The Lancet via YouTube: Some of the most significant advances in clinical artificial intelligence (AI) have occurred within oncology, including the development of models for cancer diagnosis and early detection, computational pathology and radiological imaging, patient monitoring and survival prediction, and drug discovery. However, AI systems are often developed using clinical datasets from cohorts that lack diversity and, when tested, can exhibit unfavorable biases against minority groups. There is a need to prioritize health equity in the development and implementation of AI in cancer care by creating diverse and representative clinical datasets, mitigating algorithmic biases, and optimizing the way humans interact with AI to maximize its benefits for all.
During this webinar from The Lancet and The Lancet Digital Health, a panel of experts explore the current state of AI in oncology and future opportunities for its application. Our panel place a particular focus on achieving fairness and equity for patient groups who might be at risk of health disparities or biases propagated by clinical AI.
