Fujitsu and Tokyo Shinagawa Hospital to Develop AI Technology to Support Diagnosis of COVID-19 Pneumonia

Fujitsu and Tokyo Shinagawa Medical center to acquire an AI-based know-how to facilitate the analysis of COVID-19 pneumonia. The new AI know-how will learn from past CT scans of COVID-19 pneumonia people, cutting down the load on medical doctors making diagnoses. Fujitsu will consider potential commercialization of this know-how for broader healthcare applications

Fujitsu and Tokyo Shinagawa Medical center this week declared the start of a joint R&D venture for AI know-how to aid diagnostic imaging by way of upper body CT (Computed Tomography), which signifies a promising candidate for the successful analysis of COVID-19 pneumonia.

Picture credit history: Pixabay (Absolutely free Pixabay license)

The recently-proposed know-how supports medical doctors doing diagnostic imaging on people suspected of acquiring COVID-19 pneumonia, presenting the probability of infection through a numerical, 3-dimensional visualization of the spread of shadows in the lungs employing upper body CT imaging.

Implementing AI to this analysis could appreciably minimize the load on medical doctors who diagnose COVID-19 pneumonia by automating a system that typically needs the visible confirmation of hundreds of upper body CT pictures for each individual.

Fujitsu and Tokyo Shinagawa Medical center foresee that the process will supply early detection of cases of COVID-19 pneumonia based on upper body CT picture conclusions, even in cases in which the chance of infection is established to be small on initial examination.

Figure 1: Envisioned flow of novel coronavirus infection tests (AI diagnostic imaging created through this joint R&D exertion). Picture credit history: Fujitsu

This joint research will enrich AI diagnostic aid know-how for novel coronavirus pneumonia, and Fujitsu in the end aims to commercialize the know-how as a healthcare answer for frontline healthcare experts.


When managing people who are strongly suspected of acquiring a novel coronavirus infection, the analysis and therapy strategy are established holistically based on PCR examination final results as well as other examination final results together with blood assessments and upper body CT examination (Figure 1). Even if the PCR examination is destructive, the analysis of new-onset coronavirus pneumonia can be produced by other assessments, and imaging with upper body CT is an crucial consideration.

Medical practitioners make lung condition diagnoses based on the qualities of the shadows in the lesion on the patient’s upper body CT picture. In addition to figuring out abnormalities, they visually verify hundreds of upper body CT pictures for each individual to fully grasp the 3-dimensional distribution of shadows all over the lung. A expanding desire for systems that minimize the load on medical doctors and aid fast conclusion-making has accompanied the spread of the COVID-19 virus. It is also crucial to discover novel coronavirus pneumonia from a CT scan of the patient’s upper body if there is a small probability of a novel coronavirus infection all through the examination and PCR tests is not done.

In reaction to the will need for diagnostic aid, Fujitsu and Tokyo Shinagawa Medical center will jointly embark on the progress of AI-assisted diagnostic imaging know-how for upper body CT scans, which are deemed to be an successful strategy in diagnosing novel coronavirus pneumonia.

About the joint progress of AI imaging aid know-how

Past CT picture information of COVID-19 pneumonia people from Tokyo Shinagawa Medical center will support in the progress of AI know-how that detects irregular shadow patterns in the lungs. As the AI learns from this information about the chance of COVID-19 pneumonia, Fujitsu and Tokyo Shinagawa Medical center will work collectively to evaluate the usefulness of the know-how.

When diagnosing COVID-19 pneumonia, patterns of irregular opacities in the lungs as well as the spread of shadows throughout the complete lung is crucial information.

Designs of irregular shadows are detected employing AI created by Fujitsu Laboratories, Ltd. The lung is divided into four regions on the CT picture: the proper lung periphery, the proper lung center, the remaining lung center, and the remaining lung periphery. The distribution of shadows in the vertical route in every single spot is shown in a histogram(1). This will permit the progress of new AI that can quantify the 3-dimensional spread of shadows, when employing the detected irregular shadow patterns and shadow distributions to discover COVID-19 pneumonia.

While demonstrating the chance of COVID-19 infection with the AI know-how, the intention is to shorten the sum of time medical doctors devote visually confirming the 3-dimensional spread of the shadow from hundreds of upper body CT pictures and to permit even non-specialists to proficiently diagnose COVID-19 pneumonia.

Future Programs

By means of this joint research and progress, Fujitsu and Tokyo Shinagawa Medical center intention to supply know-how that leverages a wide variety of information for analysis of COVID-19 pneumonia to attain enhanced diagnostic aid from upper body CT picture analysis.

Fujitsu is looking at the commercialization of this know-how as a healthcare answer. By linking it with electronic healthcare file information, the purpose is to aid analysis by medical doctors based on upper body CT pictures, when also expanding the software spot of this know-how.

Tokyo Shinagawa Medical center aims to add to society by integrating many reports carried out in the medical center and employing them for analysis and therapy of novel coronavirus pneumonia.

(1) Histogram

A form of statistical graph that allows visible recognition of information distribution conditions.

About Fujitsu Ltd

Fujitsu is the major Japanese information and interaction know-how (ICT) company offering a entire vary of know-how products, remedies and expert services. Somewhere around a hundred thirty,000 Fujitsu persons aid clients in more than 100 nations. We use our knowledge and the electricity of ICT to shape the potential of society with our clients. Fujitsu Constrained (TSE:6702) claimed consolidated revenues of three.9 trillion yen (US$35 billion) for the fiscal calendar year ended March 31, 2020. For more information, please see

Source: ACN Newswire