AiCE (Advanced Intelligent Clear-IQ Engine)

2020: Start of an AI system

At the end of July, 2020 Canon Medical announced the beginning of use of an AI system for recovery of low-quality MRT-pictures of AiCE. It supplements a technology platform of the company on the basis of deep learning and is compatible to the Vantage Orian 1.5T and Vantage Galan 3T MR MR-systems.

According to the press release, the AiCE system uses algorithms of deep learning to distinguish true signals from noise. Suppressing noise and at the same time strengthening a signal, it is capable to recover images and to increase quality of indistinct pictures. AiCE allows to receive high-quality images and to smoothly integrate new MR-scanners into daily practice.

Quality improvement of MRT-pictures

Vantage Orian was developed for performance improvement, ensuring comfort of patients and the help in adoption of clinical solutions, - the managing director of Canon Medical Systems USA Jonathan Furuyama explained in the press release. - Now, having added the AiCE system, we expanded possibilities of MRT. Using possibilities of AI for routine visualization, we allowed our clients to resort to techniques which were considered as impractical in the clinical relation earlier.

The AiCE system was trained using a huge number of high-quality images. Differentiating signals and noise, AiCE allows to carry out expanded scanning and reconstruction of images with preserving of the correct anatomical structures. At the same time AiCE is considered the first-ever system of reconstruction with algorithms of deep learning which is completely integrated directly into protocols of MR-scanners. It allows to provide continuous workflow.

After zasleplenny comparison of images which were recovered by AiCE with images which were recovered by normal methods, doctors came to a conclusion that AiCE provides better and accurate pictures.^[1]

2019: Announcement of technology

At the beginning of March, 2019 the Canon Medical Systems company provided technology of reconstruction of images using deep convolution neural network (DCNN) for a computer tomography.

The technology which received the name AiCE (Advanced Intelligent Clear-IQ Engine) applies deep learning to removal of "noise" in pictures of two productive scanners of Canon — Acquilion Precision and Aquilion One Gensis. It allows to reduce a radiation dose by 20–40%, depending on area of a body, or to receive more high-quality image at the same dose. Besides, the new technology allows to calculate superthin cuts at very high image quality and at very low dose that is extremely useful for such scanners as Aquilion Precision.

Canon provided technology of reconstruction of KT-pictures through neural network

The program studies at high-quality images from databases of real patients, providing optimal quality. Then the received results are checked and loaded into scanners to provide rather fast speed of recovery of images in clinical conditions. By the beginning of March, 2019 the AiCE technology is available on the Aquilion P platform.

In addition to AiCE, Canon Medical Systems also provided new features of Global Illumination for the modular platform of viewing images Vitrea Advanced Visualization. The technology provides delivery of photorealistic three-dimensional images for studying of human anatomy, allowing users to consider in detail fine details, for example sinews. It is supposed that the new feature will allow often cooperating doctors of different specialties, for example, to oncologists and surgeons, it is better to understand each other. Besides, this function will give to doctors the chance in detail to explain to patients their status and also will provide additional confirmation during the opening and forensic medical examination.^[2]

Notes