Reducing Doctors' Workload in CT Image Diagnosis
Have you ever received a computed tomography (CT) scan? A CT scan is an examination that uses digital imaging and x-rays to take cross-sectional images of the body, which is especially effective for inspecting organs in the chest, including the heart, aorta, bronchial tubes, and lungs, and those in the abdomen, including the liver and kidneys.
When diagnosing a disease, doctors compare many CT images. In particular, interpreting and diagnosing a group of diseases called diffuse lung diseases, including interstitial pneumonia, emphysema, and other numerous diseases, based on CT images requires extensive knowledge and experience. This has created demand for a technology that efficiently retrieves similar cases from the past.
There are existing technologies that allow a doctor to specify an area of focus in a certain slice image, and retrieve other patients with similar slice images. These technologies have been useful when the abnormal shadows are concentrated in one place, as with early-stage lung cancer. In the case of diffuse lung diseases, however, in which the abnormal shadows are spread across the entire lung, retrievals employing this method could find cases that appeared similar in slice images but would not necessarily look the same in three dimensions. To rule out such cases, doctors had to manually search for similar cases from literature or other sources, taking up a great deal of time.
Segmenting the Organ into Three-dimensional Areas, and Recognizing Abnormal Shadow Candidates Using AI
Fujitsu Laboratories Ltd. has developed a technology to retrieve similar cases in which abnormal shadows have spread in three dimensions from a computed tomography (CT) database of previously taken images*.
When determining similarities for diagnosis, doctors segment the organ into three-dimensional areas, such as periphery, core, top, bottom, left and right, and look at the spread of abnormal shadows in each area. Focusing on the way that doctors do, Fujitsu Laboratories has developed an AI-based technology that can accurately retrieve similar cases in which abnormal shadows have spread in three dimensions. The technology automatically segments the complex interior of the organ into areas through image analysis, and recognizes abnormal shadow candidates in each area using AI.
Specifically, this technology first applies machine learning to recognize abnormal shadow candidates from lung CT images, and estimates the boundaries of the core and the periphery of the lung based on the relatively clear parts of the lung and, in succession, segments the lungs into core and peripheral areas. Next, following the axis of the body up and down, the technology creates histograms of the number of abnormal shadow candidates located in the core and peripheral areas, then looks at the three-dimensional spread of abnormal shadows to retrieve similar cases.
*: Technology to recognize abnormal shadow candidates was jointly developed with Fujitsu R&D Center, Co., Ltd.
Retrieving Similar Cases with an 85% Accuracy Rate, and Having the Possibility of Shortening the Diagnostic Time to One-Sixth
In joint research with Hiroshima University, this technology was tested using CT images of diffuse lung diseases, and as a result, it was able to retrieve similar cases with an accuracy rate of about 85%**. This technology is expected to increase the efficiency and accuracy of diagnostic tasks, while having the possibility of shortening the diagnostic time to as great as one-sixth of previous levels.
This technology could be applied not only to the diagnosis of diffuse lung diseases, but also to other imaging diagnostic techniques, including head CTs and stomach CTs, as well as MRIs (Magnetic Resonance Imaging) and ultrasounds. Fujitsu Laboratories will conduct numerous field trials for a variety of cases to contribute to the increased efficiency of medical care.
**: Joint research with Professor Kazuo Awai of the Department of Diagnostic Radiology, Institute and Graduate School of Biomedical Sciences, Hiroshima University