Jason Olsthoorn

Inter-observer and intra-observer reliability for lung cancer target volume delineation in the 4D-CT era

BACKGROUND AND PURPOSE To investigate inter-observer and intra-observer target volume delineation (TVD) error in 4D-CT imaging of thoracic tumours. MATERIALS AND METHODS Primary and nodal gross tumour volumes (GTV) of 10 lung tumours on the 10 respiratory phases of a 4D-CT scan were contoured by six radiation oncologist observers. Inter-observer and intra-observer variability were assessed by the coefficient of variation (COV) and the volume overlap index (VOI). ANOVA was performed to assess differences in inter-observer and intra-observer variability based on patient case difficulty, respiratory phase, physician seniority, and physician observer. RESULTS VOI analysis determined that inter-observer was a more significant source of error than intra-observer variability. VOI improved with the use of 4D-CT as compared to conventional CT. ANOVA analysis for COVs found case difficulty (easy versus difficult) to be significant for inter-observer primary tumour and intra-observer nodal disease delineation. Physician seniority, respiratory phase, and individual physician were not found to be significant for TVD error. CONCLUSION Variability in TVD is a major source of error in 4D-CT treatment planning. Development of measures to reduce inter-observer and intra-observer TVD variability are necessary in order to deliver high quality radiotherapy.

An evaluation of an automated 4D-CT contour propagation tool to define an internal gross tumour volume for lung cancer radiotherapy

BACKGROUND AND PURPOSE To evaluate an automated 4D-CT contouring propagation tool by its impact on the inter- and intra-physician variability in lung tumour delineation. MATERIALS AND METHODS In a previous study, six radiation oncologists contoured the gross tumour volume (GTV) and nodes on 10 phases of the 4D-CT dataset of 10 lung cancer patients to examine the intra- and inter-physician variability. In this study, a model-based deformable image registration algorithm was used to propagate the GTV and nodes on each phase of the same 4D-CT datasets. A blind review of the contours was performed by each physician and edited. Inter- and intra-physician variability for both the manual and automated methods was assessed by calculating the centroid motion of the GTV using the Pearson correlation coefficient and the variability in the internal gross tumour volume (IGTV) overlap using the Dice similarity coefficient (DSC). RESULTS The time for manual delineation was (42.7±18.6)min versus (17.7±5.4)min when the propagation tool was used. A significant improvement in the mean Pearson correlation coefficient was also observed. There was a significant decrease in mean DSC in only 1 out of 10 primary IGTVs and 2 out of 10 nodal IGTVs. Intra-physician variability was not significantly impacted (DSC>0.742). CONCLUSIONS Automated 4D-CT propagation tools can significantly decrease the IGTV delineation time without significantly decreasing the inter- and intra-physician variability.

Strong mode-mode interactions in internal solitary-like waves

Classical linear theory presents vertically trapped internal waves of different modes as completely uncoupled. This description carries over to the simplest weakly nonlinear theory for internal solitary waves, the Korteweg-de Vries theory. The balance between weakly nonlinear and dispersive effects in this theory allows for soliton solutions, meaning that waves emerge from collisions without changing form. However, exact mode-1 internal solitary waves have been shown to exhibit departures from soliton behaviour during overtaking collisions. We present a numerical investigation of the strong modal coupling between mode-1 and mode-2 internal solitary-like waves during head-on and overtaking collisions. We begin by presenting a “clean” theoretical setup using an exact theory (the Dubreil Jacotin Long equation) for the mode-1 wave and weakly nonlinear theory for the mode-2 wave to initialize the numerical model. During the collision, the mode-2 wave is significantly deformed by the mode-1 wave-induced currents, and indeed, by the end of the collision, the mode-2 wave has lost coherence almost entirely. We discuss how the collisions change as the amplitude of the mode-1 wave decreases, as the mode-1 wave becomes broad crested, and when multiple pycnoclines preclude mode-2 wave breaking and the formation of quasi-trapped cores in the mode-2 waves. We demonstrate where viscous dissipation occurs during the collisions, finding it slightly enhanced in the near pycnocline region, but not to the point where it can explain the loss of coherence. Subsequently, we use linear theory to demonstrate that it is a combination of the pycnocline deformation and the shear across the pycnocline centre due to the mode-1 waves, which alters the structure of the mode-2 waves and leads to the loss of coherence. In fact, the shear is vital, and with only a deformed pycnocline, mode-2 wave structure is only slightly altered. We present the results of a direct numerical simulation on experimental scales in which both mode-1 and mode-2 waves are generated by stratified adjustment. This simulation confirms that the numerical results should be readily observable in the laboratory. We conclude by revisiting existing weakly nonlinear theory for collisions, finding a surprising twist on the well established notions of “weak” and “strong” collisions.

On the dynamics of vortex-wall interaction in low viscosity shear thinning fluids

We apply a pseudospectral method to numerically study the dynamics of vortices found within a low viscosity non-Newtonian fluid with a Carreau fluid rheology. The application of a Carreau fluid rheology avoids the commonly observed complications in power-law models at zero strain-rate. We find that fluids with a shear thinning rheology will preserve the small scale features of the flow. In particular, for vortex-solid wall interactions, shear thinning fluids can exhibit behavior associated with Newtonian fluids at a much higher Reynolds number. This can include secondary vorticity generation, and multiple vortex-bottom collisions each marked by periods of higher bottom shear rates. Using a variety of experimentally determined parameters from the literature, we argue that these results have direct application to many non-Newtonian fluids, including non-Newtonian fluid mud layers found on lake and ocean bottoms.