Back

Computed Tomography 3D Segmentation Based on Gray-Level Histograms

Guimerans-Otero, C. T.; Martin-Rodriguez, F.; Isasi de Vicente, F.; Fernandez-Barciela, M.

2022-08-30 radiology and imaging
10.1101/2022.08.26.22279285 medRxiv
Show abstract

3D imaging technologies like CT (Computed Tomography) and MRI (Magnetic Resonance Imaging) have represented a great advance for diagnosis. These studies are composed of many 2D grayscale images that represent slices of patients body, normally orthogonal to body main axis (from head to feet, normally axis Z in Cartesian representation). These slices (normally called axial slices) are good for many diagnosis issues, but sometimes it is interesting to consider the whole study as a 3D volume. This paper is about segmenting 3D volumes obtained from medical studies (mainly CT). We base ourselves on studying the statistical distribution of gray levels so that we can segment different tissues and treating them as separate 3D objects. Note than in a CT image, gray level is basically proportional to tissue density and this technique should be good to distinguish hard tissues like bones or teeth from soft ones like muscles or skin.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.