MICCAI 2023 Workshop Event
Clinically-oriented and Responsible AI for Medical Data Analysis
(Care-AI)
October 8, 2023 | Vancouver, Canada
Overview
Over the past decade, Artificial Intelligence (AI) and Machine Learning (ML) have advanced many fields including finance, education, and healthcare. In healthcare, AI/ML has shown promising performance and efficiency in medical image analysis, computer-aided diagnosis, and computer assisted intervention systems. However, there is a growing concern regarding the potential risks occurring as a result of poor design and development of AI applications. To address this pressing concern, responsible AI (RAI) has recently attracted increasing attention. RAI can be defined as the process of designing, implementing, and deploying AI algorithms that are fair, reliable, generalizable, explainable, robust, and secure. These principles and values are of paramount importance in high-stake fields such as healthcare.
Call for Papers
In this workshop, we will examine the technical and research progress towards accountable and
responsible AI in medical image analysis, computer-aided diagnosis, and computer assisted
intervention systems. The workshop will bring researchers and scientists from diverse
communities, such computer science, machine learning, medicine, social science, and security
& privacy community, into one place to overcome the challenges of developing responsible
AI-based clinical decisions, and discuss possible extensions of current notations and methods
of explainability, robustness, privacy, and fairness toward clinically-oriented and responsible AI
in healthcare.
Topics of interest include, but are not limited to:
- Theoretical and technical approaches for mitigating bias in AI algorithms
- Unbiased and fair benchmark medical datasets and evaluation metric for developing any AI-based approach
- Theoretical and technical approaches for explainability/interpretability
- Theoretical and technical approaches to ensure secure AI algorithms that protect patient’s privacy during different stages of algorithm development pipeline
- Privacy-preserving AI algorithms on medical data including images, videos, 1D signals, graphs, and text
- Developing AI models considering the competing constraints against each other for interpretability, reproducibility, fairness, privacy, safety, and robustness
- Developing trustworthy AI-based approaches for medical image analysis, clinical decision and computer assisted intervention
- Integrating insights from clinical domain expertise into the design, deployment, and maintenance of ML-based clinical decision-making systems
- Analysis of the current obstacles to utilizing responsible and trustworthy AI in the current practical healthcare system
- Medical machine learning under robustness constraints such as adversarial attacks, distribution shift, environment changes, etc.
Important Dates
- Paper submissions due: Jul 1, 2023 (AoE)
- Notification of paper decisions: Aug 5, 2023
- Camera ready papers due: Aug 12, 2023
- Workshop date: Oct 8, 2023
Paper Submission
Both short and full-length papers covering the topics of interest are welcomed.
