Got Bad Handwriting? Artificial Intelligence Can Help.
Apr 30, 2021
Read Time: 1 min
by Novarad
We’ve already discussed a few ways that artificial intelligence (AI) can help radiologists, but there’s another way that you probably haven’t thought of: reading doctors’ handwriting!
Infamous penmanship and ingenious tech for solving it
The cliché that doctors have terrible handwriting is decades old, and there are a few possible reasons. Bad habits may start in medical school when students frantically try to keep notes during class. In practice, handwriting may deteriorate over the course of the day as hands become tense or cramped from use.
AI could be the solution to illegible writing. Mercy Radiology, a network of 13 clinics in New Zealand, has been using robotic process automation (RPA) to transcribe physician notes, with an accuracy rate of up to 98 percent!
Not only does this method save time for staff, but it also reduces transcription errors. The robots “learn” to recognize letters, and as they consume more and more data, improve their ability to process even the most illegible writing.
AI's reading prowess benefits the bottom line
A robot called “Matilda” now handles 98 percent of Mercy’s invoices automatically. It can accomplish in two hours what takes humans 8-10 hours. With this faster invoicing, the network’s cash flow has improved by $200,000 per month.
But this tech doesn’t just make for a funny story—it could be a lifesaver. One Johns Hopkins study found that 250,000 U.S. deaths per year may be due to medical error. Any technology that reduces the risk of medical errors may also help prevent injuries and even deaths.
Start with AI for free and scan chest CTs for COVID
Ready to take advantage of AI’s potential? Novarad has launched the free COVID-19 AI Diagnostic Assistant to help identify and diagnose patients with COVID-19.
Using sophisticated AI algorithms, this tool can analyze a chest CT scan and produce an accurate, sensitive report via QR code that indicates if the patient's scans are consistent with COVID-19.