A portable induction heating system for implanted prosthesis disinfection

Abstract

Prosthetic implants, such as knee or hip prostheses, have significantly improved the wellbeing of citizens with articulation problems. However, there are still significant challenges in these procedures, being recurring infections one of the most challenging. In this context, hyperthermia has been studied as an effective alternative to antibiotics for biofilm eradication. Among the different heating alternatives, induction heating arises as the idoneous for this application due to its contactless nature and the metallic alloys used in prosthesis. This paper details the design of a portable induction heating system for implanted prosthesis disinfection.

Keywords

Induction heating power electronics biomedical applications prosthesis disinfection

1. Introduction

Prosthesis surfaces are prone to the development of resistant microbial biofilms [1]. These can create resistant and recurrent infections that, in many cases, requires removing the original prosthesis. Currently, antibiotic delivered together with surgical cleaning are the main and only treatments available, with limited effectivity due to the highly resistant characteristics of biofilm. In this context, hyperthermia has been studied as an effective alternative to antibiotics for biofilm eradication. Among the different heating alternatives, induction heating arises as the idoneous for this application due to its contactless nature and the metallic alloys used in prosthesis. This enables direct heating of the surface to treat with minimum impact to surrounding tissues and avoiding risks associated to contact. For this reason, induction heating has been studied for biofilm eradication [2]. Induction heating has already been proposed for a number of biomedical applications, including ferromagnetic implants [3] and nanoparticle heating [4] for different purposes including cancer treatment. Previous studies have proposed the development of induction-heated autoclaves for sterilization [5]. More specifically, it has been identified as a promising method for removing biofilm in prosthesis [2,6,7]. However, most induction heating used nowadays are based on industrial/domestic systems that are not suitable for clinical application, or use solenoidal coils [8] that provides little flexibility for the surgeon to provide an effective treatment.

In this context, this paper details the design of a portable induction heating system [9–12] for implanted prosthesis disinfection. It is important to note that other systems under research use solenoidal coils, leading to more complex procedures not applicable in many body parts. In this paper, however, a hand-held portable device is presented with improved versatility.

The remainder of this paper is organized as follows. Next section details the proposed device, including the main equations governing the power converter. In the next section, an electromagnetic analysis is performed providing the main equations and FEA results for the proposed system. Finally, the main experimental results are presented, including in-vivo experimentation and the main conclusions are summarized.

2. Proposed device

Despite the fact that induction heating has been identified as promising hyperthermia technique, and it is being applied in other biomedical applications such as nano-particle hyperthermia [13,14], there is no commercial device in the market with the required performance and size, and compatible with the surgical scenario. Although there exists portable induction heating equipment, these devices are highly inefficient and do not have surgical grade isolation or sterilization capabilities. Due to the lack of this technology, current applications of induction heating are limited to in vitro studies, which often uses widely available induction heating cookers as testbeds. This fact severely constraints research possibilities and prevents the translation of this technique to the clinical scenario.

In this context, this paper proposes the development of a new medical device characterized for being highly efficient, from electronic and electromagnetic point of views, which enables a portable sealed implementation that qualifies the proposed device for surgical usage, being the first induction heating device intended for surgical use. Figure 1 shows a representation of the targeted device.

Fig. 1.

Proposed portable IH system.

In order to generate the required alternating current to perform the induction heating process, a voltage-source inverter is used. Figure 2. shows the full-bridge (a) and half-bridge (b) inverters in series resonant configuration [15–17]. This implementation provides a highly-efficiency and versatile implementation. However, the resonant network requires a well-tuned resonant tank to ensure soft-switching. Considering the nature of this application and the fact that it is prone to sudden changes in the equivalent electrical impedance due to device movements, a non-resonant implementation, i.e. without C_r, has been selected.

Fig. 2.

Voltage source inverters: full-bridge (a) and half-bridge (b) series resonant inverters and (c) phase-shift control.

The proposed converter is controlled using phase-shift control (Fig. 2(c)) in a full bridge configuration. The steady-state output voltage can be represented by the following Fourier series, where the amplitud of the voltage is calcultated as follows $\begin{eqnarray}\hat{V}_{oh}=\frac{V_{i}}{h{\pi}}\sqrt{a_{h}^{2}+b_{h}^{2}}.\end{eqnarray}$ (1)

Assuming phase-shift control, parameters a_h and b_h can be easily calculated as a function of the phase-shift 𝛼, obtaining the following expression for the first harmonic $\begin{eqnarray}\hat{V}_{o1}=\frac{4U}{{\pi}}\cos \frac{{\alpha}}{2};\quad {\varphi}_{v1}=\frac{{\alpha}}{2}.\end{eqnarray}$ (2)

Then, the output current can be calculated as follows, assuming that there is no mean voltage applied to the load $\begin{eqnarray}i_{o}(t)=\mathop{\underbrace{I_{o,avg}}_{}}_{=∼0}+\mathop{\sum }_{h=1}^{\infty }\hat{I} _{oh}\sin (h{\omega}_{s}t+{\varphi}_{ih});\quad \hat{I} _{oh}=\frac{\hat{V}_{oh}}{Z_{oh}},\quad {\varphi}_{ih}={\varphi}_{vh}-{\varphi}_{h}.\end{eqnarray}$ (3)

Finally, for a given frequency, the system impedance is calculated as follows: $\begin{eqnarray}Z_{oh}=\sqrt{R_{eq}^{2}+(h{\omega}_{s}L_{eq})^{2}};\quad {\varphi}_{h}=\arctan \frac{h{\omega}_{s}L_{eq}}{R_{eq}}.\end{eqnarray}$ (4)

By using these expressions, the output power of the proposed system as a function of the phase-shift as $\begin{eqnarray}P_{o}=R_{eq}∼I_{o,rms}^{2}=\sum _{h=1}^{\infty }R_{eq}\frac{\hat{I} _{oh}^{2}}{2}=\mathop{\sum }_{h=1}^{\infty }\frac{R_{eq}\hat{V}_{oh}^{2}/2}{R_{eq}^{2}+(h{\omega}_{s}L_{eq})^{2}}.\end{eqnarray}$ (5)

Consequently, the induction coil will be designed to obtain an inductance and resistance such that the required induction heating power is delivered in a suitable frequency range considering audible noise, electromagnetic compatibility, and efficiency.

The temperature required is in the range between 60 °C, to achieve biological effect, and 80 °C, to avoid necrosis in the bone or adjacent tissue. The system is manually controlled to achieve this temperature, which is monitored using an external thermal camera. Under these conditions, the desired output power obtained from preliminary experimental results with human knee prosthesis is 100 W. Considering 230 V mains voltage, the coil is designed to obtain R_eq = 520 mΩ and L_eq = 59 μH values. Figure 3 shows the proposed converter output power variation with the main control parameters, i.e., switching frequency and phase shift α, as well as the main waveforms at maximum and medium output power.

Fig. 3.

Simulation of the proposed full-bridge topology: Power variation with the control parameter frecuency (a) and alpha (b), and main waveforms at 50-kHz with full power (c) and medium power with alpha 25% of the period (d). Red waveform denotes output current and blue waveform denotes inverter output voltage.

3. Electromagnetic analysis

The coupled electromagnetic and thermal problem is governed by solving Maxwell and Fourier equations. The computation of the magnetic flux density and the induced current density is obtained by solving an eddy-current problem in the frequency domain. Considering an impressed current density J_coil and disregarding free charge densities and propagation phenomenon, the fields can be represented by means of the magnetic vector potential A. Applying the Coulomb condition [18], the vector potential in the harmonic formulation satisfies the following diffusion equation: $\begin{eqnarray}{\nabla}^{2}\mathbf{A}-j{\omega}{\sigma}\mathbf{A}=-{\mu}\mathbf{J}_{\text{coil}}\end{eqnarray}$ (6) where 𝜎 is the electrical conductivity of the material. This equation is solved by means of the finite element (FE) method. The solution of the vector potential equation requires external current sources and boundary conditions prescribed as zero-value fields far away from the sources. Moreover, in order to obtain some of the aforementioned magnitudes, such as dissipated power density, power factor, equivalent impedance and efficiency, electric E and magnetic H fields are required. These fields can be straightforwardly obtained from the vector potential as follows: $\begin{eqnarray}\displaystyle \mathbf{E}\text{} & =\text{} & \displaystyle -j{\omega}\mathbf{A}\end{eqnarray}$ (7) $\begin{eqnarray}\displaystyle \mathbf{H}\text{} & =\text{} & \displaystyle \frac{1}{{\mu}}\cdot {\nabla}\times \mathbf{A}.\end{eqnarray}$ (8) At harmonic regime, the power dissipated in the workpiece can be obtained by integrating the real part of Poynting’s vector flux in its surface: $\begin{eqnarray}\dot{Q}=\text{Re}\left[\int _{S}[(\mathbf{E}\times \mathbf{H}^{\ast })\cdot \hat{\mathbf{n}}]dS\right]\end{eqnarray}$ (9) where $\hat{\mathbf{n}}$ is the normal vector to the interest surface. $\dot{Q}$ is the magnitude required to couple the electromagnetic and thermal analysis and the star in the magnetic field means the complex conjugated value. This magnitude is also related with the equivalent resistance of the load, which is part of the electrical circuit connected to the electronic voltage source inverter, as it is shown in Fig. 2(a). This resistance is defined as follows: $\begin{eqnarray}R_{eq}=\text{Re}\left[\int _{S}[(\overline{\mathbf{E}}\times \overline{\mathbf{H}}^{\ast })\cdot \hat{\mathbf{n}}]dS\right]\end{eqnarray}$ (10) where the bars over the electrical and magnetic fields means per unit of ampere magnitudes.

The electrical equivalent circuit of the system also includes an inductor representing the magnetic energy stored in the system. The value of this inductor can be obtained by integrating the electrical field along the trajectory of the windings: $\begin{eqnarray}V_{\text{coil}}=-\oint _{\text{windings}}\mathbf{E}\cdot \mathbf{dl}\Rightarrow L_{eq}=\text{Im}(V_{\text{coil}})/{\omega}\end{eqnarray}$ (11) where 𝜔 is the angular frequency of the currents.

Figure 4 shows some FEA results with different configurations of both the electromagnetic coupler and the prosthesis model. Several coil shapes and sizes were tested. The results with a solenoidal coil are shown in Fig. 4(a). This configuration could be convenient for clinical treatments, however, the power dissipated could be not enough. A configuration with two coils was also analyzed (Fig. 4(b)) and the obtained power dissipated was also low for effective treatments. As a result, electromagnetic couplers with improved capability to focus the magnetic field were also tested, as it is shown in Fig. 4(c). The dissipated power level was considerably higher than in the other cases and, consequently, this configuration was the basis of the practical implementation. This arrangement is based on a PQ40 3C90 ferrite core and a solenoidal coil placed in the central leg. Both size and material were chosen to meet portability and high efficiency requirements. The winding consists of n = 30 turns of a litz wire comprising n_s = 550 strands of diameter 𝜙_s = 50 μm. This wire is appropriate for working up to frequencies of f = 200 kHz with low losses. On the other hand, the electromagnetic properties considered for the prosthesis were 𝜎 = 0.52 MS/m, 𝜇_r = 1, 𝜀_r = 1, which corresponds to the titanium aluminum vanadium (TiAlVa) alloy.

Fig. 4.

Finite element analysis of induction heating of representative prosthesis geometries using different coil configurations. (a) Solenoidal coil and simplified prosthesis model. (b) Disk coils and simplified prosthesis model. (c) PQ-based electromagnetic coupler and disk prosthesis model. (d) Shape-adapted coil.

Finally, some prospective analyses where the coil adapts to the shape of a prosthesis were also carried out, as shown in Fig. 4(d). The preliminary results also show that the power dissipation is not enough for practical applications. Consequently, arrangements that focus the field are chosen for the preliminary tests.

4. Experimental results

In order to test the proposed method for implanted prosthesis disinfection using induction heating, an experimental prototype has been built (Fig. 5) whose main parameters are summarized in Table 1. The prototype includes an electromagnetic compatibility (EMC) filter, a rectifier and an inverter to supply the coil. The power can be controlled by the surgeon using a variable switch. The temperature is monitorized using an external thermal camera. The power converter has been implemented using a full-bridge topology with silicon carbide power MOSFETs for high efficiency and power density. The control signals are controlled by a MSP430 microcontroller from Texas Instruments. To obtain the desired electrical equivalent at a practical application of 1 cm, a 30 turns coil is designed. A button allows the surgeon to turn on the heating, and a potentiometer allows to control the applied power. In the proposed converter, the output power is controlled by adapting the phase angle between the full-bridge legs. Finally, an ETD ferrite core has been selected to build the coil. Figure 6 shows a detail of the power board and the coil as well as the knee prosthesis to be heated. All the prototype is immersed in a thermal polymer for electrical isolation and to provide appropriate thermal management.

Table 1
Main power converter parameters

Parameter Value

Supply characteristics 230 V, less than 1 A.

Output power 100 W

Equivalent resistance 520 mΩ

Equivalent inductance 59 μH

Inductor core PQ40 3C90 ferrite core

Inductor wire 30-turns Litz wire comprising 550 strands of 50 μm diameter

Power devices UJ3C065030T3S

Switching frequency 50 kHz

Control strategy Phase shift 0–180°

Control device MSP430FR2310

Parameter	Value
Supply characteristics	230 V, less than 1 A.
Output power	100 W
Equivalent resistance	520 mΩ
Equivalent inductance	59 μH
Inductor core	PQ40 3C90 ferrite core
Inductor wire	30-turns Litz wire comprising 550 strands of 50 μm diameter
Power devices	UJ3C065030T3S
Switching frequency	50 kHz
Control strategy	Phase shift 0–180°
Control device	MSP430FR2310

Fig. 5.

Prototype block diagram.

Fig. 6.

Experimental prototype: top side (a) and bottom (b) views, and (c) prosthesis.

Figure 7 shows the main waveform for different output power, controlled by changing the phase angle between the half-bridge legs that composes the full-bridge converter. This enables adapting the voltage applied to the coil and, consequently, the applied output power. These waveforms proves the correct operation of the prototype in the complete operation range.

Fig. 7.

Experimental waveforms for maximum (a), medium (b) and minimum (c) output power. From top to bottom: control signals, output voltage, and output current.

Finally, the proposed device has been applied to in-vitro and in-vivo experiments. In-vitro experiments have demonstrated the ability to reduce S. aureus and E. coli concentrations [19]. Figure 8 shows several examples of the device being used to heat a human knee prosthesis (a) as well as during in-vivo experiments in rabbit tissue, proving the effectiveness of the proposed device.

Fig. 8.

Experimental preliminary results: heating effects in a human knee prosthesis reaching 70 ^°C (a), and in-vivo experiment in rabbit tissue (b).

5. Discussion

Nowadays, implanted prosthesis infection is a substantial social and health problem that involves great suffering for the patient and medical costs. While physical cleaning and antibiotic treatments remains ineffective in a significant number of cases, this paper has proposed an effective and versatile tool to combat biofilms in implanted prosthesis. The main novelty of the proposed device is the design of a highly efficient power converter that can be implemented in a high-power density device that can be used with a single hand by the surgeon, providing a versatile tool to achieve hyperthermia. The proposed system uses a planar coil that, combined with a versatile power converter control, allows to deliver the required heating power to the prosthesis areas to be treated. The proposed system has been designed and implemented, and it has allowed to develop in-vitro experiments to assess its accuracy. As it has been presented before, it enables to significantly reduce the concentration of S. aureus and E. Coli bacteria after the treatment, two of the most common species found in severe prosthesis infection. The results have been equally effective both for TiAlV and CrCoMo allows, two of the most common materials in knee prosthesis. These promising results open the window for clinical application and for the proposed system to become a useful tool for implanted prosthesis infection treatment.

The currently existing systems are often based on solenoidal coils and external power converters that are use in research but are not suitable for clinical applications. The proposed system is an integrated handheld device thanks to the use of a high-performance coil made of Litz wire and with improved coupling by using a ferrite flux concentrator. Besides, a high-power-density and high-efficiency power converter has been implemented, providing both low power losses for high integration and a high degree of flexibility due to the non-resonant operation. Unlike most resonant induction heating converters, this allows for flexible and safe operation regardless the induction heating load changes due to the movements made by the surgeon. As a result of this design, the proposed device is considered the first handheld, completely integrated, and sterilizable induction heating device suitable for clinical application.

Currently, the system is manually controlled by the surgeon. It is placed at a distance of 1 cm to avoid contact for hygiene reasons. To improve the magnetic coupling, a ferrite core is used in the device. It is expected that field concentrations in some parts of the prosthesis may occur, but it will be compensated by the movement of the device and conductive heat dispersion.

While the experimental results prove that the proposed system can effectively heat prosthesis to achieve the desired thermal treatment, there are still significant challenges to be addressed. The main limitation is the temperature measurement system, which nowadays is external due to challenges with emissivity calibration due to the different tissues and fluids involved in the process. Future research lines are oriented towards the integration of contactless temperature sensors that may help to ensure both safety and effectiveness without the need of an external thermal camera.

6. Conclusions

In this paper, a portable induction heating device for implanted prosthesis disinfection has been proposed. The design is based on full-bridge implementation that features silicon carbide devices for high-efficiency and high-power-density implementation. Due to its high performance, the device can be immersed in a thermally conductive polymer to obtain a completely sealed implementation ready for usage in surgery procedures. The proposed design has been implemented and tested, showing promising results in in-vitro experiments.

Footnotes

Acknowledgements

This work was partly supported by Projects PID2022-136621OB-I00, PDC2021-120898-I00, TED2021-129274B-I00, CPP2021-008938 and ISCIII PI21/00440, co-funded by MCIN/AEI/10.13039/501100011033 and by EU through FEDER and NextGenerationEU/PRTR programs, by the DGA-FSE.

References

Gbejuade

H.O.

Lovering

A.M.

and Webb

J.C.

, The role of microbial biofilms in prosthetic joint infections, Acta Orthop86(2) (2015), 147–158, doi:10.3109/17453674.2014.966290, (in eng).

Pijls

B.G.

Sanders

I.M.J.G.

Kuijper

E.J.

and Nelissen

R.G.H.H.

, Synergy between induction heating, antibiotics, and N-acetylcysteine eradicates Staphylococcus aureus from biofilm, International Journal of Hyperthermia37(1) (2020), 130–136, doi:10.1080/02656736.2019.1710269.

Stauffer

P.R.

Sneed

P.K.

Hashemi

and Phillips

T.L.

, Practical induction heating coil designs for clinical hyperthermia with ferromagnetic implants, IEEE Transactions on Biomedical Engineering41(1) (1994), 17–28, doi:10.1109/10.277267.

Kwizera

E.A.

Stewart

Mahmud

M.M.

and He

, Magnetic nanoparticle-mediated heating for biomedical applications, J Heat Transfer144(3) (2022), doi:10.1115/1.4053007, (in eng).

Kameda

Ohkuma

Sano

Batbayar

Terashima

and Terada

, Development of a compact induction-heated autoclave with a dramatically shortened sterilization cycle in orthodontic clinics, Orthodontic Waves73(2) (2014), 55–60, doi:10.1016/j.odw.2014.03.001.

Wang

Vachon

Prasad

Pybus

C.A.

Lapin

Chopra

, Alternating magnetic fields and antibiotics eradicate biofilm on metal in a synergistic fashion, NPJ Biofilms and Microbiomes7(1) (2021), 68, doi:10.1038/s41522-021-00239-y.

Pijls

B.G.

Sanders

Kujiper

E.J.

and Nelissen

, Induction heating for eradicating Staphylococcus epidermidis from biofilm, Bone Joint Res9(4) (2020), 192–199, doi:10.1302/2046-3758.94.Bjr-2019-0274.R1, (in eng).

Greenberg

Tepper

and Chopra

, System for application of alternating magnetic fields to reduce infection, Patent Appl. (2022)), [Online]. Available: EP22382889.8 PCT/EP2023/076644.

Lucía

Maussion

Dede

and Burdío

J.M.

, Induction heating technology and its applications: Past developments, current technology, and future challenges, IEEE Transactions on Industrial Electronics61(5) (2014), 2509–2520, doi:10.1109/TIE.2013.2281162, (in English).

10.

Sarnago

Lucía

Mediano

and Burdío

J.M.

, Analysis and design of high-efficiency resonant inverters for domestic induction heating applications, International Journal of Applied Electromagnetics and Mechanics44(2) (2014), 201–208, doi:10.3233/JAE-141760, (in English).

11.

Lucía

Ó.

Sarnago

Acero

and Burdío

J.M.

, Induction heating appliances: Toward more sustainable and smart home appliances, in: Energy Smart Appliances: Applications, Methodologies, and Challenges Moreno-Muñoz

and Giacomini

(eds), Wiley-IEEE Press, USA, 2023.

12.

Lucía

Domínguez

Sarnago

and Burdío

J.M.

, Induction Heating, in: Control of Power Electronic Converters and Systems Blaabjerg

(ed.), 1 edn, Vol. 2, Elsevier, USA, 2018, ch. 21.

13.

Gomez-Polo

Larumbe

Perez-Landazabal

J.I.

and Pastor

J.M.

, Analysis of heating effects (magnetic hyperthermia) in FeCrSiBCuNb amorphous and nanocrystalline wires, Journal of Applied Physics111(7) (2012), 07A314-07A314-3, doi:10.1063/1.3672850.

14.

Sarnago

Lucia

and Burdío

J.M.

, High frequency and power density gallium nitride based inverter for magneto fluid hyperthermia, in: IEEE Annual Conference of the Industrial Electronics Society, 2016, pp. 5222–5225. doi:10.1109/IECON.2016.7794026.

15.

Sarnago

Lucía

Mediano

and Burdío

J.M.

, Analytical model of the half-bridge series resonant inverter for improved power conversion efficiency and performance, IEEE Transactions on Power Electronics30(8) (2015), 4128–4143, doi:10.1109/TPEL.2014.2359576.

16.

Lucía

Burdío

J.M.

Millán

Acero

and Puyal

, Load-adaptive control algorithm of half-bridge series resonant inverter for domestic induction heating, IEEE Transactions on Industrial Electronics56(8) (2009), 3106–3116, doi:10.1109/TIE.2009.2022516, (in English).

17.

Guillén

Sarnago

Lucía

Ó.

and Burdío

J.M.

, Acoustic noise analysis of multiplexed strategies in multi-output converters for induction cooktops, International Journal of Applied Electromagnetics and Mechanics63(s1) (2020), 59–68, doi:10.3233/JAE-209115.

18.

Michalski

K.A.

and Nevels

R.D.

, On the use of the Coulomb gauge in solving source-excited boundary value problems of electromagnetics, IEEE Transactions on Microwave Theory and Techniques36(9) (1988), 1328–1333, doi:10.1109/22.3679.

19.

Cordero

Gómez

Medel

and Esteban

, Desinfección in vitro de componentes metálicos colonizados por biofilms maduros de microorganismos habituales en infección protésica, presented at the Sociedad Española de Cirugía Ortopédica y Traumatología, 2022.