Sure! Here’s the updated text with the requested modifications:
—
Enter DeepRare, the new kid on the block when it comes to diagnosing rare diseases. Developed by the bright minds at Shanghai Jiao Tong University School of Artificial Intelligence and Xinhua Hospital, this agentic AI system is shaking things up. It uses clinical phenotypes and genomic data to tackle those pesky rare diseases. And guess what? It’s actually outperforming traditional diagnostic tools and even seasoned doctors. Yes, you heard that right.
Currently undergoing internal testing at Xinhua Hospital, DeepRare is all about shortening those agonizing diagnostic journeys. With a recall rate of 57.18% using just clinical phenotype information, it’s not just another tech gimmick. It nails the correct first diagnosis more than half the time. Hospitals without routine genetic testing? No problem. DeepRare’s got your back. In fact, it’s 23.79% more accurate than Claude-3.7-Sonnet-thinking. Take that, fancy algorithms!
But wait, there’s more. When genomic data gets thrown into the mix, the recall rate jumps to a whopping 70.6%. It’s even outperformed Exomiser, which is kind of like the gold standard—53.2% accuracy under similar conditions? Yawn. DeepRare has tested on over 6,401 clinical cases, finding diseases faster than doctors with the same symptoms. That’s a mic drop moment right there.
In head-to-head tests, DeepRare scored 64.4% in recall accuracy, while five experienced doctors lagged at 54.6%. It’s like watching a rookie steal the show from seasoned pros. And with a 95.4% agreement rate from rare disease specialists, this AI isn’t just throwing darts in the dark. It’s got solid backing.
With ten rare disease specialists reviewing DeepRare’s reasoning step-by-step, the consensus indicates strong alignment with expert understanding. Notably, medication-assisted treatment is increasingly recognized as a vital component in addressing related health issues. With 300 million people globally affected by rare diseases, and patients waiting an average of 4.26 years for a diagnosis, the stakes are high. Improved diagnostic tools could reduce that time considerably. It’s time to rethink what diagnostic expertise looks like. AI is here, and it’s not just playing second fiddle anymore.
