1887
banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
Real-time automatic fiducial marker tracking in low contrast cine-MV images
Rent:
Rent this article for
USD
10.1118/1.4771931
/content/aapm/journal/medphys/40/1/10.1118/1.4771931
http://aip.metastore.ingenta.com/content/aapm/journal/medphys/40/1/10.1118/1.4771931

Figures

Image of FIG. 1.
FIG. 1.

Flowchart of the proposed framework for tracking markers in low contrast MV images.

Image of FIG. 2.
FIG. 2.

(a) The input MV image. (b) Histogram of gray levels in the input MV image on the left.

Image of FIG. 3.
FIG. 3.

(a) The partial histogram of Fig. 2(b) . We use the minimum pixel value as the initial position of the kernel, denoted by the leftmost dot. The sequence of kernel locations converges at the local maximum, which corresponds to the foreground region in Fig. 2(a) . (b) Transformation function obtained using the iterative procedure in (a).

Image of FIG. 4.
FIG. 4.

(a) Enhanced image using the proposed method. Arrows highlight the implanted markers. (b) Histogram of the enhanced image.

Image of FIG. 5.
FIG. 5.

Comparison of local entropies obtained from the original images and the contrast-enhanced images of patient 1 (a) and patient 2 (b). The block size represents the length of one side of a square region centered on a fiducial marker.

Image of FIG. 6.
FIG. 6.

(a) MV image of a patient with implanted cylindrical fiducial markers. Expected position of a selected marker is highlighted. (b) We search for marker candidates in a 41 × 41 search region (dotted box). The maker candidates found in the search region are highlighted by black color. (c) Detected marker position is highlighted. (d), (e), (f) are magnified images of (a), (b), (c), respectively.

Image of FIG. 7.
FIG. 7.

Sequence of MV treatment beam imaging. The frames 1, 7, 8, 15, 16, 23, 24, and 25 are shown. These frames are selected to show how the treatment field changed in this MV sequence.

Image of FIG. 8.
FIG. 8.

The number of mean shift iterations for the MV sequence shown in Fig. 7 .

Image of FIG. 9.
FIG. 9.

Manually localized and automatically tracked marker positions for the MV sequence shown in Fig. 7 .

Image of FIG. 10.
FIG. 10.

The timing diagram of our system (four markers).

Image of FIG. 11.
FIG. 11.

Selected frames from a sequence of MV treatment beam images. These frames are selected to show how the MLC blocked region is changed during treatment. (Left column) Contrast-enhanced MV images and (right column) field mask.

Tables

Generic image for table
TABLE I.

Detection rates for the patients studied.

Generic image for table
TABLE II.

Comparison of tracking error and manual localization error.

Generic image for table
TABLE III.

Treatment gantry angles for the patients studied.

Generic image for table
TABLE IV.

Comparison of tracking errors and manual localization errors at different gantry angles.

Loading

Article metrics loading...

/content/aapm/journal/medphys/40/1/10.1118/1.4771931
2012-12-19
2014-04-16
Loading

Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: Real-time automatic fiducial marker tracking in low contrast cine-MV images
http://aip.metastore.ingenta.com/content/aapm/journal/medphys/40/1/10.1118/1.4771931
10.1118/1.4771931
SEARCH_EXPAND_ITEM