Quality assurance guidelines for superficial hyperthermia clinical trials : II. Technical requirements for heating devices.

Quality assurance (QA) guidelines are essential to provide uniform execution of clinical trials with uniform quality hyperthermia treatments. This document outlines the requirements for appropriate QA of all current superficial heating equipment including electromagnetic (radiative and capacitive), ultrasound, and infrared heating techniques. Detailed instructions are provided how to characterize and document the performance of these hyperthermia applicators in order to apply reproducible hyperthermia treatments of uniform high quality. Earlier documents used specific absorption rate (SAR) to define and characterize applicator performance. In these QA guidelines, temperature rise is the leading parameter for characterization of applicator performance. The intention of this approach is that characterization can be achieved with affordable equipment and easy-to-implement procedures. These characteristics are essential to establish for each individual applicator the specific maximum size and depth of tumors that can be heated adequately. The guidelines in this document are supplemented with a second set of guidelines focusing on the clinical application. Both sets of guidelines were developed by the European Society for Hyperthermic Oncology (ESHO) Technical Committee with participation of senior Society of Thermal Medicine (STM) members and members of the Atzelsberg Circle.

Introduction

These quality assurance (QA) guidelines for the application of superficial hyperthermia clinical trials were developed at the request of the Atzelsberg Circle for Clinical Hyperthermia of the interdisciplinary working group of hyperthermia “Interdisziplinäre Arbeitsgruppe Hyperthermie” (IAH) [1] of the German Cancer Society (“Deutsche Krebsgesellschaft”) to the European Society for Hyperthermic Oncology (ESHO) [2]. ESHO delegated this task to the ESHO technical committee (TC), who formulated the guidelines with participation of experienced members of the Society for Thermal Medicine (STM) [3]. In addition, the manufacturers providing equipment for superficial hyperthermia were invited to provide their feedback on the QA guidelines during public sessions at the 2014 and 2015 annual meetings of ESHO or alternatively by personal communication.

The QA guidelines seek to establish a minimum level of treatment quality in hyperthermia treatments delivered in all multi-institutional studies initiated by the Atzelsberg Circle or under the auspices of the ESHO. The goal of this effort is to establish QA guidelines for the application of superficial hyperthermia, similar to the QA guidelines for administration of deep hyperthermia defined earlier [4, 5] and as a long awaited follow up to previous superficial hyperthermia QA guidelines provided by the Radiation Therapy Oncology Group and ESHO [6-8].

These QA guidelines for the application of superficial hyperthermia clinical trials consists of two parts:

Part I [9] provides detailed instructions on treatment documentation, defines a good hyperthermia treatment, and identifies the clinical conditions where a certain hyperthermia system can or cannot adequately heat the tumor volume.

Part II, i. e., this document, provides quality assurance requirements for heating equipment as well as detailed instructions how to characterize and document the performance of hyperthermia devices in order to apply reproducible, uniform, and high quality hyperthermia treatments. These characteristics help to establish the maximum size and depth of tumors that can be heated adequately.

Implementation of these QA guidelines should facilitate correct assessment of whether a tumor can or cannot be heated with the specific heating device(s) available in a hyperthermia center and, thus, enable a decision as to whether a patient is or is not eligible to participate in a clinical study.

Many different systems are used to apply superficial hyperthermia, either built commercially or in academic research laboratories. Each system has unique characteristics and advantages (as summarized in Appendix A), which may result in a large heterogeneity in the quality of applied hyperthermia treatments between various hyperthermia centers.

To assure proper performance of a superficial hyperthermia applicator, the spatial thermal pattern should be evaluated under well-controlled conditions prior to the first clinical use of the applicator and at regular intervals as part of a clinical QA program. Due to time constraints in the clinical workflow and the large diversity of superficial heating systems, which include electromagnetic (EM), radiofrequency (RF), ultrasound (US), and infrared (IR) technologies, these new QA guidelines require characterization of the heating performance of hyperthermia equipment with temperature (rise) simulations and measurements in homogeneous muscle tissue-equivalent phantoms. This is a more pragmatic approach than that used in previous guidelines/recommendations, which were based on power deposition patterns quantified in W/kg and described as Specific Absorption Rate (SAR) distributions. For homogenous nonperfused phantom models, the SAR and temperature rise in a defined period of time are directly related, since no convective heat losses are present [10].

Definitions and characteristic features of a superficial hyperthermia system

Applicators

Superficial applicators, used today, consist of

External EM antennas or waveguides,

External EM capacitive electrodes (“capacitive” RF systems),

External US transducers (US systems), and

External noncontacting IR heating systems.

In general, these devices deposit energy to heat a limited volume of tissue close to the heating device. A brief summary of the operating principles of various systems for heating superficial tissue is given in Part I of the QA guidelines [9], while more detailed equipment options are given in Appendix A. Further information can be found in reviews [11-14].

Superficial hyperthermia applicator terminology

Single applicator : A single radiating aperture connected to a power amplifier having independent control of output power from 0–100%. The applicator can be an EM radiator [16-24], US transducer [14], IR lamp [15], or the active electrode of a capacitive system [52, 53].

Multi-element applicator array: To produce effective heating of large area tumors, several single applicators can be combined into larger arrays with separate power control of each element to enable 2D power steering [25, 29, 34, 36]. Examples include patient-customized arrays of separately placed applicators such as 2–6 lucite cone applicator (LCA) [18] or two adjacent contact flexible microstrip applicator (CFMA) [19].

Multi-element array applicator: Alternatively a heat applicator can be constructed with a fixed spatial array of independently controlled heating elements, such as spiral microstrip [26-28], conformal microwave array (CMA) [30-33], Microtherm 1000 planar array [35] applicators, or a fixed array of ultrasound apertures, such as four and sixteen element devices [60, 61].

In order to generate sufficiently uniform and repeatable heating, special care must be devoted to interapplicator spacing and noncoherent phase excitation in order to minimize cross-coupling and phase interaction between adjacent applicators and to achieve a contiguous SAR/temperature distribution above the 50% iso-SAR or 50% isotemperature level across the tissue target between applicator elements.

Single applicators and multi-element applicators should fulfill the following criteria or be characterized as follows:

Every single radiating aperture should be capable of producing effective superficial HT under the aperture (Section Test conditions).

Every radiating element of an applicator array or array applicator must be powered noncoherently in order to avoid cross-coupling effects between the elements.

Multiple coherently radiating elements may be considered as a single independent radiating element if a common wave front is created parallel to the surface.

The therapeutic region of an array applicator or an applicator array is defined analogously to the definition of therapeutic region for a single applicator.

It is strongly recommended to include thermometry measurement points underneath each independently powered element of an array applicator or applicator array.

Technical considerations for the water bolus

The function of the temperature-controlled water-circulating bolus is to couple electromagnetic or ultrasound energy into the patient and to control the skin surface temperature. IR systems generally operate without a water bolus and can achieve some control of skin surface temperatures through forced convection of air.

General requirements for the water bolus

An optimal contact area of the water bolus with the skin of the target surface requires that the bolus is sufficiently large to smoothly follow the skin contour [37-40] and that the bolus extends beyond the radiating aperture. This provides better coupling of EM/US field to tissue without distorting the radiation pattern; at the same time it puts greater demands on reproducible alignment of the applicator relative to the tumor margin. An adequate bolus design is required since the dimensions and temperature of the water bolus significantly affect the applicator power deposition (SAR) pattern, thermal effective field size (TEFS), and thermal depth profiles. For EM radiative applicators, the bolus must be filled with deionized water, whereas for capacitive heating saline bolus is generally preferred. For US transducers, the water must be degassed as well. In all cases, the water must be circulated through the bolus with a circulation pump and the temperature controlled within the range of 151 to 45 °C. A bubble trap, impurity filter and a flow indicator is recommended to be included in the circulation system. The input and output flow connectors of the water bolus are located on opposite sides. Large applicator arrays may require dual-input–dual-output ports in order to maintain acceptable temperature homogeneity (<1 °C variation) across the bolus surface [39-41]. The water bolus can be filled with low density semirigid porous foam, plastic spacers, or thin rubber pins to avoid collapse of the bolus while pressed against the skin surface.

Specific features of the water bolus

Some manufacturers provide a water bolus that is an integral part of the applicator with a rigid plastic frame that mounts to the applicator aperture [19, 22, 27–29, 33, 35]. While this makes bolus position more reproducible relative to the applicator, the consequence of a fixed-frame set-up is often a convex bolus shape with reduced contact area between water bolus and skin that is smaller than the radiating aperture. This may cause a localized E‑field discontinuity and associated high tissue temperature rise near the sharp transition in EM coupling at the water bolus–air interface if that occurs under a high field region of the radiating aperture [41]. Ultrasound systems require deionized and degassed water, with the degree of degassing required inversely proportional to frequency; this is most critical for systems operating at 0.5–1.5 MHz where a dissolved oxygen content below 0.1 ppm is required to avoid lossy propagation and outgassing of air from the coupling bolus during ultrasound transmission. In EM capacitive heating systems, saline (0.1–1.5%) is generally used [43, 44] to improve impedance matching between electrodes and muscle tissue, as well as to spread RF currents over the contact area and thereby reduce skin burns at the edge of the electrode.

Guidelines for proper design and evaluation of the water bolus

Size: EM-radiative applicators: the water bolus should extend at least beyond the perimeter of the applicator aperture and ideally at least 2–5 cm outside the perimeter. Phantom experiments of the applicator with various bolus sizes are recommended.

EM-capacitive systems: the TEFS is highly dependent on the size of the contact area of the coupling bolus and tissue. Especially at the edge of the bolus, care must be taken to provide smooth transitions at the bolus–skin contact at the peripheral rim of the water bolus [42-44].

Thickness: For EM-radiative applicators the optimal water bolus thickness depends on type of applicator, its dimensions as well as target depth [41] and the thickness should not exceed critical values to avoid oscillation modes in the bolus layer [45, 46]. These modes occur for larger sized applicators at frequencies above 100 MHz [47] and also for suboptimal contact between bolus and applicator [20].

EM-capacitive systems utilize a frequency below 100 MHz. They require a minimum water bolus thickness depending on applicator size to minimize the effect of concentrated current at the edge of the metal electrodes. Their performance is rather insensitive to a variation in water bolus thickness provided the minimum thickness requirement is satisfied [43, 44].

For ultrasound applicators with properly degassed water bolus, thickness has negligible effect on energy propagation (losses) since the attenuation in water is very low, but may affect beam pattern shape and position of the focal zone depending on the device and configuration. Phantom experiments or direct SAR or intensity measurements must be designed specific for each applicator with various bolus thicknesses in the clinically relevant range.

Variations in bolus thickness (±50%) are inevitable in clinical conditions due to the irregular skin surface contour. In order to assure that users know how to respond effectively to patient complaints and to unexpected temperature heterogeneity, the impact of these variations on field homogeneity and effective field size should be assessed experimentally or by simulations.

Good contact between bolus and phantom: Wet gauze can be used to improve heat transfer at the bolus–skin interface for EM devices [41] but proper care must be taken to ensure that no air gaps occur in the interface [48]. Air bubbles and air gaps need to be removed with acoustic gel or water coupling for ultrasound devices.

The water temperature has a profound impact on the effective penetration depth: The use of cold water will increase the effective penetration depth of the maximum temperature and the therapeutic extent up to 1–2 cm [41, 49]. Experimental evaluation with phantom or pennes bioheat equation(PBHE)-based models to estimate the actual penetration depth is therefore recommended for specific energy modalities and bolus configurations. These models can also be applied for estimating the impact of water bolus dimensions and flow.

Characterization of heating properties

Temperature rise criteria for adequate heating

ESHO-TC defines a heating device to be adequate/appropriate if a temperature rise (TR) of at least 6 °C above the starting temperature can be achieved in 6 min at 1 cm depth in a muscle-tissue equivalent phantom. Devices with heating depths less than 1 cm that are dedicated to use on limited depth superficial disease must fulfill the same criterion at a depth of 0.5 cm. The required temperature increase of 1 °C/min equals a power deposition rate (SAR) of ±60 W/kg [50].

The thermal effective field size (TEFS) is defined as the area within the 50% of maximum TR contour (i. e., ∆T ≥ 3 °C) in the 1 cm deep plane under the aperture, see Fig. 1. Similarly, the thermal effective penetration depth (TEPD) is defined as the depth at which the maximum TR is 50% of the maximum TR (i. e., ∆T ≥ 3 °C in 6 min) at 1 cm depth as shown in Fig. 2. Note that the maximum TR at 1 cm depth does not necessarily occur in the central cross section plane through the applicator as illustrated in Fig. 3.

Fig. 1

Example of the thermal effective field size (TEFS). Calculated normalized temperature rise (TR) distribution at 1 cm depth in a two-layered phantom, with fat layer thickness of 10 mm overlying the muscle phantom. Heating time t = 6 min, P = 175 W. The black solid line indicates the applicator aperture, while the black dashed line represents the water bolus. The maximum TR in the 1 cm deep plane in muscle-tissue equivalent phantom is Tmax1cm = 7.6 °C. The TEFS isotherm then quantifies the area with TR ≥ 3.8 °C

Fig. 2

Example of the thermal effective penetration depth (TEPD). Simulated temperature increase (blue – left axis) and specific absorption rate (SAR; red – right axis) as a function of depth in the center of two-layered phantom, with fat layer (blue) thickness of 10 mm. Heating time t = 6 min with P = 175 W and bolus surface temperature identical to the initial phantom temperature. The maximum temperature rise (TR) in the 1 cm deep plane in muscle-tissue equivalent phantom is Tmax1cm = 7.6 °C. The TEPD is thus the depth where the TR is 3.8 °C, i. e., 39 mm from the tissue surface. The maximum SAR at 1 cm depth in the muscle is approximately 75%. The resulting effective penetration depth (EPD) derived according to the traditional definition [7] is indicated by the arrow. Observe that the effective heat penetration is essentially at nearly the same depth for both definitions

Fig. 3

Calculated normalized specific absorption rate (SAR) distribution of a 10 × 10 cm lucite cone applicator (LCA) at 1 cm depth in a homogeneous muscle tissue phantom. The color scale from blue to dark red represents a 10% SAR increase for every color transition. Adapted from [66]

For consistency, the measurements to determine TEFS must be made within 15 s after the end of heating to minimize thermal conduction smearing of the SAR pattern within the phantom. If a multi-element applicator array is used, elements must be arranged such that a continuous region of TR > 50% of the maximum TR exists. This assessment should be made at a depth of 1 cm. In order to determine TEPD, analogous measurements at other depths (>1 cm) are needed until TEPD is found.

Note that besides characterizing the temperature distribution, the efficiency of power transfer within the complete heating system (cables, applicator, and bolus) should be measured. This information can be used as a reference value to compare the clinical applied power levels between hyperthermia centers and assess whether realistic powers are applied to assure the required increase in tumor temperature. However, it is also important that the user is aware of how much power is lost in cables, applicator, and bolus. Efficiency measurements can be accomplished either with a calorimetric2 method or the total absorbed power may be calculated from rate of TR measurements.

Note that TR estimation alone is not sufficient for a complete characterization of the system. Numerical modeling and/or experimental assessment in terms of other parameters like SAR are recommended to achieve a more fundamental characterization of the system and to assist the user with determining the relevant parameters for the actual clinical setup. SAR can be determined with the TR measurement procedure described above if a short measurement time of 1 min or less is used to prevent blurring of the SAR pattern by heat conduction. This will require a sufficiently high power output to achieve an adequate temperature rise (e. g., >3 °C) in that short time interval.

The following SAR criteria will then apply for applicator characterization [7]:

In these guidelines, we recommend deriving both TEFS and TEFD from TR measurements. However, the values derived from SAR, as defined below, may also be useful, e. g., for comparisons with numerical calculations:

The effective field size (EFS) is defined by the area within the 50% of maximum SAR contour in the 1 cm deep plane under the aperture.

The effective penetration depth (EPD) is defined by the depth where the SAR falls to 50% of the maximum SAR at 1 cm depth. Note that the maximum SAR may not be in the plane through the main central axes of the applicator.

Generic phantoms

Two models are necessary to characterize heating of two typical disease conditions: (1) superficial chestwall disease that extends from the tissue surface to moderate depth (generally <15 mm), and (1) tumors at moderate depth underlying normal skin and fat.

For reliable phantom measurements the general rule is that the phantom size should extend beyond the applicator by at least 10 cm in all directions. The thickness of the phantom should be at least three times the expected penetration depth. Applying this rule leads to the following recommendations for phantom dimensions:

Superficial chestwall disease,

EM-radiative systems: The block phantom should consist of homogeneous muscle tissue equivalent material at least 10 cm thick and extending laterally 10 cm beyond the physical dimensions of the applicator or 5 cm beyond the bolus, whichever is larger.

EM capacitive systems: The phantom thickness should be at least 15 cm with lateral dimensions at least twice as large as the electrodes.

IR systems: The phantom size should extend beyond the expected TEFS by at least 5 cm and should have a thickness of 5 cm.

US systems: The phantom should consist of homogeneous muscle or “soft tissue” equivalent material extending laterally 5 cm beyond the physical dimensions of the applicator or the bolus, whichever is larger, and the thickness should be at least three times the expected penetration depth.

Tumors at moderate depth underlying normal skin and fat,

EM radiative systems: To model tumors underlying normal fat, a 1 cm thick layer of fat tissue-equivalent phantom should be placed on top of the muscle phantom which should then have a thickness of at least 9 cm.

EM capacitive systems: To model tumors underlying normal fat, a 1 cm thick layer of fat tissue-equivalent phantom should be placed on top of the muscle phantom which should then have a thickness of at least 14 cm.

IR systems: Similar to EM phantom with fat layer thickness of 0.5 and 1 cm.

US systems: Similar to EM phantom with muscle phantom thickness at least three times the expected penetration depth.

In all phantoms, the muscle phantom should be composed of at least two layers separated by a thin (≤0.1 mm) polyester, cheesecloth, or mylar film. The US phantoms should be coupled with liquid or gel, and separated by thin <50 um polyethylene or mylar film. The thickness of the upper layer should be 1 cm. For more extensive evaluation of distributions at different depths in the phantom, the block of muscle phantom may be separated into additional layers, with thicknesses of 0.5, 1, 1.5, 2.5 cm in case of EM devices operating at a frequency of 915 MHz or higher and 1, 1.5, 2.5, 4.5 cm for other frequencies. The fat phantom should consist of two layers with thickness of 0.5 cm.

If the power deposition pattern will be characterized as a function of depth with an IR camera, the muscle phantom should be split in two halves along a vertical cross-section along the center of the phantom. The applicator should be positioned with its central axis aligned with the vertical measurement plane of the phantom. The applicator should be rotated 90° between successive measurements to determine the pattern along both central (major and minor) axes of the applicator. A 1 cm thick fat phantom layer split down the middle should be added to the top of the muscle phantom to model tumor underlying normal fat.

For all temperature and power deposition measurements, a proper phantom must be used with the correct electrical, optical, or ultrasound properties and correct thermal properties. Tissue equivalence of the phantom material can be obtained by using freshly made materials following a validated recipe or demonstrating equivalent properties via measurements.

Phantom recipes for electromagnetic equivalent phantoms at commonly used frequencies are specified in Appendix B; a limited set of suggestions of EM–phantom mixtures for less frequently used frequencies, ultrasound phantoms, and infrared phantoms are also provided.

Test conditions

The thermal distribution of the applicator should be characterized under standardized operating conditions and at all frequencies in clinical use. The two generic reference phantom set-ups described in Section Generic phantoms should be used for this characterization.

All measurements should be carried out at room temperature, i. e., the phantom model should be at equilibrium with room temperature at the start of each experiment. A bolus of appropriate dimension should be placed on top of the phantom with liquid circulating at the same room temperature as the rest of phantom. The distribution of temperature rise must be measured in three orthogonal planes crossing the center of applicator as shown in Fig. 4.

Fig. 4

Thermal effective field size (TEFS) profiles should be measured in three orthogonal planes crossing the center of the applicator

The horizontal plane at 1 cm depth must always be measured by an IR camera using a horizontally split phantom with a removable top layer of 1 cm which is removed for the IR temperature measurement.

The vertical planes (xz, yz) can be measured with one of three alternative methods as shown in Fig. 5:

Fig. 5

Illustration of three alternative options to obtain thermal profile with depth: a thermal camera view of vertical plane containing peak temperature rise (TR); b reconstruction of vertical distribution from thermal camera views of multiple horizontal planes; and c multiple measurements along a single axis depth probe

with IR camera using a vertically split phantom or

with IR camera in multiple horizontal planes and post processing of data from multiple planes,

with temperature probes in catheters running through the phantom at the defined depths. Although this method can be used, the ESHO-TC prefers the use of methods (a) or (b) which produce much higher resolution characterization of the distribution.

Alternative method (a): Use of a vertically split phantom is highly recommended for best resolution of TR distribution.

Alternative method (b): To reconstruct the vertical TR distribution from multiple horizontal plane data, the measurement planes must include depths of 0.5 (for ≥915 MHz only), 1, 1.5, 2.5, 4.5 cm and additional depth planes as needed to measure SAR < 10% of max SAR under the applicator.

Alternative method (c): Multiple stationary temperature probes must be used to obtain a minimum resolution of 1 cm. The absolute TR at each depth measured immediately after switching on EM, US, or IR power must be used to plot the temperature depth profile as a means to determine TEPD.

With any of the methods, if the vertical plane measurements show a deviation of the temperature maximum from the central plane, an additional measurement must be performed through the vertical plane along the cross-section of the applicator that includes the true maximum TR. In addition, at least one temperature sensor must be placed at 1 cm depth in the center of the central plane of the applicator as a reference point regardless of the test set-up applied. The absolute TR measured at this location is indicative of the efficiency of the applicator to deposit a high SAR at this position and combined with applicator input power allows simple quantitative comparison between institutes of the temperature increase obtained in the fixed 6 min heating interval. The SAR is related to the temperature increase in six minutes of heating according the simple formula: SAR [W/kg] = 67 × dT/dt [°C/min] [50].

Required specifications of the IR thermal imaging

As discussed in method (b), IR thermal imaging is required for characterization of the temperature distribution in the horizontal plane at 1 cm depth in phantom. The resolution of readings and relative accuracy (NETD) of the thermographic camera should be 0.1 °C and 0.05 °C or better, respectively. The spatial resolution should be 1–2 mm for phantoms less than 30 cm2, and 2–3 mm for larger phantoms. The camera should be placed at a proper height to cover the entire phantom surface and test images should be recorded with rulers in the field of view to ensure correct rendition of dimensions.

Disclaimer

This publication is based on literature and other sources of information judged to be reliable by the authors representing the ESHO-TC. However, the authors, ESHO-TC, and editors disclaim any warranty or liability based on or relating to the contents of this publication. The authors and ESHO-TC do not endorse any products, manufacturers, or suppliers. Nothing in this publication should be interpreted as implying such endorsement. Several companies were invited to provide feedback on the document but have not participated actively either at ESHO-TC meetings or in the writing of this document. Some companies were invited to provide feedback on the document but have not participated actively either at ESHO-TC meetings or in the writing of this document. The authors and ESHO-TC alone are responsible for the content and writing of this paper.

Table A.1

General performance of heating systems

		Thermal effective penetration depth	Dealing with problems at the interfacesa	Spatial power control		Thermal effective field size	Ease of use and positioning	Comfort to the patient	Temperature control	Commercial
				Dimensions	Resolution
EM waveguide	Single applicator	+	++	1D	–	- to +	0	0	+	yes
	Array	+	++	2D	++	0 to ++	-	0	++	no
EM microstrip	Single applicator	0	++	1D	–	0	+	+	+	yes
	Array	+	++	2D	++	++	+ to ++	+	++	yes
EM capacitive		+	–	1D	–	+	++	+	–	yes
US	Single applicator	++	0	1D	–	-	0	+	+	no
	Array	++	0	3D	++	-	+	+	++	no
IR		–	++	2D	+	++	++	++	+	yes

The performance is indicated by following grades: ++ very good, + good, o fair, - poor, – very poor

aDealing with the problems at the interfaces (fat layer under the skin and/or muscle/bone interface)

Table B.1

Literature values of relevant dielectric [71, 74], acoustic [72, 74], and thermal properties [74] of muscle and fat

		US	RF/microwave				Physical
		2–8 MHz	13.56 MHz	27 MHz	433 MHz	915 MHz
Muscle	εr [−]	–	138	95.9	56.87	55.00	–
	σ [S/m]	–	0.628	0.654	0.81	0.95	–
	α/f [dB/cm/MHz]	1.1	–	–	–	–	–
	c [m/s]	1588.4	–	–	–	–	–
	ρ[kg/m3]	–	–	–	–	–	1090
	cp [J/kg/K]	–	–	–	–	–	3421
	k [W/m/K]	–	–	–	–	–	0.49
Fat	εr [−]	–	25.4	18	11.6	11.3	–
	σ [S/m]	–	0.055	0.06	0.08	0.11	–
	α/f [dB/cm/MHz]	0.6	–	–	–	–	–
	c [m/s]	1440.2	–	–	–	–	–
	ρ[kg/m3]	–	–	–	–	–	911
	cp [J/kg/K]	–	–	–	–	–	2348
	k [W/m/K]	–	–	–	–	–	0.21

Where ε r [−] is relative dielectric constant, σ [S/m] is electric conductivity, α/f [dB/cm/MHz] is attenuation per frequency, c [m/s] speed of sound, ρ [kg/m3] is material density, c p is specific heat capacity [J/kg/K], k is thermal conductivity [W/m/K]

Table B.2

Muscle phantom recipes. The quantities are given for preparation of 1 l of phantom

	13 & 27 MHz	433 MHz	915 MHz	433 MHz	915 MHz
Reference	[79]	Adjusted recipe of [78]		Adjusted recipe of [77]
TX 150b/TX 151 (Superstuff)	–	19.3 g	18.3 g	–
Agar powder	40 g	20.9 g	20.9 g	16 g	16 g
Formaldehyde 8%	–	6.7 g	6.7 g	–	–
NaCl (salt)	2.4 g	3.8 g	4.8 g	10 g	7 g
Deionized Water	≈956 g	701.6 g	701.6 g	570 g	570 g
Sugar	–	–		400 g	447 g
Polyethylene powder	–	50.1 g	57.6 g	–
Permittivity	79.0	60.7	52.8	58.7	52.4
Conductivity S/m	0.44–0.6a	0.85	1.07	0.67	0.98

aThe lower value is recently measured, whereas the higher value is taken from the literature [79]. The difference can be explained by the absence of highly toxic preservative, NaN3. Hence, it is recommended to use 3.3 g NaCl per liter phantom to achieve an estimated value of 0.6 S/m

bTX 150 & TX 151 (Oil Center Research, Lafayette, LA, USA); * In the original work [70], highly toxic NaN3 was used as preservative. Other, less toxic preservatives were given with the mass equivalent quantities of NaCl, to compensate the conductivity decrease. Here, the recipe is adopted for use with formaldehyde 8%. The phantom can be however prepared even without this preservative. In such a case, an equivalent amount of water should be used.

Table B.3

Fat tissue equivalent phantom recipes

	13 & 27 MHz	433 MHz	915 MHz		433 MHz	915 MHz
Reference	[81]	[22, 80]			[82]
Laminac 4110	79%	69.30 g		85.2%	–
Aluminum powder	20.72%	29.46 g		14.5%	–
Shawinigan/Acetylene black	0.28%	0.30 g		0.24%	–
XC-72 Carbon powder	–	0.94 g		–	–
MEK peroxide/P 102 resin	7cc/2 kg	I part per 200 parts resin		0.375%
Flour	–	–			500
Oil	–	–			225
Saline (0.9% NaCl per liter)	–	–			25
Dielectric constant	19	5.5	5.61		7.1	9.1
Conductivity/S/m	0.028	0.04	0.0665		0.11	0.16

Table C.1

Routine quality assurance procedures

Interval	Daily		Quarterly3 months		6–12 months
Topic	Visual check	Acceptance criterion	Check function	Acceptance criterion	Check function	Acceptance criterion
Temperature probes	Broken probes	Accuracy ±0.2 °C	Full range calibration	Accuracy ±0.2 °C over 37–45 °C	–	–
Water bolus	Leakage, intact structure	Good circulation	–	–	–	–
Water system	Tubing and connections	Good circulation	Water exchange	Correct electrical conductivity	–	–
Power system	Damage	Power output as expected	–	–	Full calibration, power output	Power reading forward and reflected ±5%
Applicators	Damage	SAR pattern as expected	–	–	Heating pattern	TEFS, TEFD, EFS, efficiency
Control system	Interaction of commands	Proper functioning control algorithm	Back-up system	Proper function	Update software	Verify system performance
Safety	Cable connections	Correct undamaged connections	–	–	Stray radiation	Values within EU/FCC regulations