How AI Can Deliver on Its Promises in Healthcare

In 2016, AI pioneer Geoffrey Hinton famously predicted that deep learning would replace radiologists within five years. Fast forward to 2024, and we see a different reality: AI has yet to replace radiologists but can significantly enhance their work. There remains a shortage of radiologists, and AI offers the potential to help by analyzing more scans, according to UvA researcher Hoel Kervadec, who specializes in medical imaging.

Kervadec explains that AI is not about replacing radiologists but augmenting their capabilities, particularly in dealing with the high volume of scans. By utilizing AI, radiologists can focus on more complex cases, while AI handles simpler tasks. "AI can work faster and better by leveraging existing medical knowledge, but human expertise is still crucial for more complex cases," he says.

Another vital development in AI for healthcare is from professor Evangelos Kanoulas, who founded the startup Ellogon.ai. This startup uses AI tools to determine which cancer patients will benefit from immunotherapy, potentially saving costs and improving outcomes by accurately selecting patients for expensive treatments.

On the ethical side of AI development is professor Somaya Ben Allouch, focusing on how AI can be integrated into healthcare equitably and responsibly. Her work emphasizes the importance of ensuring equal access to healthcare for all groups, using AI to improve care for vulnerable populations. Ben Allouch is part of the AI for Health Equity Lab, which recently received the prestigious ELSA label for its ethical work.

The full article can be read here.