Hyperthermia treatment planning guided applicator selection for sub-superficial head and neck tumors heating

Figures & data

Figure 1. (a) Patient 3D model in SEMCAD X with highlighted CTV and GTV (homogeneous transparent patient model was used for better visualisation of CTV and GTV). Example of treatment planning setups for all studied applicators, (b) HYPERcollar, (c) Lucite cone applicator and (d) current sheet applicator.

Figure 2. CTV and GTV fully (a) within the frame of the HYPERcollar, (b) under the LCA aperture.

Figure 3. Piecewise linear temperature dependent perfusion scaling factors for muscle, fat and tumour. Please note that the perfusion at the specified temperature is obtained by multiplying the perfusion value at 37 °C with the temperature dependent scaling function.

Table 1. Summary of electromagnetic simulation settings in SEMCAD X.

Model	Global grid step (mm)	Refined grid step (mm)	Periods (−)	Size (Mcells)	Calculation time (min.)
LCA	5	1/0.5	35	15	10
CSA	5	1/0.05	25	65	30
HYPERcollar	1.25	–	15	75	8

Table 2. Dielectric properties at 434 MHz [20], thermal properties for 37 °C [19], * optimised values from ten patients treated with HYPERcollar applicator [21], the internal air and lung were modelled in temperature simulations using boundary condition.

Tissue	εr (–)	σ (S/m)	ρ (kg/m^3)	c (J/kg·K)	k (W/m·K)	ω (ml/min·kg)	Q (W/kg)
Air	1	0	1.2	–	–	–	–
Blood	–	–	1050	3617	–	–	–
Bone	13.1	0.09	1908	1312	0.32	10	0.15
Brainstem	55.1	1.05	1046	3630	0.51	559	11.4
Cartilage	45.1	0.6	1099	3568	0.49	35	0.54
Cerebellum	55.1	1.05	1045	3653	0.51	763	15.5
Cerebrum	56.8	0.75	1045	3696	0.55	763	15.5
Fat	11.6	0.08	911	2348	0.21/0.5*	32.7/255*	0.51
Lucite	2.6	0.003	1000	–	–	–	–
Lung	23.6	0.38	394	–	–	–	–
Muscle	56.7	0.8	1090	3421	0.49/0.4*	39.1/442.8*	0.96
Optical nerve	35	0.46	1075	3613	0.49	160	2.5
Sclera	57.4	1.01	1032	4200	0.58	380	5.9
Spinal cord	69	1.53	1005	4047	0.59	160	2.5
Tumour/GTV	59	0.89	1050	3950	0.51/1.5*	72.3/848*	0
Thyroid	61.3	0.89	1050	3609	0.52	5624	87
Vitreous humour	69	1.53	1005	4047	0.59	0	0
Water	78	0.04	1000	–	–	–	–

Following the Lang et al. [24] model, the rest value for tumour perfusion is 1.85 times higher than the rest value for muscle perfusion.

Figure 4. (a) CTV and GTV volume cumulative histograms for 24 CTVs and 17 GTVs, (b) mean CTV depth profiles including minimum and maximum depth interval.

Figure 5. Example of SAR and temperature cumulative histograms and depth profiles for Lucite cone applicator (LCA), current sheet applicator (CSA) and HYPERcollar applicator (HC). CTV (a) cumulative SAR histograms; (b) constant perfusion model (ω_const.), temperature histograms; (c) temperature dependent perfusion model (ω_Tdep.) temperature histograms; (d) mean SAR ± standard deviation (SD) range; (e) mean temperature ± SD range for (ω_const.) and (f) mean temperature ± SD range for (ω_Tdep.) for CTV depth (from 18 mm to 50 mm) and all three studied applicators.

Figure 6. TC25 comparison of CTV for: (a) the Lucite cone applicator (LCA) and (b) the current sheet applicator (CSA) vs. the HYPERcollar, comparisons for CTV within HYPERcollar frame are encircled with red colour. Note that for two LCA vs. HYPERcollar comparisons we received almost identical values resulting in partly overlapping of two points.

Figure 7. Example of normalised cubic-filtered SAR (%) axial slices in VEDO for (a) Lucite cone applicator, TC25 = 86%; (b) initial HYPERcollar optimisation, TC25 = 62% and (c) modified HYPERcollar optimisation which was done in order to improve TC25 = 78%, however the target-to-hotspot quotient (THQ) decreased from 1.73 for case (b) to 1.48 for case (c); for case (a) LCA THQ =1.12. Note that the CTV delineation for LCA and HYPERcollar differs slightly due to the different FDTD grid implementation in SEMCAD X.

Figure 8. (a) correlations between TC25_HYPERcollar–TC25_LCA and 22 CTV volumes, correlations between T50_HYPERcollar–T50_LCA using (b) constant perfusion model and c) temperature dependent perfusion model.

Figure 9. Dependency of TC25_LCA and 14 maximum CTV depths for CTVs fully under LCA apertures. Applying a threshold of TC25 = 75%, i.e. our inclusion criteria for deep H&N HT, the maximum CTV depth is 50 mm.

Gabriel S, Lau RW, Gabriel C. (1996). The dielectric properties of biological tissues: III. Parametric models for the dielectric spectrum of tissues. Phys Med Biol 41:2271–93.

 PubMed Web of Science ®Google Scholar

Hasgall PA, Neufeld E, Gosselin MC, et al. ITIS Database for thermal and electromagnetic parameters of biological tissues; 2013. Available from: https://www.itis.ethz.ch/virtual-population/tissue-properties/overview/.

 Google Scholar

Verhaart RF, Verduijn GM, Fortunati V, et al. (2015). Accurate 3D temperature dosimetry during hyperthermia therapy by combining invasive measurements and patient-specific simulations. Int J Hyperthermia 31:686–92.

 PubMed Web of Science ®Google Scholar

Lang J, Erdmann B, Seebass M. (1999). Impact of nonlinear heat transfer on temperature control in regional hyperthermia. IEEE Trans Biomed Eng 46:1129–38.

 PubMed Web of Science ®Google Scholar