Publication detail

Segmentation of basic tissues for assessing noise in iteratively reconstructed MDCT data

WALEK, P. JAN, J.

Original Title

Segmentation of basic tissues for assessing noise in iteratively reconstructed MDCT data

Czech Title

Segmentace základních struktur pro hodnocení šumu v iterativně rekonsruovaných MDCT datech

English Title

Segmentation of basic tissues for assessing noise in iteratively reconstructed MDCT data

Type

conference paper

Language

en

Original Abstract

This paper describes fully automatic, fast and reliable algorithm for segmentation of basic tissues in iteratively reconstructed multidetector X-ray CT thoracic, abdominal and brain images. Iterative reconstruction becomes a part of routine in medical imaging and assessing of noise properties of iteratively reconstructed data in diverse tissues is required. Proposed segmentation is based on simple thresholding with fully automatic deter-mination of thresholds which are derived from the most significant peaks in histograms of Hounsfield units. As thresholding is unable to distinguish trabecular bone and soft tissue additional segmentation based on classification is proposed.

Czech abstract

Příspěvek popisuje plně automatickou, rychlou a spolehlivou metodu pro segmentaci základních tkáňových struktur v iterativně rekonstruovaných CT obrazech hrudníku, břicha a hlavy. Iterativní rekonstrukce se stávají nástrojem běžně využívaným v klinické praxi a je nutné provést hodnocení vlastností šumu v takto rekonstruovaných datech. Aby bylo možné hodnotit vlastnosti šumu v různých tkáních, je navržen segmentační algoritmus založený na prahování s automatickým stanovením prahů z histogramu. Prahování není schopno rozlišit mezi trabekulární částí kosti a měkkou tkání a z tohoto důvodu je navržena další segmentace založena na klasifikaci.

English abstract

This paper describes fully automatic, fast and reliable algorithm for segmentation of basic tissues in iteratively reconstructed multidetector X-ray CT thoracic, abdominal and brain images. Iterative reconstruction becomes a part of routine in medical imaging and assessing of noise properties of iteratively reconstructed data in diverse tissues is required. Proposed segmentation is based on simple thresholding with fully automatic deter-mination of thresholds which are derived from the most significant peaks in histograms of Hounsfield units. As thresholding is unable to distinguish trabecular bone and soft tissue additional segmentation based on classification is proposed.

Keywords

X-ray Computed Tomography, Tissue Segmentation, Iterative Reconstruction, Dose Reduction.

RIV year

2012

Released

09.09.2012

ISBN

978-1-61804-119-7

Book

Proceedings of the WSEAS International Conference on Visualization, Imaging and Simulation (Vis'12)

Pages from

211

Pages to

216

Pages count

6

BibTex


@inproceedings{BUT93827,
  author="Petr {Walek} and Jiří {Jan}",
  title="Segmentation of basic tissues for assessing noise in iteratively reconstructed MDCT data",
  annote="This paper describes fully automatic, fast and reliable algorithm for segmentation of basic tissues in
iteratively reconstructed multidetector X-ray CT thoracic, abdominal and brain images. Iterative reconstruction
becomes a part of routine in medical imaging and assessing of noise properties of iteratively reconstructed data
in diverse tissues is required. Proposed segmentation is based on simple thresholding with fully automatic deter-mination of thresholds which are derived from the most significant peaks in histograms of Hounsfield units. As
thresholding is unable to distinguish trabecular bone and soft tissue additional segmentation based on classification
is proposed.",
  booktitle="Proceedings of the WSEAS International Conference on Visualization, Imaging and Simulation (Vis'12)",
  chapter="93827",
  howpublished="print",
  year="2012",
  month="september",
  pages="211--216",
  type="conference paper"
}