Systematic review of pre-clinical and clinical devices for magnetic resonance-guided radiofrequency hyperthermia

Figures & data

Figure 1. Selection process for inclusion of papers with criteria reported in the text.

Table 1. Table summarizes original studies on MR-guided RF HT devices.

Device	MRI insert configuration (HT-only/ dual-function)	Intended application (preclinical/clinical) and body site	MRI system Information	Validation status Not validated versus validated (preclinical/clinical)	Heating validation metric (Qualitative/quantitative)	MR compatibility validation metric (qualitative/quantitative	MRT sequence and parameters used for quantitative validation	MRT accuracy validation metric (Qualitative/quantitative)
BSD2000-3D-MRI: 12 dipole pairs (100MHZ) [40,42–47]	HT-only	Clinical – lower extremities and the pelvis	1.5T Body Coil RX/TX chain Simultaneous heating and imaging	Preclinical: cylindrical and elliptical phantoms	Quant.: Mean_Tincrease,max = 5.9 ± 0.4 °C, Mean_steering error = 0.4 ± 0.2 cm [44]	Not reported	SPGR sequence with TR: 600 ms TE: 20 ms FA:50 Reso: not reported [42] SPGR sequence with TR: 600 ms TE: 5, 20 ms FA:50 3.9 × 3.9 × 10 mm^3 [40] SPGR sequence with TR: 750 ms TE: 4, 20 ms FA:50 Reso: 3.9 × 3.9 × 10 mm^3 [43]	Quant.: ΔTave = ±0.4 °C [42] Tresolution = ±0.5–1 °C [40] ΔSARave <10% [43]
				Clinical: 24 patients:		Qual.: MRgHT is feasible. [45,46]	SPGR sequence with TR: 600 ms TE: 20 ms FA:50 Reso: 3.9 × 3.9 × 10 mm^3 [45]	Quant.: p = 0.04: MRTave,tumor versus clinical response [45]
				15× rectum/bladder [46]	Qual.: Tave,MR of the tumor is a useful variable to evaluate the quality and effectivity of treatments [46]
				9× high grade sarcoma – lower extremities/pelvis [45]	Not reported			R^2 = 0.96: MRT versus conventional invasive thermometry^a [45]
MAPA: 4 flared pairs (140MHz) [48–50]	HT-only	Clinical – extremities	1.5T Head Coil RX – Body Coil RX/TX chain Simultaneous heating and imaging	Preclinical: cylindrical phantom [48]	Not reported	Qual: No artifacts were observed	GRE sequence with TR: 34 ms TE: 20 ms FA:50 Reso: 0.9 × 0.9 × 10 mm^3 [49,50]	Quant: ΔTave < 0.2 °C, R^2 = 0.998 [49]
				Clinical: 14 patients: (leg sarcomas)	Not reported			Quant:
				4 patients [49]		Not reported		R^2 = 0.84
				10 patients [50]		Quant: SNR tumor = 115, SNR muscle = 90		ΔTave, tumor < 0.62 °C
MRlabCollar (434 MHz) [52,53,55]	HT-only	Preclinical – H&N	1.5T Body Coil RX/TX chain Sequential heating and imaging	Preclinical: cylindrical phantom	Quant.: Tincrease,max = 4.5 °C [53] R^2 = 0.76: Thermal distribution size and shape in MRT versus simulations [52] Qual: capability of the focused heating was confirmed [53]	Qual: no severe distortions in ROI in the MRT image [53]	SPGR sequence with TR: 34 ms TE: 14.9, 17.3, 19.7 ms FA:29 Reso: 1.6 × 1.6 × 10 mm^3 [55]	Quant: ΔTave<0.14 °C, ΔTmax = 0.22 °C [55]
MRCollar (434 MHz) [56]	HT-only	Clinical – H&N	1.5T Not reported	Preclinical: reduced scale setup (continuation of MRlabcollar)	Not reported	Qual: No large artifacts for the modules containing the antennas	Not reported	Not reported
RF capacitive (26 MHz) [57,58]	HT-only	Clinical – brain (nanoparticle)	3T Body Coil RX/TX chain Sequential heating and imaging	Preclinical: cylindrical phantom	Not reported	Not reported	GRE sequence with TR: 110 ms TE: 10 ms FA:60 Reso: 1.6 × 1.6 slice thickness not reported mm^3 [57,58]	Quant: ΔTave< 0.5 °C [57,58]
APA (915 MHz) [73]	HT-only	Clinical – deep seated tumor	4.7T Body Coil RX/TX chain Simultaneous heating and imaging	Preclinical: cylindrical phantom	Qual: ability to control of the position of the focal point [73]	Not reported	Not reported	Not reported
CSA (100 MHz) [60]	HT-only	Clinical – deep-seated pelvic and abdominal	3T Body Coil RX/TX chain Sequential heating and imaging	Preclinical: cuboid phantom	Not reported	Not reported	GRASS sequence with TR: 160 ms TE: 20 ms FA:40 Reso: 2 × 2×10 mm^3 [60]	Quant: ΔTave = ±1 °C, ΔTSD= 0.4 °C [60]
CWA (100 MHz) [61,62]	HT-only	Clinical – pelvis and abdomen	1.5T Body Coil RX/TX chain Simultaneous heating and imaging	Preclinical: cylindrical phantom	Qual.: good ability of field pattern steering [61,62]	Qual.: MR-compatibility has been verified in preliminary tests [61]	SPGR sequence with TR: 600 ms TE: 20 ms FA:50 3.9 × 3.9 × 10 mm^3 [62]	Quant.: ΔTave =0.45 °C [62]
Intravascular MRI guidewire (180 MHz) [67]	HT-only	Clinical – vessel wall	1.5T Body Coil TX chain/ Intravascular MRI guidewire RX Simultaneous heating and imaging	Preclinical: cylindrical phantom	Not reported	Qual.: The SNR profile was comparable to the pattern of simulated RF power distribution	SPGR sequence with TR: 68 ms TE: 15 ms FA:30 0.9 × 0.9 × 8 mm^3	Qual.: The simulated RF power distribution was comparable to that of MRT
				Aorta of rabbits	Quant.: Tincrease, max ≈7 °C			Quant.: ΔTSD= 1.2 °C
Slot & Dipole-type (2.45 GHz) [68]	HT-only	Preclinical – animal thigh muscle	7T Body Coil RX/TX chain Sequential heating and imaging	Preclinical: cuboid phantom	Not reported	Not reported	SPGR sequence with TR: 20 ms TE: 8 ms FA:20 1.2 × 1.2 × 3 mm^3	Quant.: ΔT2SD = +0.318/−0.339 °C
				10 male rabbits: thigh muscle	Not reported			ΔT2SD = +0.693/−0.661 °C
				In 6 rabbits: VX2 tumor cells were inserted	Quant.: Tincrease_ave_tumor = 42.60 ± 0.14 °C Tincrease_ave_skin = 43.27 ± 0.45 °C		SPGR sequence with TR: 20 ms TE: 8 ms FA:20 0.8 × 0.8 × 3 mm^3	ΔT2SD = +0.564/−0.526 °C
Coaxial-type (2.45 GHz) [69]	HT-only	Preclinical – local tumors in small animals	7T Body Coil RX/TX chain Simultaneous heating and imaging	Preclinical: cylindrical phantom	Quant.: Tincrease ≥ 3 °C	Not reported	SPGR sequence with TR: 40 ms TE: 4 ms FA:25 2.3 × 2.3 × 1 mm^3	Quant. : ΔTmax≤ ±0.6 °C
				2 experimental mice with implanted tumors	Not reported		SPGR sequence with TR: 40 ms TE: 4 ms FA:25 2.3 × 2.3 × 1 mm^3	Qual.: Feasibility of in vivo MRgHT
Hexagonal flared-type (130 MHz) [63–66]	HT-only	Preclinical – whole body	1.5T Body Coil RX/TX chain Simultaneous heating and imaging	Preclinical: cylindrical phantom	Not reported	Qual.: Simultaneous heating and imaging without deterioration of SNR [63]	GRE sequence with TR: 34 ms TE: 20 ms FA:25 0.6 × 0.6 × 4 mm^3	Quant.: ΔTave =0.92 °C, TRMS=1.23 °C [63] Tresolution =0.22 °C [64]
				2 canines: brain and muscle		Not reported		Tresolution =0.6 °C [64]
Yagi-Uda-based (2.45 GHz) [70,72]	HT-only	Preclinical – superficial tumors in small animals (murine)	7T Body Coil RX/TX chain Simultaneous heating and imaging	Preclinical: cylindrical phantom [70]	Qual.: ability to create a small heating focus (<1cm^3). Satisfactory agreement between simulated versus MRT distribution	Qual.: The applicator is MR-compatible. No disturbances are visible in the area of interest	Not reported	Not reported
				4 mice: tumor on leg [73]	Quant.: Tincrease =7 °C	Qual.: No artifact in MR images
Electronic switching: Coil/C-shaped dipole (Larmor freq.) [75]	Dual-function	Preclinical – not mentioned	3T Body Coil RX/TX chain Sequential heating and imaging	Preclinical: cylindrical phantom	Qual.: Ability for steerable heating. Repeatable heating patterns in subsequent experiments.	Not reported	SPGR sequence with TR: 100 ms TE: 12 ms FA:35 Reso: 3.75 × 3.75 ×  10 mm^3	Quant.: Tincrease,max ≈4.3 °C in 45 min
Thermal MR: Bow-tie dipoles (Larmor freq.) [76,77]	Dual-function	Preclinical – not mentioned	7T Body Coil RX/TX chain Sequential heating and imaging	Preclinical: cylindrical phantom	Qual.: ability to generate and steer deep-seated temperature focus [76] Quant.: Tincrease ∼11 °C: center of a head sized phantom [77]	Not reported	Not reported	Not reported
Statistics of preclinical devices	5× HT only 2× Dual-function	7× Preclinical devices		12× Preclinical validations	5× Quant 2× only Qual 5× Not reported	0× Quant 4× only Qual 8× Not reported	5× SPGR 2× GRE	8× Quant 1× only Qual 3× Not reported
Statistics of clinical devices	8× HT only	8× Clinical devices		9× Preclinical validations	2× Quant 2× only Qual 5× Not reported	0× Quant 5× only Qual 4× Not reported	6× SPGR 2× GRE 1× GRASS	6× Quant 1× only Qual 2× Not reported
				4× Clinical validations	0× Quant 1× only Qual 3× Not reported	1× Quant 2× only Qual 1× Not reported	2× SPGR 2× GRE	4× Quant 0× only Qual 0× Not reported

Note that in view of clarity, for each parameter, only the best achieved data is summarized in this table. Unless stated, all reported accuracies were measured versus fiber-optic temperature probe. R²: correlation coefficient, ΔT_ave: average temperature error, ΔT_SD: standard deviation of temperature error, ΔT_max: maximum temperature error, T_resolution: thermal resolution, SAR: specific absorption rate, SNR: signal to noise ratio. The statistics at the end of the table are highlighted with gray to distinguish that area from the devices.

^aThe accuracy of the MRT was measured against Bowman probes (high-impedance thermistors).

Figure 2. The number of publications over years after applying the search string and exclusion criterion 1 provided in the text (Material and methods). An increased number of publications on MR-guided RF HT (red bars) can be seen after 2001 (development of the BSD2000-3D-MRI) and after 2014 (development of the MRlabcollar and Thermal MRI).

Table 2. Summary of proposed validation stages and metrics for standardization of the validation of MRgHT devices.

Topic	Sub-topic	Quantification metric
Heating ability	Focusing	Option 1: Ttarget, increase Option 2: SARmax for 1 W input power Option 3: 50% iso-SAR contour
	Focus steering	Option 1: relocation of maximum heating point (mm) Option 2: relocation of center of the 50% iso-SAR contour (mm)
Compatibility	B0	No standardized metrics were found in literature. Proposal: average and maximum ΔB0 changes with/without the RF HT device present
	B1+	No standardized metrics were found in literature. Proposal: average and maximum ΔB1+ changes with/without the RF HT device present
	Decoupling	No standardized metrics were found in literature. Proposal: Option 1: Coupling between the systems (S-matrix). Option 2: Image quality degradation metrics during contemporaneous operation of MRI and RF HT
	Image noise	Image signal to noise ratio (SNR)
	Geometric distortion	DICE score, normalized cross-correlation of landmarks in calibration phantoms
MRT accuracy	ΔT Accuracy	Root mean squared error (RMSE)
	ΔT bias and precision	Mean error (ME) and Standard error (SE)
	Correlation	Coefficient of determination (R^2) and probability (p-value)

Gellermann J, Wlodarczyk W, Feussner A, et al. Methods and potentials of magnetic resonance imaging for monitoring radiofrequency hyperthermia in a hybrid system. Int J Hyperthermia. 2005;21:497–513.

 PubMed Web of Science ®Google Scholar

Mulder HT, Curto S, Paulides MM, et al. Systematic quality assurance of the BSD2000-3D MR-compatible hyperthermia applicator performance using MR temperature imaging. Int J Hyperthermia. 2018;35:305–313.

 PubMed Web of Science ®Google Scholar

Gellermann J, Wlodarczyk W, Ganter H, et al. A practical approach to thermography in a hyperthermia/magnetic resonance hybrid system: validation in a heterogeneous phantom. Int J Radiat Oncol Biol Phys. 2005;61:267–277.

 PubMed Web of Science ®Google Scholar

Gellermann J, Weihrauch M, Cho CH, et al. Comparison of MR‐thermography and planning calculations in phantoms. Med Phys. 2006;33:3912–3920.

 PubMed Web of Science ®Google Scholar

Gellermann J, Hildebrandt B, Issels R, et al. Noninvasive magnetic resonance thermography of soft tissue sarcomas during regional hyperthermia: correlation with response and direct thermometry. Cancer. 2006;107:1373–1382.

 PubMed Web of Science ®Google Scholar

Gellermann J, Wlodarczyk W, Hildebrandt B, et al. Noninvasive magnetic resonance thermography of recurrent rectal carcinoma in a 1.5 tesla hybrid system. Cancer Res. 2005;65:5872–5880.

 PubMed Web of Science ®Google Scholar

Delannoy J, LeBihan D, Hoult DI, et al. Hyperthermia system combined with a magnetic resonance imaging unit. Med Phys. 1990;17:855–860.

 PubMed Web of Science ®Google Scholar

Delannoy J, Chen C-N, Turner R, et al. Noninvasive temperature imaging using diffusion MRI. Magn Reson Med. 1991;19:333–339.

 PubMed Web of Science ®Google Scholar

Carter DL, MacFall JR, Clegg ST, et al. Magnetic resonance thermometry during hyperthermia for human high-grade sarcoma. Int J Radiat Oncol Biol Phys. 1998;40:815–822.

 PubMed Web of Science ®Google Scholar

Numan WCM, Hofstetter LW, Kotek G, et al. Exploration of MR-guided head and neck hyperthermia by phantom testing of a modified prototype applicator for use with proton resonance frequency shift thermometry. Int J Hyperthermia. 2014;30:184–191.

 PubMed Web of Science ®Google Scholar

Paulides MM, Bakker JF, Hofstetter LW, et al. Laboratory prototype for experimental validation of MR-guided radiofrequency head and neck hyperthermia. Phys Med Biol. 2014;59:2139–2154.

 PubMed Web of Science ®Google Scholar

Tarasek MR, Pellicer R, Hofstetter LW, et al. Validation of MR thermometry: method for temperature probe sensor registration accuracy in head and neck phantoms. Int J Hyperthermia. 2014;30:142–149.

 PubMed Web of Science ®Google Scholar

Paulides MM, Drizdal T, Van Rhoon GC, et al. Novel applicator design for MR guided RF hyperthermia in head and neck cancers: heating performance and RF coupling. Paris (France): International Society for Magnetic Resonance in Medicine (ISMRM); 2018.

 Google Scholar

Kim KS, Lee SY. Nanoparticle-mediated radiofrequency capacitive hyperthermia: a phantom study with magnetic resonance thermometry. Int J Hyperthermia. 2015;31:831–839.

 PubMed Web of Science ®Google Scholar

Kim KS, Hernandez D, Lee SY. Time-multiplexed two-channel capacitive radiofrequency hyperthermia with nanoparticle mediation. Biomed Eng Online. 2015;14:95.

 Web of Science ®Google Scholar

Kowalski ME, Behnia B, Webb AG, et al. Optimization of electromagnetic phased-arrays for hyperthermia via magnetic resonance temperature estimation. IEEE Trans Biomed Eng. 2002;49:1229–1241.

 PubMed Web of Science ®Google Scholar

Hoffmann W, Rhein KH, Wojcik F, et al. Performance and use of current sheet antennae for RF-hyperthermia of a phantom monitored by 3 tesla MR-thermography. Int J Hyperthermia. 2002;18:454–471.

 PubMed Web of Science ®Google Scholar

Nadobny J, Wlodarczyk W, Westhoff L, et al. Development and evaluation of a three-dimensional hyperthermia applicator with water-coated antennas (WACOA). Med Phys. 2003;30:2052–2064.

 PubMed Web of Science ®Google Scholar

Nadobny J, Wlodarczyk W, Westhoff L, et al. A clinical water-coated antenna applicator for MR-controlled deep-body hyperthermia: a comparison of calculated and measured 3-D temperature data sets. Ieee Trans Biomed Eng. 2005;52:505–519.

 PubMed Web of Science ®Google Scholar

Qiu B, El-Sharkawy A-M, Paliwal V, et al. Simultaneous radiofrequency (RF) heating and magnetic resonance (MR) thermal mapping using an intravascular MR imaging/RF heating system. Magn Reson Med. 2005;54:226–230.

 PubMed Web of Science ®Google Scholar

Demura K, Morikawa S, Murakami K, et al. An easy-to-use microwave hyperthermia system combined with spatially resolved MR temperature maps: phantom and animal studies. J Surg Res. 2006;135:179–186.

 PubMed Web of Science ®Google Scholar

Curto S, Faridi P, Shrestha TB, et al. An integrated platform for small-animal hyperthermia investigations under ultra-high-field MRI guidance. Int J Hyperthermia. 2018;34:341–351.

 PubMed Web of Science ®Google Scholar

Samulski TV, MacFall J, Zhang Y, et al. Non-invasive thermometry using magnetic resonance diffusion imaging: potential for application in hyperthermic oncology. Int J Hyperthermia. 1992;8:819–829.

 PubMed Web of Science ®Google Scholar

MacFall JR, Prescott DM, Charles HC, et al. 1H MRI phase thermometry in vivo in canine brain, muscle, and tumor tissue. Med Phys. 1996;23:1775–1782.

 PubMed Web of Science ®Google Scholar

Raaijmakers EAL, Mestrom RMC, Sumser K, et al. An MR-compatible antenna and application in a murine superficial hyperthermia applicator. Int J Hyperthermia. 2018;34:697–703.

 PubMed Web of Science ®Google Scholar

Sumser K, Geerman A, Haeck J, et al. Development and in vivo validation of an MRCompatible temperature controllable superficial hyperthermia applicator for small animal studies. Proceedings of the 1st World Conference on Biomedical Applications of Electromagnetic Fields (EMF-Med); 2018 Sep 10–13; Split, Croatia. IEEE; 2018.

 Google Scholar

Yeo D YX, Wu J, Hofstetter LW, et al. Investigation of a dual-function applicator for RF hyperthermia and MRI. Aarhus (Denmark): European Society for Hyperthermia Oncology (ESHO); 2011.

 Google Scholar

Winter L, Özerdem C, Hoffmann W, et al. Design and evaluation of a hybrid radiofrequency applicator for magnetic resonance imaging and RF induced hyperthermia: electromagnetic field simulations up to 14.0 Tesla and Proof-of-Concept at 7.0 Tesla. PLoS One. 2013;8:e61661.

 PubMed Web of Science ®Google Scholar

Winter L, Oezerdem C, Hoffmann W, et al. Thermal magnetic resonance: physics considerations and electromagnetic field simulations up to 23.5 Tesla (1GHz). Radiat Oncol. 2015;10:201.

 Web of Science ®Google Scholar