When it comes to reading MRI scans, the greatest source of misreadings come from human error.
A new study says the use of artificial intelligence can do away with these errors, all while making the reading time significantly faster than it would be if carried out by a human.
Rapid AI readings
The new research, published in Circulation: Cardiovascular Imaging, describes how analyzing heart function on cardiac MRI scans typically takes about 13 minutes for a human doctor. Using artificial intelligence, the study says, allows for scans to be analyzed in as little as 4 seconds.
"Cardiovascular MRI offers unparalleled image quality for assessing heart structure and function; however, current manual analysis remains basic and outdated," said study author Charlotte Manisty, M.D. Ph.D, in a press release.
"Automated machine learning techniques offer the potential to change this and radically improve efficiency, and we look forward to further research that could validate its superiority to human analysis," she continued.
Time-saving and life-saving AI
In the UK, where the study was carried out, it is estimated that more than 150,000 cardiac MRI scans are performed on a yearly basis.
Due to the high amount of scans, and how much accumulated time could be saved on all of these, the researchers believe that using AI to read scans has the potential to save 54 clinician-days every year at each UK health center where the scans are carried out.
On top of this, the AI has great potential for eradicating human error in the scan reading process.
"Our dataset of patients with a range of heart diseases who received scans enabled us to demonstrate that the greatest sources of measurement error arise from human factors," Manisty explains.
"This indicates that automated techniques are at least as good as humans, with the potential soon to be 'super-human' -- transforming clinical and research measurement precision."
For their research, the scientists trained a machine learning neural network AI to read cardiac MRI scans with the results of almost 600 patients.
The AI was then tested for precision in comparison with an expert and trainee on 110 separate patient cases from multiple centers.
The team of researchers' findings showed that there was no significant difference in accuracy between the AI and human readings, while those of the AI were carried out much faster.