Russia Develops New Tomographic Image Software

  • This software is developed for the reconstruction of the x-ray tomographic images in real time.
  • The Kremlin’s agenda is to have all the systems in Russia domestically designed and made.
  • Smart Engines have higher speed than the other software on the global market.

Russian company Smart Engines made an announcement via Russian CNews. The company developed new software, called Smart Tomo Engine. This software is developed for the reconstruction of x-ray tomographic images in real time. The software can be applicable in the airport scanning, medical, security, law enforcement and the government agencies.

The X-Ray Images.

The Kremlin’s agenda is to have all the systems in Russia domestically designed and made.

Furthermore, the initial tests of the Smart Tomo Engine software for tomographic reconstruction on the Elbrus-SV system showed that the combination of a Russian processor and software at the early stages of development is more than enough for real-time tomography.

According to the CNews interview with the General Director of the Smart Engines, Vladimir Lazarov:

“In accordance with the European classification of technology readiness, it can be evaluated as TRL7. This means that our system has shown operability and efficiency in conjunction with a real tomograph in operation on a serial computer. The software developed by us is a ready  to implement product that can be integrated both at the stage of developing fundamentally new tomographs, and adapted to existing samples in operation.”

The Smart Tomo Engine software platform provides reconstruction of a three-dimensional digital image of an object based on a set of its transmission tomographic images in the x-ray range.

Currently, the tomography medical industry non-invasive diagnostics based on computed tomography radiography (CT) allows you to form an image of the shape and internal structure of living and non-living objects using x-rays.

To form the three-dimensional structure of the image with x-ray CT, x-ray images taken at different angles are used. A set of projections undergoes special processing, reconstruction, which recreates the internal morphological structure of the object.

The platform is based on a special library of tomographic reconstruction, which, through the API, provides functions for reading tomographic projections, directly tomographic reconstruction using one of the three algorithms, as well as saving the results in the medical DICOM or PNG format. The software also includes a graphical interface for two-dimensional visualization of tomographic images and reconstruction results.

The Russian Academy of Sciences (RAS) consists of the national academy of Russia; a network of scientific research institutes from across the Russian Federation; and additional scientific and social units such as libraries, publishing units, and hospitals.

For layer-by-layer two-dimensional reconstruction, the classical FBP (Filtered Back Projection) algorithm with reverse projection and linear filtering, frequency DFR (Direct Fourier Reconstruction) using fast Fourier transform for filtering and reverse projection, as well as domestic HFBP (Hough FBP), where the Brady algorithm for fast calculation of the Hough transform is used for reverse projection, and the Derish method is used to speed up linear filtering.

Moreover, Smart Engines have higher speed than the other software on the global market. The Smart Engines held trials conjointly with the Russian Academy of Sciences. Tomo Smart Engine works in systems running Russian OS “Elbrus,” Microsoft Windows, Apple macOS and various Linux distributions, and is also compatible with various processor architectures, including VLIW-architecture “Elbrus” and x86-x86-64.

The test reconstruction was performed on synthetic data and real images obtained when shooting the may beetle on a microtomograph of the FEDERAL research center of the Russian Academy of Sciences. Synthetic dataset “Shepp-Logan 3D” was obtained by mathematical modeling using a fan scheme and layer-by-layer calculation of the projection from a three-dimensional Shepp-Logan phantom. The size of the projection made 511х51 pixels.

Only $1/click

Submit Your Ad Here

Christina Kitova

I spent most of my professional life in finance, insurance risk management litigation.

Leave a Reply