AI has an astonishing potential to assist clinical decision-making and revolutionize global health, but this can only be possible if the use of AI in healthcare takes into account the needs of diverse populations. When AI is biased, it is prone to reinforcing inequality, which can lead to injustices in healthcare, making already vulnerable patients more susceptible to fatal outcomes and misdiagnoses.
In this session, we'll target some of the main challenges that need to be addressed in order to move towards fairness in AI for healthcare, including gender and racial bias, data gaps, and ethical concerns. We will divide into breakout groups, where we will brainstorm recommendations to solve these challenges. We will also explore how AI can help make invisible minorities visible and reduce inequalities in healthcare, and how open practices can be used in the fight against bias.
If only a handful of participants attend the session, we will discuss all challenges together. If more participants attend, we'll proceed to divide attendees into break-out groups, where each group will brainstorm solutions for a specific challenge. Each break-out group will have a moderator to guide and incentivize the discussion. We'll prepare virtual templates with post-its to record everyone's contributions and visually present them at the end of the session to the entire group.
We're hoping that many efforts and discussions will continue after Mozfest. Share any ideas you already have for how to continue the work from your session.:Session findings and contributions will be recorded and published as part of our ongoing project "Identifying and mitigating bias in deep learning for biomedical data through open science practices", which is currently funded by the Mozilla Foundation (Mozilla science grant). We'll maintain constant communication with participants to inform them about the progress being made, and we will invite them to contribute to our project, which involves creating a collaborative database to increase the ethnic variability of skin cancer samples, sharing code and techniques to test algorithmic bias in skin cancer detection, and an online workshop to communicate our findings to the wider community.
Healthcare researcher & data enthusiast focused on implementing open science practices to measure key social determinants of health, improve epidemic surveillance and response, and reduce health inequalities.