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.
A planning quality evaluation tool for prostate adaptive IMRT based on machine learning
Rent:
Rent this article for
USD
10.1118/1.3539749
/content/aapm/journal/medphys/38/2/10.1118/1.3539749
http://aip.metastore.ingenta.com/content/aapm/journal/medphys/38/2/10.1118/1.3539749
View: Figures

Figures

Image of FIG. 1.
FIG. 1.

The relationship between the dose and the distance-to-target for voxels of a patient rectum. (a) shows a scatter plot of the dose at each individual voxel. vs. its distance-to-target; (b) presents the cumulative DVH of the rectum; (c) plots the cumulative DTH of the rectum.

Image of FIG. 2.
FIG. 2.

Results of PCA applied to the training database. (a)–(d) show the statistical dispersion of the DVH/DTH for the bladder/rectum from 198 plans using the box-and-whisker plot: box (central 50%), vertical line (lower and upper 25%), and plus (outliers). (e)–(h) present with the Scree plot the truncated principle component scores (PCS) that account for 90% variance.

Image of FIG. 3.
FIG. 3.

Reconstruction of the bladder DVH curve using truncated PCS. The left figure shows contribution from the first two PCS. The right illustrates the reconstruction of the DVH curve compared with the original data.

Image of FIG. 4.
FIG. 4.

Comparison of fitting errors using MVNLR and SVR. The fitting error in SVR is managed by the tolerance tube: , which is chosen empirically to approximate the average fitting error using MVNLR

Image of FIG. 5.
FIG. 5.

Model evaluation on rectum using 14 cases outside the training pool. For 11 out of 14 plans, the DVHs generated by the treatment planning system (TPS) are within the prediction bands. The “triangle” shows the result after replanning for case #13.

Image of FIG. 6.
FIG. 6.

Model evaluation on bladder using 14 cases outside the training pool. Except for case #13, the DVHs generated by the treatment planning system (TPS) agree with the prediction. The triangle shows the result after replanning for case #13.

Image of FIG. 7.
FIG. 7.

The correlations among the bladder volume, its first DTH PCS, and its first DVH PCS. The left figure clearly indicates that an increase in bladder volume will inversely increase , the mean distance variance. The right indicated the trend between and the mean dose variance .

Loading

Article metrics loading...

/content/aapm/journal/medphys/38/2/10.1118/1.3539749
2011-01-11
2014-04-24
Loading

Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: A planning quality evaluation tool for prostate adaptive IMRT based on machine learning
http://aip.metastore.ingenta.com/content/aapm/journal/medphys/38/2/10.1118/1.3539749
10.1118/1.3539749
SEARCH_EXPAND_ITEM