Integrating clinical access limitations into iPDT treatment planning with PDT-SPACE

Abstract

PDT-SPACE is an open-source software tool that automates interstitial photodynamic therapy treatment planning by providing patient-specific placement of light sources to destroy a tumor while minimizing healthy tissue damage. This work extends PDT-SPACE in two ways. The first enhancement allows specification of clinical access constraints on light source insertion to avoid penetrating critical structures and to minimize surgical complexity. Constraining fiber access to a single burr hole of adequate size increases healthy tissue damage by 10%. The second enhancement generates an initial placement of light sources as a starting point for refinement, rather than requiring entry of a starting solution by the clinician. This feature improves productivity and also leads to solutions with 4.5% less healthy tissue damage. The two features are used in concert to perform simulations of various surgery options of virtual glioblastoma multiforme brain tumors.

Fig. 1.

Paraview visualizations of all tumor models viewed from the front of the brain. All meshes are 175 $mm$ in height.

Fig. 2.

Example of injection constraint on a brain tumor model with a parallel placement of 19 sources satisfying the constraint.

Fig. 3.

Example of a valid injection and an invalid injection.

Fig. 4.

Problem definitions for determining the maximum distance of displacement d.

Fig. 5.

Example visualization of initial source placement generation with an injection constraint on T1. a) Definition of $l_{ref}$ and $z_{ref}$ . b) A 2D coordinate system in a plane perpendicular to $z_{ref}$ . sources are placed at each grid cross point along the $z_{ref}$ direction.

Fig. 6.

Oriented Bounding Box of an ellipsoid.

Fig. 7.

Example visualization of initial source placement generation with no injection constraint on T1.

Fig. 8.

OBB and sample initial source placement result for T1. The longest and shortest dimensions of the OBB define the major and minor axes of the tumor, respectively.

Fig. 9.

Overall (a) and tumor model specific (b) correspondence graphs between initial source placement overdosage ( $m{m}^3$ ) (horizontal axis) and final placement overdosage ( $m{m}^3$ ) (vertical axis). Note this graph consists of the combination of all points obtained for T1 $\sim$ T9.

Fig. 10.

The T1 tumor (red volume), manual initial source placement (white lines inside the tumor) and injection constraint (white circle) used in testing.

Fig. 11.

Runtime ( $h$ ) and OAR overdosage ( $m{m}^3$ ) results with and without injection constraints for 9 tumor cases. Each result for a tumor is the average of 10 PDT-SPACE runs with different random seeds.

Fig. 12.

OAR overdosage ( $m{m}^3$ ), runtime ( $h$ ) and number of sources of manually generated initial source placement (MANUAL), automatically generated initial placement using the same number of sources and source lengths as manual initial placement (AUTO-D), and automatically generated initial placement with unconstrained number of sources (AUTO-U). a) shows results with no injection constraint and b) shows results with an injection constraint. Each result for a tumor is the average of 10 PDT-SPACE runs with different random seeds.

Fig. 13.

Relationship between the number of sources required (horizontal axis) vs. the total energy required (vertical axis). Vertical axes shown in log-scale. Showing unit-less relative total energy; actual total energy should be further scaled with practical PS and light diffuser assumptions.