AI-based 3D CT scans have a high level of precision when it comes to measuring how much the coronavirus has infiltrated the respiratory system. (image: MEDICAL IP)
SEOUL, April 2 (Korea Bizwire) — South Korea has introduced the country’s first commercial artificial intelligence (AI) medical imaging analysis technology that can accurately determine the severity of pneumonia in a coronavirus patient.
The MEDIP PRO Technology developed by MEDICAL IP, an AI medical imaging analysis firm, is currently being used in 24 countries around the world to test for the coronavirus, just two weeks after the technology was first revealed to the world free-of-charge.
The technology recreates two-dimensional images such as CT and MRI scans into 3D, allowing doctors to easily and accurately determine the symptoms and distinguish the organs, which is sometimes difficult when using a 2D image.
This technology is garnering attention around the world because of the coronavirus outbreak, allowing doctors to determine the severity of pneumonia of a coronavirus patient more easily than an X-ray scan, using its signature deep learning technology.
A joint research team from Seoul National University Hospital, the First Hospital of Lanzhou University in China, and Self-Defense Forces Central Hospital in Japan reported in a recent study that MEDIP PRO is very effective in the diagnosis of pneumonia among coronavirus patients.
The research team recreated lung CT scans of 17 coronavirus patients (14 South Koreans, 3 Chinese) into AI-based 3D footage for analysis, and compared them with the analysis of X-ray scans by eight radiologists.
The 3D CT scans found 186 lesions in the lungs, while the X-ray scans revealed only 19 of them. In addition, only 55.8 percent of the pneumonia symptoms found on the 3D scan were seen on the X-ray scan.
The assessment showed that AI-based 3D CT scans have a high level of precision when it comes to measuring how much the coronavirus has infiltrated the respiratory system.
Overall, the level of sensitivity and specificity of an X-ray scan of a coronavirus patient only reaches 25 and 90 percent of a 3D CT scan, respectively.
Sensitivity refers to the chance of detecting positive cases of coronavirus, while specificity detects negative cases.
“The study revealed that measuring the size and mass of pneumonia using 3D CT imaging plays an important role in the a precise diagnosis of the symptom’s severity,” said Dr. Yun Soon-ho from Seoul National University Hospital.
“This technology will be very useful in times like these, when a global pandemic is at hand and there are only a limited amount of medical resources available.”
H. M. Kang (hmkang@koreabizwire.com)
