This news release is issued by FUJIFILM Corporation in Japan.
Fujifilm makes no representation that products on this news release are commercially available in all countries and regions. Approved uses of products vary by country and region.
Specifications and appearance of products are subject to change without notice.
TOKYO, May 19, 2020 — FUJIFILM Corporation (President: Kenji Sukeno) is commencing a research study to develop Artificial Intelligence (AI)-based technology to aid in the diagnosis and treatment assessment of patients with COVID-19-induced pneumonia. The technology for quantifying the lesions of interstitial pneumonia*1, co-developed with Kyoto University (the Department of Respiratory Medicine, Graduate School of Medicine, Professor Toyohiro Hirai), will be applied to the project. The company will now embark on a joint research study with local medical institutions treating COVID-19 patients, starting with the Kanagawa Cardiovascular and Respiratory Center (Yokohama, Japan).
The spread of the novel coronavirus, which causes COVID-19, has emerged as a serious issue around the world. The world has yet to see clear judging criteria for determining the effectiveness of various treatment options, currently explored by doctors. In order to confirm the progression of pneumonia and the effectiveness of treatments, doctors need to examine hundreds of chest CT images for each patient to visually check the characteristics of ever-changing lesions and it puts a serious strain on specialists. There are expert opinions that COVID-19-induced pneumonia presents similarly to interstitial pneumonia in diagnostic images and has diverse variations in lesion patterns. Fujifilm’s CT quantification technology for interstitial pneumonia is powered by an AI-based software that examines CT images to identify bronchi, blood vessels and normal lungs in lung field*2 as well as seven types of lesions such as reticular opacities, ground-glass opacities and honeycomb lungs*3, and automatically carries out categorization and measurement to quantify lesions of interstitial pneumonia. It also divides the lung field into 12 zones*4 and shows the volume and ratio of lesions for each of the zones so that clinicians can examine the distribution and progression of lesions within the lung field in details.
Fujifilm began collaborating with Kyoto University in the spring of 2018 and applied Fujifilm’s AI technology to categorize and quantify lesions of interstitial pneumonia to case data held by Kyoto University. The cycle of evaluating its identification performance and providing feedback for improvement was repeated numerous times for enhancement, resulting in an advanced level of precision in lesion type identification.
Fujifilm will apply this CT quantification technology for interstitial pneumonia to develop the technology that helps evaluate the progress of patients with COVID-19-induced pneumonia and determine the effectiveness of treatments. In addition, the technology is expected to contribute to accelerating the development and evaluation of drug candidates for treating pneumonia induced by COVID-19.
Fujifilm has been working on developing AI technology that can be used for assisting medical diagnostic imaging, facilitating workflow at the medical frontline and delivering maintenance services for medical equipment. The company’s AI technology for use in these domains is now marketed under the brand name, “REiLI.” In order to deliver solutions that meet various needs and workflows at the medical frontline, Fujifilm will continue to carry out technological development to achieve fast-paced development of solutions aimed at assisting doctors in diagnostic imaging and streamlining their workflow.
(Under development) CT images of a patient who has developed pneumonia as a complication of COVID-19 and analysis results
A: CT image’s axial view (transverse plane), B: Sagittal view (longitudinal plane)
C: Coronal view (vertical plane into ventral and dorsal, D: 3D image, each showing the results of lesion identification
E: Graphs showing the presence of a specified type of lesions by zone and its volume
Corporate Communications Division