A fast referenceless MR thermometry technique

The single reference variable flip angle sequence with a multi-echo stack of stars acquisition (SR-VFA-SoS) simultaneously measures temperature change using proton resonance frequency (PRF) shift and T1-based thermometry methods. This work evaluates SR-VFA-SoS thermometry in MR-guided focused ultrasound in an in vivo rabbit model. Simultaneous PRF shift thermometry and T1-based thermometry were obtained in a New Zealand white rabbit model (n = 7) during MR-guided focused ultrasound surgery using the SR-VFA-SoS sequence at 3 T. Distinct locations in muscle (n = 16), fat (n = 12), or the interface of both tissues (n = 23) were heated. The T1-temperature coefficient of fat was determined using least-squares fitting of inversion recovery-based T1 maps of untreated fat harvested from the animal and was applied to the in vivo measured heat-induced T1 changes to create temperature maps. Using k-space weighted image contrast reconstruction, temporal resolution of 1.71 s was achieved for simultaneous thermometry at 1.5 × 1.5 × 2 mm voxel resolution. PRF shift thermometry was not sensitive to heating in fat. T1 changes were observed in fat at the ultrasound focus. The mean T1-temperature coefficient for fat was determined to be 1.9%/°C ± 0.2%/°C. Precision was 0.76°C ± 0.18°C for PRF shift thermometry in muscle and 1.93°C ± 0.60°C for T1-based thermometry in fat. Sonications in muscle showed an increase in T1 of 2.4%/°C ± 0.9%/°C. The SR-VFA-SoS sequence was shown to simultaneously measure temperature change using PRF shift and T1-based methods in an in vivo model, providing thermometry for both aqueous and fat tissues.

Referenceless proton resonance frequency (PRF) shift thermometry provides a means to measure temperature changes during minimally invasive thermotherapy that is inherently robust to motion and tissue displacement. However, if the referenceless method is used to determine temperature changes during prostate ablation, phase gaps between water and fat in image regions used to determine the background phase can confound the phase estimation. We demonstrate an extension to referenceless thermometry which eliminates this problem by allowing background phase estimation in the presence of phase discontinuities between aqueous and fatty tissue. In this method, images are acquired with a multi‐echo sequence and binary water and fat maps are generated from a Dixon reconstruction. For the background phase estimation, water and fat regions are treated separately and the phase offset between the two tissue types is determined. The method is demonstrated during in vivo thermal ablation of canine prostate.

The purpose of this study is to demonstrate the first work of T 1-based magnetic resonance thermometry using magnetic resonance fingerprinting (dubbed MRFT). We compared temperature estimation of MRFT with proton resonance frequency shift (PRFS) thermometry on ex vivo bovine muscle. We demonstrated MRFT’s feasibility in predicting temperature on ex vivo bovine muscles with deep brain stimulation (DBS) lead. B 0 maps generated from MRFT were compared with gold standard B 0 maps near the DBS lead. MRFT and PRFS estimated temperatures were compared in the presence of motion. All experiments were performed on a 3 Tesla whole-body GE Premier system with a 21-channel receive head coil (GE Healthcare, Milwaukee, WI). Four fluoroptic probes were used to measure the temperature at the center of a cold muscle (probe 1), the room temperature water bottle (probe 2), and the center and periphery of the heated muscle (probes 3 and 4). We selected regions of interest (ROIs) around the location of the probes and used simple linear regression to generate the temperature sensitivity calibration equations that convert T 1 maps and Δs maps to temperature maps. We then repeated the same setup and compared MRFT and PRFS thermometry temperature estimation with gold standard probe measurements. For the MRFT experiment on DBS lead, we taped the probe to the tip of the DBS lead and used a turbo spin echo sequence to induce heating near the lead. We selected ROIs around the tip of the lead to compare MRFT temperature estimation with probe measurements and compared with PRFS temperature estimation. Vendor-supplied B 0 mapping sequence was acquired to compare with MRFT-generated B 0 maps. We found strong linear relationships (R 2 > 0.958) between T 1 and temperature and Δs and temperatures in our temperature sensitivity calibration experiment. MRFT and PRFS thermometry both accurately predict temperature (RMSE < 1.55 °C) compared to probe measurements. MRFT estimated temperature near DBS lead has a similar trend as the probe temperature. Both B 0 maps show inhomogeneities around the lead. MRFT estimated temperature is less sensitive to motion.

Magnetic resonance (MR) thermometry is a valuable method for characterizing thermal fields generated by high intensity focused ultrasound (HIFU) transducers in tissue phantoms and excised tissues. However, infiltration of noise signals generated by external rf sources into the scanner orifice limits the ability of the scanner to measure temperature rise during the heating or ablation phase. In this study, magnetic resonance interferometry (MRI) monitored HIFU ablations are performed on freshly excised porcine liver samples, at varying sonication times, 20 s, 30 s, and 40 s at a constant acoustic intensity level of 1244 W/cm2. Temperature throughout the procedure was measured using proton resonant frequency MR thermometry. Without filtering, reliable temperature measurements during the heating phase could not be obtained since temperature maps appeared blurred and analysis was impossible. Also, measurements acquired during the cooling phase decayed manifested an unrealistically slow rate of temperature decay. This abnormally slow rate was confirmed with computational results. A low-pass RC filter circuit was subsequently incorporated into the experimental setup to prevent infiltration of noise signals in the MRI orifice. This modified RC filter circuit allowed noninvasive measurement of the HIFU induced temperature rise during the heating phase followed by temperature decay during cooling. The measured data were within 13% agreement with the temperature rise computed by solving the acoustic and heat equations.

Independent assessment of MRI-induced temperature change and SAR distributions in phantoms

To present a three-dimensional (3D) segmented echoplanar imaging (EPI) pulse sequence implementation that provides simultaneously the proton resonance frequency shift temperature of aqueous tissue and the longitudinal relaxation time (T1 ) of fat during thermal ablation. The hybrid sequence was implemented by combining a 3D segmented flyback EPI sequence, the extended two-point Dixon fat and water separation, and the double flip angle T1 mapping techniques. High-intensity focused ultrasound (HIFU) heating experiments were performed at three different acoustic powers on excised human breast fat embedded in ex vivo porcine muscle. Furthermore, T1 calibrations with temperature in four different excised breast fat samples were performed, yielding an estimate of the average and variation of dT1 /dT across subjects. The water only images were used to mask the complex original data before computing the proton resonance frequency shift. T1 values were calculated from the fat-only images. The relative temperature coefficients were found in five fat tissue samples from different patients and ranged from 1.2% to 2.6%/°C. The results demonstrate the capability of real-time simultaneous temperature mapping in aqueous tissue and T1 mapping in fat during HIFU ablation, providing a potential tool for treatment monitoring in organs with large fat content, such as the breast.

There is growing interest in performing hyperthermia treatments with clinical magnetic resonance imaging-guided high-intensity focused ultrasound (MR-HIFU) therapy systems designed for tissue ablation. During hyperthermia treatment, however, due to the narrow therapeutic window (41–45 °C), careful evaluation of the accuracy of proton resonant frequency (PRF) shift MR thermometry for these types of exposures is required.Purpose: The purpose of this study was to evaluate the accuracy of MR thermometry using a clinical MR-HIFU system equipped with a hyperthermia treatment algorithm.Methods: Mild heating was performed in a tissue-mimicking phantom with implanted temperature sensors using the clinical MR-HIFU system. The influence of image-acquisition settings and post-acquisition correction algorithms on the accuracy of temperature measurements was investigated. The ability to achieve uniform heating for up to 40 min was evaluated in rabbit experiments.Results: Automatic centre-frequency adjustments prior to image-acquisition corrected the image-shifts in the order of 0.1 mm/min. Zero- and first-order phase variations were observed over time, supporting the use of a combined drift correction algorithm. The temperature accuracy achieved using both centre-frequency adjustment and the combined drift correction algorithm was 0.57° ± 0.58 °C in the heated region and 0.54° ± 0.42 °C in the unheated region.Conclusion: Accurate temperature monitoring of hyperthermia exposures using PRF shift MR thermometry is possible through careful implementation of image-acquisition settings and drift correction algorithms. For the evaluated clinical MR-HIFU system, centre-frequency adjustment eliminated image shifts, and a combined drift correction algorithm achieved temperature measurements with an acceptable accuracy for monitoring and controlling hyperthermia exposures.

Magnetic Resonance Thermometry (MRT) is demonstrating huge abilities to guide laser interstitial thermal therapy (LITT) in several organs, such as the brain. Among the methods to perform MRT, Proton Resonance Frequency (PRF) shift holds significant benefits, like tissue independence. Despite its potential, PRF shift-based MRT holds significant challenges affecting the accuracy of reconstructed temperature maps. In particular, susceptibility artifacts due to gas-bubble formation are an important source of error in temperature maps in MRT-guided LITT. This work presents the characterization of the susceptibility artifacts in MRT-guided LITT and the measurement of its size. LITT was performed in gelatin-based phantoms, at 5 W, 2 W, 1 W, and 0.5 W under MRI guidance with a 1.5 T clinical MRI scanner. Temperature images were obtained with a 3D EPI (Echo-planar imaging) prototype sequence. Areas of temperature errors were defined as zones of negative temperature variation <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\boldsymbol{&lt; -2^{\circ}\mathrm{C}}$</tex> . Moreover, we have analyzed the artifact shape in sagittal, axial and coronal planes. The analysis demonstrates a double-lobe shape for the susceptibility artifact mainly distributed in the sagittal plane. Also, the higher laser power caused a bigger artifact area. Temperature errors of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\boldsymbol{\sim 80^{\circ}\mathrm{C}}$</tex> proved the necessity to avoid susceptibility artifact generation during MRT-guided LITT. The analysis of the influence of the laser power on the artifact has suggested that using low laser power (0.5 W) helps avoid this measurement error.

The T1 and T2 temperature dependence of female breast adipose tissue was investigated at 1.5 T in order to evaluate the applicability of relaxation-based MR thermometry in fat for the monitoring of thermal therapies in the breast. Relaxation times T1 , T2 and T2TSE (the apparent T2 measured using a turbo spin echo readout sequence) were measured in seven fresh adipose breast samples for temperatures from 25 to 65 °C. Spectral water suppression was used to reduce the influence of the residual water signal. The temperature dependence of the relaxation times was characterized. The expected maximum temperature measurement errors based on average calibration lines were calculated. In addition, the heating-cooling reversibility was investigated for two samples. The T1 and T2TSE temperature (T) dependence could be fitted well with an exponential function of 1/T. A linear relationship between T2 and temperature was found. The temperature coefficients (mean ± inter-sample standard deviation) of T1 and T2TSE increased from 25 °C (dT1/dT = 5.35 ± 0.08 ms/°C, dT2TSE/dT = 3.82 ± 0.06 ms/°C) to 65 °C (dT1 /dT = 9.50 ± 0.16 ms/°C, dT2TSE/dT = 7.99 ± 0.38 ms/°C). The temperature coefficient of T2 was 0.90 ± 0.03 ms/°C. The temperature-induced changes in the relaxation times were found to be reversible after heating to 65 °C. Given the small inter-sample variation of the temperature coefficients, relaxation-based MR thermometry appears to be feasible in breast adipose tissue, and may be used as an adjunct to proton resonance frequency shift (PRFS) thermometry in aqueous tissue (glandular + tumor).

High intensity focused ultrasound (HIFU) is a non-invasive thermal ablation technique. To perform the ablation procedure safely, temperature monitoring is employed to preserve healthy tissues while simultaneously ensuring that the targeted region is completely ablated. Ultrasound (US) thermometry techniques have the advantages of cost-effectiveness and portability over other medical imaging modalities such as MRI. We propose a 3D US thermal monitoring method for HIFU ablation. A US element and sampling device are used to acquire time-of-flight (TOF) information, from which we reconstruct speed of sound (SOS) images to detect the temperature increase during the ablation. We use a physics-based HIFU simulation to segment the ablated region of interest (ROI) to cope with the sparsity of the recorded data. HIFU thermal ablations were performed under MR monitoring on a phantom and the results from the proposed method were compared with MR thermometry. On average, the difference between those two datasets was 1.3°C in the ROI around the ablation focal point, which verifies the feasibility of the proposed method.

The characterization of high intensity focused ultrasound (HIFU) fields is important for both clinical treatment planning as well as for regulation of HIFU medical devices. In previous work, we have used a 100-μm fiber optic probe hydrophone (FOPH) to measure pressure waveforms from a 2-MHz HIFU source with 42-mm aperture and 44-mm focal length. The formation of shock waves with peak positive pressure of up to 80-MPa were measured and modeled in transparent tissue-mimicking gel phantoms and boiling was achieved in milliseconds [Canney MS, et al., J. Acoust. Soc. Am., 120:3110 (2006)]. In this work, the FOPH was also used to measure temperature changes in tissue phantoms from HIFU at peak focal intensities of 5000-20,000 W/cm2. Temperature measurements were obtained by first low-pass filtering the voltage signal measured from the FOPH to remove the acoustic part of the measurement. Then, calibration of voltage to temperature was performed using results from a separate calibration experiment. Experimental measurements were compared with numerical modeling using a KZK-type model for acoustic propagation coupled with a heat transfer model. In summary, temperatures of 100°C were measured at the HIFU focus in milliseconds, in agreement with modeling [Work supported by NIH DK43881, NSBRI SMS00402, and RFBR.]

BackgroundClinical treatment of cardiac arrhythmia by radiofrequency ablation (RFA) currently lacks quantitative and precise visualization of lesion formation in the myocardium during the procedure. This study aims at evaluating thermal dose (TD) imaging obtained from real-time magnetic resonance (MR) thermometry on the heart as a relevant indicator of the thermal lesion extent.MethodsMR temperature mapping based on the Proton Resonance Frequency Shift (PRFS) method was performed at 1.5 T on the heart, with 4 to 5 slices acquired per heartbeat. Respiratory motion was compensated using navigator-based slice tracking. Residual in-plane motion and related magnetic susceptibility artifacts were corrected online. The standard deviation of temperature was measured on healthy volunteers (N = 5) in both ventricles. On animals, the MR-compatible catheter was positioned and visualized in the left ventricle (LV) using a bSSFP pulse sequence with active catheter tracking. Twelve MR-guided RFA were performed on three sheep in vivo at various locations in left ventricle (LV). The dimensions of the thermal lesions measured on thermal dose images, on 3D T1-weighted (T1-w) images acquired immediately after the ablation and at gross pathology were correlated.ResultsMR thermometry uncertainty was 1.5 °C on average over more than 96% of the pixels covering the left and right ventricles, on each volunteer. On animals, catheter repositioning in the LV with active slice tracking was successfully performed and each ablation could be monitored in real-time by MR thermometry and thermal dosimetry. Thermal lesion dimensions on TD maps were found to be highly correlated with those observed on post-ablation T1-w images (R = 0.87) that also correlated (R = 0.89) with measurements at gross pathology.ConclusionsQuantitative TD mapping from real-time rapid CMR thermometry during catheter-based RFA is feasible. It provides a direct assessment of the lesion extent in the myocardium with precision in the range of one millimeter. Real-time MR thermometry and thermal dosimetry may improve safety and efficacy of the RFA procedure by offering a reliable indicator of therapy outcome during the procedure.

Both pefluorocarbon microbubbles and nanodroplets have been investigated as enhancers of high intensity focused ultrasound (HIFU) thermal ablation, however microbubbles often lead to surface or skin lesions. We have designed and investigated a dual-perfluorocarbon (PFC) nanodroplet which has the benefits of sufficiently small size to extravasate from tumors, enhanced stability at body temperature, and sufficiently low acoustic threshold for vaporization. In vitro, microbubbles enhanced thermal depostion at the target site by 21%, but were found to cause surface heating up to 60.2±2.2°C. Nanodroplets caused no more surface heating (10.1±1.1°C) than the temperature rise observed in agent-free controls (9.8±0.8°C), and enhanced heating at the target by 51%. Circulation time of the nanodroplets was investigated in vivo. HIFU (1 MHz, 4.06 MPa, CW, 15 seconds) was applied to rat livers (n=3) up to 95 minutes after nanodroplet injection, and any thermal enhancement was detected simultaneously by MR thermometry. Temperature rises of up to 55 degrees above body temperature were observed out to 95 minutes. HIFU applied to control livers without nanodroplets induced only a 22°C maximal temperature rise. These results suggest that the nanodroplets are sufficiently stable to enhance HIFU ablation in vivo for at least 1.5 hours and could reduce focused ultrasound surgical procedure times by as much as 5 fold by more quickly ablating a larger region of tissue, without compromising safety.

In MRI-guided high-intensity focused ultrasound (MR-HIFU) therapy, temperature-dependent proton resonance frequency (PRF) shift is a key factor to quantify and visualize the spatial heating pattern in treated and surrounding tissue. However, during treatment, multiple scanner-related changes can impact the accuracy of the temperature measurements obtained with the PRF shift method and cause an over/under estimation of temperature, which can be a major safety issue for treatments involving real-time temperature control. Hence, it is necessary to apply corrections to ensure accurate temperature measurements during heating. Prior to image acquisition, the central MR frequency F0 can be measured to adjust the F0 of next image acquisition. After acquisition, corrections can be applied to the acquired images to remove scanner-related influences, most importantly phase drift. Different phase drift correction algorithms such as conventional and polynomial adaptive drift correction estimate the background phase by fitting a linear or polynomial to the image phase outside the treatment area and perform the correction accordingly. The goal of this study was to understand the performance of these algorithms for long heating durations as would be experienced during hyperthermia or transurethral HIFU (>20 minutes).

MR-guided high-intensity focused ultrasound (HIFU) for treatment of uterine fibroids and painful bone metastases uses the proton resonant frequency shift (PRF) for temperature monitoring in water-based tissues. However, PRF fails to detect temperature changes in tissues with high lipid content, such as fat and bone marrow. Previous studies have shown a change in T2 of subcutaneous fat, red and yellow bone marrow during treatments with focused ultrasound. The lack of calibration data for 3T acquisitions, however, makes it difficult to convert T2 values into maps of tissue temperature. In this study we investigated the dependence of T2 temperature on temperature in porcine adipose tissue and bovine yellow bone marrow at thermal equilibrium at 3T.