AI beats radiographers in early detection of lung cancer
Google’s AI model spotted subtle lung lesions that five out of six radiologists missed
Warnings of the extent to which artificial intelligence (AI) will infiltrate our daily lives in the near future abound — from mass job insecurity, to romantic relationships and even our sex lives. However, it’s not all bad. Part of AI’s realm of infiltration is health care — and possibly for the good.
At the annual I/O Google conference held at its headquarters on May 7-9, AI was a top priority. Among some of the inroads the tech giant has made in the field of AI, Google announced that it has created AI that can potentially identify lung cancer a year before a human doctor can.
Using a computed tomography (CT) scan, Google’s AI model spotted subtle lung lesions that five out of six radiologists missed. Spotting lung cancer this far in advance has life-saving potential and can increase the survival rate of patients by 40%.
In a different trial being carried out by UK start-up Medopad in conjunction with Chinese tech company Tencent, inroads are being made into detecting Parkinson’s disease. Using an app powered by AI, researchers are attempting to detect Parkinson’s within minutes by having patients carry out a series of physical tests via an app.
According to The Telegraph, the trial is being carried out at a private mental health clinic in the UK with 40 patients. Data collected from the tests the users complete will be analysed to diagnose Parkinson’s within minutes or help patients monitor the condition after they’ve been diagnosed.
If successful, the app could be rolled out for use in hospitals before the end of the year and will not only result in swift identification but will also mean that tests can be carried out in the comfort of a patient’s home.