Accelerated Diffusion-Weighted MRI of Rectal Cancer Using a Residual Convolutional Network.

This work presents a deep-learning-based denoising technique to accelerate the acquisition of high b-value diffusion-weighted MRI for rectal cancer. A denoising convolutional neural network (DCNN) with a combined L1-L2 loss function was developed to denoise high b-value diffusion-weighted MRI data acquired with fewer repetitions (NEX: number of excitations) using the low b-value image as an anatomical guide. DCNN was trained using 85 datasets acquired on patients with rectal cancer and tested on 20 different datasets with NEX = 1, 2, and 4, corresponding to acceleration factors of 16, 8, and 4, respectively. Image quality was assessed qualitatively by expert body radiologists. Reader 1 scored similar overall image quality between denoised images with NEX = 1 and NEX = 2, which were slightly lower than the reference. Reader 2 scored similar quality between NEX = 1 and the reference, while better quality for NEX = 2. Denoised images with fourfold acceleration (NEX = 4) received even higher scores than the reference, which is due in part to the effect of gas-related motion in the rectum, which affects longer acquisitions. The proposed deep learning denoising technique can enable eightfold acceleration with similar image quality (average image quality = 2.8 ± 0.5) and fourfold acceleration with higher image quality (3.0 ± 0.6) than the clinical standard (2.5 ± 0.8) for improved diagnosis of rectal cancer.

High-intensity focused ultrasound ablation of muscle in an anticoagulated swine model.

INTRODUCTION: Image-guided non-invasive high-intensity focused ultrasound (HIFU) has been gaining recognition in treating musculoskeletal tumors and desmoids. However, there is no consensus on the appropriate perioperative management for patients on ongoing anticoagulation who undergo HIFU ablation. MATERIAL AND METHODS: Image-guided HIFU treatment was performed in swine on an ongoing oral anticoagulation protocol (N = 5) in two treatment sessions seven days apart. On day one, a total of twenty locations were ablated, and on day eight, ten more muscle ablations were performed, and the animals were euthanized. Imaging, clinical examination, and histopathology were performed to investigate treated tissue for bleeding. RESULTS: Imaging, clinical examination, and histopathology revealed either no bleeding or, in some samples, only small scattered cavities (0.2-2 mm in diameter) filled with blood. CONCLUSION: Noninvasive HIFU ablation of muscle may not require a coagulation profile within normal limits.

Accelerating Prostate Diffusion-weighted MRI Using a Guided Denoising Convolutional Neural Network: Retrospective Feasibility Study.

PURPOSE: To investigate the feasibility of accelerating prostate diffusion-weighted imaging (DWI) by reducing the number of acquired averages and denoising the resulting image using a proposed guided denoising convolutional neural network (DnCNN). MATERIALS AND METHODS: Raw data from the prostate DWI scans were retrospectively gathered between July 2018 and July 2019 from six single-vendor MRI scanners. There were 103 datasets used for training (median age, 64 years; interquartile range [IQR], 11), 15 for validation (median age, 68 years; IQR, 12), and 37 for testing (median age, 64 years; IQR, 12). High b-value diffusion-weighted (hb DW) data were reconstructed into noisy images using two averages and reference images using all 16 averages. A conventional DnCNN was modified into a guided DnCNN, which uses the low b-value DW image as a guidance input. Quantitative and qualitative reader evaluations were performed on the denoised hb DW images. A cumulative link mixed regression model was used to compare the readers' scores. The agreement between the apparent diffusion coefficient (ADC) maps (denoised vs reference) was analyzed using Bland-Altman analysis. RESULTS: Compared with the original DnCNN, the guided DnCNN produced denoised hb DW images with higher peak signal-to-noise ratio (32.79 ± 3.64 [standard deviation] vs 33.74 ± 3.64), higher structural similarity index (0.92 ± 0.05 vs 0.93 ± 0.04), and lower normalized mean square error (3.9% ± 10 vs 1.6% ± 1.5) (P < .001 for all). Compared with the reference images, the denoised images received higher image quality scores from the readers (P < .0001). The ADC values based on the denoised hb DW images were in good agreement with the reference ADC values (mean ADC difference ranged from -0.04 to 0.02 × 10-3 mm2/sec). CONCLUSION: Accelerating prostate DWI by reducing the number of acquired averages and denoising the resulting image using the proposed guided DnCNN is technically feasible. Supplemental material is available for this article. © RSNA, 2020.

Volumetric 3D assessment of ablation zones after thermal ablation of colorectal liver metastases to improve prediction of local tumor progression.

PURPOSE: The goal of this study was to develop and evaluate a volumetric three-dimensional (3D) approach to improve the accuracy of ablation margin assessment following thermal ablation of hepatic tumors. METHODS: The 3D margin assessment technique was developed to generate the new 3D assessment metrics: volumes of insufficient coverage (VICs) measuring volume of tissue at risk post-ablation. VICs were computed for the tumor and tumor plus theoretical 5- and 10-mm margins. The diagnostic accuracy of the 3D assessment to predict 2-year local tumor progression (LTP) was compared to that of manual 2D assessment using retrospective analysis of a patient cohort that has previously been reported as a part of an outcome-centered study. Eighty-six consecutive patients with 108 colorectal cancer liver metastases treated with radiofrequency ablation (2002-2012) were used for evaluation. The 2-year LTP discrimination power was assessed using receiver operating characteristic area under the curve (AUC) analysis. RESULTS: A 3D assessment of margins was successfully completed for 93 out of 108 tumors. The minimum margin size measured using the 3D method had higher discrimination power compared with the 2D method, with an AUC value of 0.893 vs. 0.790 (p = 0.01). The new 5-mm VIC metric had the highest 2-year LTP discrimination power with an AUC value of 0.923 (p = 0.004). CONCLUSIONS: Volumetric semi-automated 3D assessment of the ablation zone in the liver is feasible and can improve accuracy of 2-year LTP prediction following thermal ablation of hepatic tumors. KEY POINTS: • More accurate prediction of local tumor progression risk using volumetric 3D ablation zone assessment can help improve the efficacy of image-guided percutaneous thermal ablation of hepatic tumors. • The accuracy of evaluation of ablation zone margins after thermal ablation of colorectal liver metastases can be improved using a volumetric 3D semi-automated assessment approach and the volume of insufficient coverage assessment metric. • The new 5-mm volume-of-insufficient-coverage metric, indicating the volume of tumor plus 5-mm margin that remained untreated, had the highest 2-year local tumor progression discrimination power.

Ablation of the sacroiliac joint using MR-guided high intensity focused ultrasound: a preliminary experiment in a swine model.

BACKGROUND: Dysfunction of the Sacroiliac Joint (SIJ) is one of the key sources of low back pain. For prolonged pain relief, some patients undergo fluoroscopic guided radio-frequency (RF) ablation of SIJ, during which a number of RF probes are inserted to create thermal lesions that disrupt the posterior sacral nerve supply. This procedure is minimally invasive, laborious, time-consuming and costly. To study if High Intensity Focused Ultrasound (HIFU) ablation is a feasible alternative approach to SIJ pain treatment, we performed experiments using HIFU to ablate SIJ in the swine model. METHODS: Three female Yorkshire swine (36, 35.2 and 34 kg) underwent bilateral Magnetic Resonance guided HIFU (MRgHIFU) ablation of the SIJs. Treatment assessment was performed using contrast-enhanced imaging, histopathology and evaluation of pain and changes in ambulation and gait. RESULTS: Contiguous lesions along the right and left SIJs were achieved in all animals. In one out of three animals, excessive heating of the muscle and skin tissue in the near-field resulted in unwanted muscle necrosis. No changes in animal behavior, ambulation or gait were detected. CONCLUSIONS: The initial experiments with MRgHIFU ablation of SIJs in sub-acute swine model show promise for this ablation modality as a non invasive and more precise alternative to the currently used fluoroscopically - guided RF ablations and injections.

Development of a Searchable Database of Cryoablation Simulations for Use in Treatment Planning.

PURPOSE: To create and validate a planning tool for multiple-probe cryoablation, using simulations of ice ball size and shape for various ablation probe configurations, ablation times, and types of tissue ablated. MATERIALS AND METHODS: Ice ball size and shape was simulated using the Pennes bioheat equation. Five thousand six hundred and seventy different cryoablation procedures were simulated, using 1-6 cryoablation probes and 1-2 cm spacing between probes. The resulting ice ball was measured along three perpendicular axes and recorded in a database. Simulated ice ball sizes were compared to gel experiments (26 measurements) and clinical cryoablation cases (42 measurements). The clinical cryoablation measurements were obtained from a HIPAA-compliant retrospective review of kidney and liver cryoablation procedures between January 2015 and February 2016. Finally, we created a web-based cryoablation planning tool, which uses the cryoablation simulation database to look up the probe spacing and ablation time that produces the desired ice ball shape and dimensions. RESULTS: Average absolute error between the simulated and experimentally measured ice balls was 1 mm in gel experiments and 4 mm in clinical cryoablation cases. The simulations accurately predicted the degree of synergy in multiple-probe ablations. The cryoablation simulation database covers a wide range of ice ball sizes and shapes up to 9.8 cm. CONCLUSION: Cryoablation simulations accurately predict the ice ball size in multiple-probe ablations. The cryoablation database can be used to plan ablation procedures: given the desired ice ball size and shape, it will find the number and type of probes, probe configuration and spacing, and ablation time required.

MRI-guided focused ultrasound ablation of lumbar medial branch nerve: Feasibility and safety study in a swine model.

PURPOSE: About 10-40% of chronic low back pain cases involve facet joints, which are commonly treated with lumbar medial branch (MB) radiofrequency neurotomy. Magnetic resonance imaging-guided focused ultrasound (MRgFUS), a non-invasive, non-ionising ablation modality used to treat tumours, neuropathic pain and painful bone metastasis can also be used to disrupt nerve conduction. This work's purpose was to study the feasibility and safety of direct MRgFUS ablation of the lumbar MB nerve in acute and subacute swine models. MATERIALS AND METHODS: In vivo MRgFUS ablation was performed in six swine (three acute and three subacute) using a clinical MRgFUS system and a 3-T MRI scanner combination. Behavioural assessment was performed, and imaging and histology were used to assess the treatment. RESULTS AND CONCLUSIONS: Histological analysis of the in vivo studies confirmed thermal necrosis of the MB nerve could be achieved without damaging the spinal cord or adjacent nerve roots. MRgFUS did not cause changes in the animals' behaviour or ambulation.

Closed-Bore Interventional MRI: Percutaneous Biopsies and Ablations.

OBJECTIVE: The purpose of this article is to review clinical applications and technologic development of MRI-guided percutaneous interventions performed in closed-bore MRI scanners. CONCLUSION: Interventional MRI has rapidly adapted to the closed-bore environment. New tools are being developed to facilitate the use of MRI-guided procedures, and cost-effectiveness studies are exploring the economics of interventional MRI.

Feasibility Study on MR-Guided High-Intensity Focused Ultrasound Ablation of Sciatic Nerve in a Swine Model: Preliminary Results.

INTRODUCTION: Spastic patients often seek neurolysis, the permanent destruction of the sciatic nerve, for better pain management. MRI-guided high-intensity focused ultrasound (MRgHIFU) may serve as a noninvasive alternative to the prevailing, more intrusive techniques. This in vivo acute study is aimed at performing sciatic nerve neurolysis using a clinical MRgHIFU system. METHODS: The HIFU ablation of sciatic nerves was performed in swine (n = 5) using a HIFU system integrated with a 3 T MRI scanner. Acute lesions were confirmed using T1-weighted contrast-enhanced (CE) MRI and histopathology using hematoxylin and eosin staining. The animals were euthanized immediately following post-ablation imaging. RESULTS: Reddening and mild thickening of the nerve and pallor of the adjacent muscle were seen in all animals. The HIFU-treated sections of the nerves displayed nuclear pyknosis of Schwann cells, vascular hyperemia, perineural edema, hyalinization of the collagenous stroma of the nerve, myelin sheet swelling, and loss of axons. Ablations were visible on CE MRI. Non-perfused volume of the lesions (5.8-64.6 cc) linearly correlated with estimated lethal thermal dose volume (4.7-34.2 cc). Skin burn adjacent to the largest ablated zone was observed in the first animal. Bilateral treatment time ranged from 55 to 138 min, and preparation time required 2 h on average. CONCLUSION: The acute pilot study in swine demonstrated the feasibility of a noninvasive neurolysis of the sciatic nerve using a clinical MRgHIFU system. Results revealed that acute HIFU nerve lesions were detectable on CE MRI, gross pathology, and histology.

Feasibility of intermittent pneumatic compression for venous thromboembolism prophylaxis during magnetic resonance imaging-guided interventions.

PURPOSE: Venous thromboembolism (VTE) is a common cause of morbidity and mortality in hospitalized and surgical patients. To reduce risk, perioperative VTE prophylaxis is recommended for cancer patients undergoing surgical or interventional procedures. Magnetic resonance imaging (MRI) is increasingly used in interventional oncology when alternative imaging modalities do not adequately delineate malignancies. Extended periods of immobilization during MRI-guided interventions necessitate an MR compatible sequential compression device (SCD) for intra-procedural mechanical VTE prophylaxis. Such devices are not commercially available. MATERIALS AND METHODS: A standard SCD routinely used at our institution for VTE prophylaxis during interventional procedures was used. To satisfy MR safety requirements, the SCD controller was placed in the MR control room and connected to the compression sleeves in the magnet room through the wave guide using tubing extensions. The controller pressure sensor was used to monitor adequate pressure delivery and detect ineffective low or abnormal high pressure delivery. VTE prophylaxis was provided using the above mentioned device for 38 patients undergoing MR-guided ablations. RESULTS: There was no evidence of device failure due to loss of pressure in the extension tubing assembly. No interference with the anesthesia or interventional procedures was documented. CONCLUSION: Although the controller of a standard SCD is labeled as "MR-unsafe", the SCD can be used in interventional MR settings by placing the device outside the MR scanner room. Using serial tubing extensions did not cause device failure. The described method can be used to provide perioperative mechanical VTE prophylaxis for high risk patients undergoing MR-guided procedures.

Adapting MRI acoustic radiation force imaging for in vivo human brain focused ultrasound applications.

A variety of magnetic resonance imaging acoustic radiation force imaging (MR-ARFI) pulse sequences as the means for image guidance of focused ultrasound therapy have been recently developed and tested ex vivo and in animal models. To successfully translate MR-ARFI guidance into human applications, ensuring that MR-ARFI provides satisfactory image quality in the presence of patient motion and deposits safe amount of ultrasound energy during image acquisition is necessary. The first aim of this work was to study the effect of motion on in vivo displacement images of the brain obtained with 2D Fourier transform spin echo MR-ARFI. Repeated bipolar displacement encoding configuration was shown less sensitive to organ motion. The optimal signal-to-noise ratio of displacement images was found for the duration of encoding gradients of 12 ms. The second aim was to further optimize the displacement signal-to-noise ratio for a particular tissue type by setting the time offset between the ultrasound emission and encoding based on the tissue response to acoustic radiation force. A method for measuring tissue response noninvasively was demonstrated. Finally, a new method for simultaneous monitoring of tissue heating during MR-ARFI acquisition was presented to enable timely adjustment of the ultrasound energy aimed at ensuring the safety of the MR-ARFI acquisition.

Application of Zernike polynomials towards accelerated adaptive focusing of transcranial high intensity focused ultrasound.

PURPOSE: To study the phase aberrations produced by human skulls during transcranial magnetic resonance imaging guided focused ultrasound surgery (MRgFUS), to demonstrate the potential of Zernike polynomials (ZPs) to accelerate the adaptive focusing process, and to investigate the benefits of using phase corrections obtained in previous studies to provide the initial guess for correction of a new data set. METHODS: The five phase aberration data sets, analyzed here, were calculated based on preoperative computerized tomography (CT) images of the head obtained during previous transcranial MRgFUS treatments performed using a clinical prototype hemispherical transducer. The noniterative adaptive focusing algorithm [Larrat et al., "MR-guided adaptive focusing of ultrasound," IEEE Trans. Ultrason. Ferroelectr. Freq. Control 57(8), 1734-1747 (2010)] was modified by replacing Hadamard encoding with Zernike encoding. The algorithm was tested in simulations to correct the patients' phase aberrations. MR acoustic radiation force imaging (MR-ARFI) was used to visualize the effect of the phase aberration correction on the focusing of a hemispherical transducer. In addition, two methods for constructing initial phase correction estimate based on previous patient's data were investigated. The benefits of the initial estimates in the Zernike-based algorithm were analyzed by measuring their effect on the ultrasound intensity at the focus and on the number of ZP modes necessary to achieve 90% of the intensity of the nonaberrated case. RESULTS: Covariance of the pairs of the phase aberrations data sets showed high correlation between aberration data of several patients and suggested that subgroups can be based on level of correlation. Simulation of the Zernike-based algorithm demonstrated the overall greater correction effectiveness of the low modes of ZPs. The focal intensity achieves 90% of nonaberrated intensity using fewer than 170 modes of ZPs. The initial estimates based on using the average of the phase aberration data from the individual subgroups of subjects was shown to increase the intensity at the focal spot for the five subjects. CONCLUSIONS: The application of ZPs to phase aberration correction was shown to be beneficial for adaptive focusing of transcranial ultrasound. The skull-based phase aberrations were found to be well approximated by the number of ZP modes representing only a fraction of the number of elements in the hemispherical transducer. Implementing the initial phase aberration estimate together with Zernike-based algorithm can be used to improve the robustness and can potentially greatly increase the viability of MR-ARFI-based focusing for a clinical transcranial MRgFUS therapy.

Rapid MR-ARFI method for focal spot localization during focused ultrasound therapy.

MR-guided focused ultrasound (FUS) is a noninvasive therapy for treating various pathologies. MR-based acoustic radiation force imaging (MR-ARFI) measures tissue displacement in the focal spot due to acoustic radiation force. MR-ARFI also provides feedback for adaptive focusing algorithms that could correct for phase aberrations caused by the skull during brain treatments. This work developed a single-shot echo-planar imaging-based MR-ARFI method that reduces scan time and ultrasound energy deposition. The new method was implemented and tested in a phantom and ex vivo brain tissue. The effect of the phase aberrations on the ultrasound focusing was studied using displacement maps obtained with echo-planar imaging and two-dimensional spin-warp MR-ARFI. The results show that displacement in the focal spot can be rapidly imaged using echo-planar imaging-based MR-ARFI with high signal-to-noise ratio efficiency and without any measurable tissue heating. Echo-planar imaging-based displacement images also demonstrate sufficient sensitivity to phase aberrations and can serve as rapid feedback for adaptive focusing in brain treatments and other applications.

Consistency of signal intensity and T2* in frozen ex vivo heart muscle, kidney, and liver tissue.

PURPOSE: To investigate tissue dependence of the MRI-based thermometry in frozen tissue by quantification and comparison of signal intensity and T2* of ex vivo frozen tissue of three different types: heart muscle, kidney, and liver. MATERIALS AND METHODS: Tissue samples were frozen and imaged on a 0.5 Tesla MRI scanner with ultrashort echo time (UTE) sequence. Signal intensity and T2* were determined as the temperature of the tissue samples was decreased from room temperature to approximately -40 degrees C. Statistical analysis was performed for (-20 degrees C, -5 degrees C) temperature interval. RESULTS: The findings of this study demonstrate that signal intensity and T2* are consistent across three types of tissue for (-20 degrees C, -5 degrees C) temperature interval. CONCLUSION: Both parameters can be used to calculate a single temperature calibration curve for all three types of tissue and potentially in the future serve as a foundation for tissue-independent MRI-based thermometry.