«A new study shows how training deep-learning models on patient outcomes could help reveal gaps in existing medical knowledge.»

«In the last few years, research has shown that deep learning can match expert-level performance in medical imaging tasks like early cancer detection and eye disease diagnosis. But there’s also cause for caution. Other research has shown that deep learning has a tendency to perpetuate discrimination. With a healthcare system already riddled with disparities, sloppy applications of deep learning could make that worse.»

«Now a new paper published in Nature Medicine is proposing a way to develop medical algorithms that might help reverse, rather than exacerbate, existing inequality. The key, says Ziad Obermeyer, an associate professor at UC Berkeley who oversaw the research, is to stop training algorithms to match human expert performance.»

Article written by Karen Hao



MIT Technology Review