Validation of image registration and fusion of MV CBCT and planning CT for radiotherapy treatment planning

Abstract

In areas like adaptive therapy, multi-phase radiotherapy, and single fraction palliative treatment or in the treatment of patients with metal implants where megavoltage(MV) CT could be considered as a treatment planning modality, the reduced contrast in the MV CT images could lead to limited accuracy in localization of the structures. This would affect the precision of the treatment. In this study, as an extension our previous work on bespoke MV cone beam CT (MV CBCT), we propose to register the MV CBCT with kilovoltage (kV) CT for treatment planning. The MV CBCT images registered with kV CT would be effective for treatment planning as it would account for the inadequate soft tissue information in the MV CBCT and would allow comparison of changes in patient dimensions and assist in localization of the structures. The intensity based registration algorithm of the BrainSCAN therapy planning software was used for image registration of the MV CBCT and kV CT images. The accuracy of the registration was validated using qualitative and quantitative measures. The effect of image quality on the level of agreement between the contouring done on both the MV CBCT and kV CT was assessed by comparing the volumes of six structures delineated. To assess the level of agreement between the plans after the registration, two independent plans were generated on the MV CBCT and the planning CT using the posterior fossa of the skull as the target. The dose volume histograms and conformity indices of the plans were compared. The results of this study show that treatment planning with MV CBCT images would be effective, using additional anatomical structure information derived from registering the MV CBCT image with a standard kVCT.

Estimation of the uncertainty of elastic image registration with the demons algorithm

“…In the future, elastic registration may play a role as well and therefore it is important to evaluate its accuracy. Various methods to validate image registration have been proposed, such as tracking of landmarks (Coselmon et al 2004, Brock et al 2005, Rietzel and Chen 2006, Heath et al 2007, Yin et al 2010, Castillo et al 2009, Gu et al 2010, Nithiananthan et al 2009, Shamir et al 2009, Brock 2010, Gendrin et al 2011, Murphy et al 2011, Gu et al 2011, Pruem et al 2011, Liu et al 2012, check of the alignment of contours (Heath et al 2007, Wijesooriya et al 2008 or overlap of volumes (Klein et al 2009, Hardcastle et al 2012, evaluation of the overlay of corresponding structures in a color wash or a split screen technique (Kessler 2006, Thomas et al 2011 as well as the investigation of the performance on artificially created test cases, where the ground truth on the deformation is known (Wang et al 2005, Schnabel et al 2003, Kybic and Unser 2003, Kashani et al 2007, Li et al 2009, Long et al 2010, Zhong et al 2010, Kirby et al 2011, Chang et al 2010. Szegedi et al 2012 and Szegedi et al 2013 created a real tissue phantom to evaluate image registration in the presence of motion that involves four dimensional hysteresis.…”

A novel evaluation model of image registration for cone‐beam computed tomography guided lung cancer radiotherapy

“…CBCT-guided radiotherapy positioning is influenced by various factors, primarily observer-related differences in judgment, 2,3 the quality of images used for registration, [4][5][6] different registration algorithms, [7][8][9] and the choice of registration regions. [10][11][12] Some studies have emphasized that selecting different regions of interest (ROIs) for online image registration in planning images can lead to varying displacements.…”

“…Bissonnette et al 1 confirmed that the verification and adjustment of radiotherapy setup, guided by cone‐beam computed tomography (CBCT), can effectively enhance the geometric accuracy of lung cancer radiotherapy and minimize the external margin of the planning target volume (PTV). The accuracy of CBCT‐guided radiotherapy positioning is influenced by various factors, primarily observer‐related differences in judgment, 2 , 3 the quality of images used for registration, 4 , 5 , 6 different registration algorithms, 7 , 8 , 9 and the choice of registration regions. 10 , 11 , 12 Some studies have emphasized that selecting different regions of interest (ROIs) for online image registration in planning images can lead to varying displacements.…”

Use of a megavoltage electronic portal imaging device to identify prosthetic materials

“…Attempts have been made to overcome this problem using an extended CT number range [6], orthovoltage cone beam CT [7], and megavoltage cone beam CT [8,9]. However these techniques are either uncommon or experimental and are not available in many clinical departments [10].…”