| Zugriffsnummer | 39485 |
| Dokumenttyp | Konferenzartikel |
| Sprache | Englisch |
| Titel | Gradient heating of bulk metallic implants |
| Autor(in); Institution |
Brühl, Rüdiger; 8.1, Biomedizinische Magnetresonanz, PTB-Berlin
Ihlenfeld, Albrecht; 8.1, Biomedizinische Magnetresonanz, PTB-Berlin
Ittermann, Bernd; 8.1, Biomedizinische Magnetresonanz, PTB-Berlin
|
| Quelle/Jahr | ISMRM 25th annual meeting & exhibition: 22 - 27 April 2017, Honolulu : IN: Proceedings of the Society of Magnetic Resonance in Medicine: 25 (2017), 2620-1 - 2620-3 |
| Availability | [online only] |
| Konferenzangaben | ISMRM 25th Annual Meeting & Exhibition, Honolulu, Hawaii, 22-27, April, 2017, USA |
| Zusammenfassung | Synopsis The temperature increase of an excised hip prosthesis was measured under exposure to the switched gradients of a clinical 3T scanner. For the acetabular cup, insulated or embedded in gelatin gel, temperature increases of ΔT=25.8 K and 3.8 K, respectively, were observed within 10 min. From the initial temperature increase of up to 110 mK/s a gradient-induced heating power of PG = 370 W/kg in adjacent muscle can be derived and this quantity's relation to local SAR is discussed. The results suggest that gradient-induced heating of bulk metallic implants cannot automatically be assumed to be negligible. INTRODUCTION The possibility of hazardous tissue heating near metallic implants is well established1-3 for the RF case but studies looking at gradient-switching effects are scarce.4Recently, a numerical simulation study of gradient induced heating of a hip prosthesis showed a temperature increase of up to 20 K in the implant and the surrounding tissue.5 This seemed to contradict Ref. (4) where no measurable effects were found in a realistic prosthesis. The present work aims to measure the gradient-induced power absorption and heating of a common hip prosthesis on a clinical scanner. METHODS Cup (Aesculap Plasmafit Plus3, 54 mm diameter) and stem (Aesculap Excia, 180 mm) of an excised hip prosthesis, both made from Ti-6Al-4V alloy, were investigated separately, both either insulated in polystyrene or embedded in gelatin gel (Fig. 1). Temperature was monitored by fiber-optical sensors. The implants were placed at positions P1 (x=−15cm,y=−9cm,z=−30cm)(x=−15cm,y=−9cm,z=−30cm) or P2 (x=−20cm,y=0cm,z=0cm)(x=−20cm,y=0cm,z=0cm) in a clinical 3T scanner (Siemens Verio). P1 was chosen according to manufacturer gradient maps for maximum dB/dt dB/dt in a realistic patient position, P2 corresponds to Ref. (4). The stem was oriented in the natural direction whereas the cup's symmetry axis was parallel to the z-direction. For maximum effect a self-written EPI-like sequence with continuous trapezoidal gradients (Gmax=20mT/m,f=2.0kHz,S=200Tm−1s−1)(Gmax=20mT/m,f=2.0kHz,S=200Tm−1s−1) was applied; the product EPI and bSSFP sequences were used for comparison. Experiments were performed in normal operating mode with all safety features in place; patient scanning would have been permissible. RF was off. RESULTS Fig. 2 shows the temperature change of the cup under trapezoidal z-gradients. Increases of ΔT=25.8 K and 3.8 K within 10 min were observed for the insulated and embedded cup, respectively, at P1. After 40 min ΔT was 6.0 K for the embedded implant and not yet saturated. If the adjacent tissue follows the implant temperature, as it was found in Ref. (5), a gradient-induced specific heating power PGPG can be derived via PG=k0cTPG=k0cT, where k0≡dT/dt∣t=0k0≡dT/dt∣t=0 is the initial slope of the temperature curve and cTcT the tissue's specific heat capacity. For the insulated implant k0=110mK/sk0=110mK/s is determined; the value for the embedded implant is consistent within (much larger) errors. For, e.g. muscle with6 cT=3400Jkg−1K−1cT=3400Jkg−1K−1 this results in PG=370W/kgPG=370W/kg. Heating rates for product sequences were lower but could still be substantial: k0=40mK/sk0=40mK/s (EPI, P1), k0=6.8mK/sk0=6.8mK/s (bSSFP, P1), k0=2.6mK/sk0=2.6mK/s (bSSFP, P2). Heat generation in the prosthesis stem was about an order of magnitude lower, which was partially offset by lower heat losses. At the implant position the self-written sequence generated an average dB/dt rms=49.7T/s dB/dt rms=49.7T/s, i.e. less than the FPO:B limit1 of 56T/s56T/s for implant scanning. |
