Limits...
Segmentation of Pulmonary Vascular Trees from Thoracic 3D CT Images.

Shikata H, McLennan G, Hoffman EA, Sonka M - Int J Biomed Imaging (2009)

Bottom Line: A quantitative validation was performed with more than 1000 manually identified points selected from inside the vessel segments to assess true positives (TPs) and 1000 points randomly placed outside of the vessels to evaluate false positives (FPs) in each case.On average, for both the high and low volume lung images, 99% of the points was properly marked as vessel and 1% of the points were assessed as FPs.Our hybrid segmentation algorithm provides a highly reliable method of segmenting the combined pulmonary venous and arterial trees which in turn will serve as a critical starting point for further quantitative analysis tasks and aid in our overall goal of establishing a normative atlas of the human lung.

View Article: PubMed Central - PubMed

Affiliation: Iowa Institute for Biomedical Imaging, The University of Iowa, Iowa City, IA 52242, USA.

ABSTRACT
This paper describes an algorithm for extracting pulmonary vascular trees (arteries plus veins) from three-dimensional (3D) thoracic computed tomographic (CT) images. The algorithm integrates tube enhancement filter and traversal approaches which are based on eigenvalues and eigenvectors of a Hessian matrix to extract thin peripheral segments as well as thick vessels close to the lung hilum. The resultant algorithm was applied to a simulation data set and 44 scans from 22 human subjects imaged via multidetector-row CT (MDCT) during breath holds at 85% and 20% of their vital capacity. A quantitative validation was performed with more than 1000 manually identified points selected from inside the vessel segments to assess true positives (TPs) and 1000 points randomly placed outside of the vessels to evaluate false positives (FPs) in each case. On average, for both the high and low volume lung images, 99% of the points was properly marked as vessel and 1% of the points were assessed as FPs. Our hybrid segmentation algorithm provides a highly reliable method of segmenting the combined pulmonary venous and arterial trees which in turn will serve as a critical starting point for further quantitative analysis tasks and aid in our overall goal of establishing a normative atlas of the human lung.

No MeSH data available.


Related in: MedlinePlus

Two examples of segmentation results. (a) and (c) are the results from a TLC scan. (b) and (d) are from the associated FRC scan of the same subject. All images are shown at the same scale. Note the breathing-related changes of lung vasculature caused by regional lung parenchymal expansion between the TLC and FRC lung volumes.
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2801012&req=5

fig8: Two examples of segmentation results. (a) and (c) are the results from a TLC scan. (b) and (d) are from the associated FRC scan of the same subject. All images are shown at the same scale. Note the breathing-related changes of lung vasculature caused by regional lung parenchymal expansion between the TLC and FRC lung volumes.

Mentions: Figure 8 shows volume-rendered images of the segmentation results from both TLC and FRC scans of one subject. Both results include peripheral thin segments close to the pleural surface. Since the scale of these images is the same, they also show how the lung geometry changes between the two volumes scanned. Table 1 shows a summary of the TP rate for both TLC and FRC scans. TP rate for TLC scans was 99.6% from the total of 16933 validation points and 99.5% from 15,281 points in right and left lungs, respectively. Similarly, TP rate for FRC scans was 99.0% from 13356 points and 98.2% from 10646 points in right and left lungs, respectively. Since the lung volume of FRC scans is less than that of TLC scans, there were less validation points for FRC data sets. Also, especially in the FRC scans due to increased compliance, motion artifacts may have contributed to a small increase in the error rate. Table 2 shows a summary of the FP rates for both TLC and FRC scans. A total of 11,925 points and 9620 points were used in right and left lung of the TLC scans, respectively, yielding FP rates of 1.21% in the right and 0.99% in the left lung. For FRC scans, a total of 11752 points in right lung and 8722 points in left lung were used and FP rates of 0.96% and 0.88% in the right and left lungs were obtained.


Segmentation of Pulmonary Vascular Trees from Thoracic 3D CT Images.

Shikata H, McLennan G, Hoffman EA, Sonka M - Int J Biomed Imaging (2009)

Two examples of segmentation results. (a) and (c) are the results from a TLC scan. (b) and (d) are from the associated FRC scan of the same subject. All images are shown at the same scale. Note the breathing-related changes of lung vasculature caused by regional lung parenchymal expansion between the TLC and FRC lung volumes.
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2801012&req=5

fig8: Two examples of segmentation results. (a) and (c) are the results from a TLC scan. (b) and (d) are from the associated FRC scan of the same subject. All images are shown at the same scale. Note the breathing-related changes of lung vasculature caused by regional lung parenchymal expansion between the TLC and FRC lung volumes.
Mentions: Figure 8 shows volume-rendered images of the segmentation results from both TLC and FRC scans of one subject. Both results include peripheral thin segments close to the pleural surface. Since the scale of these images is the same, they also show how the lung geometry changes between the two volumes scanned. Table 1 shows a summary of the TP rate for both TLC and FRC scans. TP rate for TLC scans was 99.6% from the total of 16933 validation points and 99.5% from 15,281 points in right and left lungs, respectively. Similarly, TP rate for FRC scans was 99.0% from 13356 points and 98.2% from 10646 points in right and left lungs, respectively. Since the lung volume of FRC scans is less than that of TLC scans, there were less validation points for FRC data sets. Also, especially in the FRC scans due to increased compliance, motion artifacts may have contributed to a small increase in the error rate. Table 2 shows a summary of the FP rates for both TLC and FRC scans. A total of 11,925 points and 9620 points were used in right and left lung of the TLC scans, respectively, yielding FP rates of 1.21% in the right and 0.99% in the left lung. For FRC scans, a total of 11752 points in right lung and 8722 points in left lung were used and FP rates of 0.96% and 0.88% in the right and left lungs were obtained.

Bottom Line: A quantitative validation was performed with more than 1000 manually identified points selected from inside the vessel segments to assess true positives (TPs) and 1000 points randomly placed outside of the vessels to evaluate false positives (FPs) in each case.On average, for both the high and low volume lung images, 99% of the points was properly marked as vessel and 1% of the points were assessed as FPs.Our hybrid segmentation algorithm provides a highly reliable method of segmenting the combined pulmonary venous and arterial trees which in turn will serve as a critical starting point for further quantitative analysis tasks and aid in our overall goal of establishing a normative atlas of the human lung.

View Article: PubMed Central - PubMed

Affiliation: Iowa Institute for Biomedical Imaging, The University of Iowa, Iowa City, IA 52242, USA.

ABSTRACT
This paper describes an algorithm for extracting pulmonary vascular trees (arteries plus veins) from three-dimensional (3D) thoracic computed tomographic (CT) images. The algorithm integrates tube enhancement filter and traversal approaches which are based on eigenvalues and eigenvectors of a Hessian matrix to extract thin peripheral segments as well as thick vessels close to the lung hilum. The resultant algorithm was applied to a simulation data set and 44 scans from 22 human subjects imaged via multidetector-row CT (MDCT) during breath holds at 85% and 20% of their vital capacity. A quantitative validation was performed with more than 1000 manually identified points selected from inside the vessel segments to assess true positives (TPs) and 1000 points randomly placed outside of the vessels to evaluate false positives (FPs) in each case. On average, for both the high and low volume lung images, 99% of the points was properly marked as vessel and 1% of the points were assessed as FPs. Our hybrid segmentation algorithm provides a highly reliable method of segmenting the combined pulmonary venous and arterial trees which in turn will serve as a critical starting point for further quantitative analysis tasks and aid in our overall goal of establishing a normative atlas of the human lung.

No MeSH data available.


Related in: MedlinePlus