Samsung applies AI to medical imaging
The South Korean tech giant, together with its medical device affiliate Samsung Medison, showcased various types of diagnostic imaging software at the Radiological Society of North America 2018 Annual Meeting (RSNA 2018) in Chicago.
Offerings include those for ultrasound, digital radiography, computed tomography, and magnetic resonance imaging (MRI).
Samsung’s S-Detect for Breast, which analyses for breast lesions using ultrasound images, and provides standardized reports and classifications, was showcased at RSNA 2018.
Samsung said the software increased the accuracy of diagnosis from 0.83 to 0.87 for doctors with four years or less experience, according to a study published by radiology professor Tommaso Bartolotta from University of Palermo in Italy.
Its AI software for digital radiography, dubbed SimGrid, provides images of lung tissue that are obscured by bone in chest X-Rays.
Its AI software called Auto Lung Nodule Detection, like its name sake, detects lung nodules more accurately.
Samsung also introduced a mobile stroke unit for computed tomography, and is developing AI technology that will allow MRIs to display information such as knee cartilage thickness and images of knee arthritis.
More companies are also looking to apply AI to healthcare. Last week, LG CNS and Lunit announced that they plan to apply AI and cloud for analysis of X-Ray videos of lungs.
Samsung is beefing up its AI capabilities by launching research centres focusing on healthcare around the world, having opened its seventh one in Montreal last month.
Samsung is planning to invest US$22 billion in AI, 5G and emerging technology going forward.