Estonia

FCT PixelShine

NEW

Imaging processing software provides clear and natural CT images through deep learning.

The content on this page is intended to healthcare professionals and equivalents.

  • High quality images without the blurring
  • More natural look CT image
  • Process a typical study in less than a minute
  • Dedicated hardware is not required
  • Approval (Deep Learning based image processing) Manufacturer : AlgoMedica, Distributed by FUJIFILM

Promotion Video

Deep learning technology was used to train FCT PixelShine to improve the reconstruction of low dose CT images.
During the training, an extensive number of noisy low and high quality standard-dose CT image pairs were presented to FCT PixelShine.
After the training, FCT PixelShine successfully acquired the mapping function between noisy low dose images and corresponding high quality standard-dose images.

Difference from Iterative Reconstruction

Reduced blurring

The above NPS curves indicates that FCT PixelShine reduces noise without blurriness. On the other hand, Iterative Reconstructions tends to introduce more blurriness by more noise reduction.
NPS curve indicates noise magnitude by area under the curve and noise texture by its curve shape.

(Comparison based on NPS curve peak)

More natural look

High noise CT images processed by FCT PixelShine look more natural, without blur, and in evaluation were preferred by radiologists to the “waxy” or smoothed images that can be created by Iterative Reconstruction algorithms offered by major CT scanner vendors.

Abdomen

FBP

IR

Lung HRCT

FBP

IR

Less Processing time

FCT PixelShine is able to process a typical study in less than a minute.

Dedicated hardware is not required

Runs on an off-the-shelf server for easy installation(Supports only FCT series)

Products Name : FCT PixelShine(Model name : PixelShine)