital phenotyping, cohort discovery, and in-silico trials. Despite the promising potential of AI in clinical applications, its regular use comes with bias and ethical challenges. As highlighted by recent studies, disparities in health care, although may start at the collection of clinical data, could be amplified with the development of AI technologies.
Topics of interest
Any original research related to ethics and bias of AI in clinical applications. The relevant AI techniques include, but are not limited to, natural language processing, medical imaging, deep learning, predictive modeling, human computer interface, Internet of Things, and more. Clinical applications include, but are not limited to, clinical decision support, clinical research, translational research, consumer applications, robotics.
Other relevant topics include: AI for health equity, AI for health disparity, transparency/interpretability/explainability of AI techniques in clinical applications, data bias, algorithmic bias, human bias of AI techniques, fairness metrics, fairness evaluation, fairness tools, reasoning, practical and technical solutions to mitigate the bias, and more.
Finally, we will consider limited types of position papers on AI ethics/bias. This would include position papers from individuals/groups that are part of a community that has historically been adversely impacted by artificial intelligence, bias, or health disparities. We will also consider position papers from institutions playing key roles in mitigating the impact of bias in clinical applications.
Fei Wang, PhD, FACMI, FAMIA, FIAHSI, ACM Distinguished Member.
Associate Professor of Health Informatics, Weill Cornell Medicine. Founding Director. WCM Institute of AI for Digital Health.
Oral Presentations & Posters
Regular Papers: 10 pages with up to 2 extra pages for references/appendices.
Will describe mature ideas, where a substantial amount of implementation, experimentation, or data collection and analysis has been completed.
Short Papers: 6 pages with up to 1 extra page for references/appendices.
Will describe innovative ideas, where preliminary implementation and validation work have been conducted.
Poster Submissions: 2 pages with up to 1 extra page for references/appendices.
Will present innovative ideas, late-breaking work, concepts, work-in-progress, early stages of research, and preliminary results from implementation and validations to academic and industrial audience.
Position Papers: 4 pages with up to 2 extra pages for references/appendices.
Will present an arguable opinion about AI ethics/bias and its impact.
The purpose of the tutorial/hackathon session is to raise awareness of the problem of bias in clinical data and AI algorithms with the ultimate goal of creating innovative approaches that can help reduce or eliminate bias in clinical data and AI. Participants may be students, researchers, and data scientists who are interested in applying AI to clinical applications. Complete this form to register the Tutorial/Hackathon Session.
Track I: Clinical Natural Language Processing
Data: de-identified clinical notes from MIMIC III
Algorithm: Rule-based NLP algorithm
Task: Understand how stigmatizing language in clinical notes varies by patients' medical condition and race/ethnicity
Pre-requisite: Complete required training and sign the data use agreement for the MIMIC III data access at https://physionet.org/content/mimiciii/1.4/
Track II: Medical Imaging
Data: Knee X-ray images from NIH OAI publicly available dataset
Algorithm: Convolutional Neural Networks
Task: Understand how knee joint segmentation and measurement varies by different racial or gender groups, and imbalanced training data
Submission and Review
Anyone who is interested in ethics and bias of AI in clinical applications is invited to submit their work to the EBAIC 2023.
Authors can log in at https://easychair.org/conferences/?conf=ieeeichi2023 and submit their papers under the "ebaic" track. All submitted papers will be peer-reviewed by domain experts.
For more information regarding paper template and review process, please visit https://ieeeichi.github.io/ICHI2023/call_for_papers.html.
All submissions will be published in IEEE Xplore and indexed in other Abstracting and Indexing (A&I) databases. Accepted papers have an oral presentation slot at the conference.
Yanshan Wang, PhD, University of Pittsburgh, Pittsburgh, PA, USA
Hongfang Liu, PhD, Mayo Clinic, Rochester, MN, USA
Ahmad P. Tafti, PhD, University of Pittsburgh, Pittsburgh, PA, USA
Kirk Roberts, PhD, The University of Texas Health Science Center at Houston, USA
David Oniani, University of Pittsburgh, Pittsburgh, PA, USA
Sonish Sivarajkumar, University of Pittsburgh, Pittsburgh, PA, USA
Rema Padman, PhD, Carnegie Mellon University, USA
Fei Wang, PhD, Weill Cornell Medicine, USA
Vikas Singh, PhD, University of Wisconsin-Madison, USA
Hossein Estiri, PhD, Harvard Medical School, USA
Deadline for all submissions: March 21st, 2023
Notification of decisions: April 11th, 2023
Camera-ready due: April 21st, 2023
Workshop date: June 26th, 2023