A framework for continuous target tracking during MR-guided high intensity focused ultrasound thermal ablations in the abdomen

Abstract

BackgroundDuring lengthy magnetic resonance-guided high intensity focused ultrasound (MRg-HIFU) thermal ablations in abdominal organs, the therapeutic work-flow is frequently hampered by various types of physiological motion occurring at different time-scales. If left un-addressed this can lead to an incomplete therapy and/or to tissue damage of organs-at-risk. While previous studies focus on correction schemes for displacements occurring at a particular time-scale within the work-flow of an MRg-HIFU therapy, in the current work we propose a motion correction strategy encompassing the entire work-flow.MethodsThe proposed motion compensation framework consists of several linked components, each being adapted to motion occurring at a particular time-scale. While respiration was addressed through a fast correction scheme, long term organ drifts were compensated using a strategy operating on time-scales of several minutes. The framework relies on a periodic examination of the treated area via MR scans which are then registered to a reference scan acquired at the beginning of the therapy. The resulting displacements were used for both on-the-fly re-optimization of the interventional plan and to ensure the spatial fidelity between the different steps of the therapeutic work-flow. The approach was validated in three complementary studies: an experiment conducted on a phantom undergoing a known motion pattern, a study performed on the abdomen of 10 healthy volunteers and during 3 in-vivo MRg-HIFU ablations on porcine liver.ResultsResults have shown that, during lengthy MRg-HIFU thermal therapies, the human liver and kidney can manifest displacements that exceed acceptable therapeutic margins. Also, it was demonstrated that the proposed framework is capable of providing motion estimates with sub-voxel precision and accuracy. Finally, the 3 successful animal studies demonstrate the compatibility of the proposed approach with the work-flow of an MRg-HIFU intervention under clinical conditions.ConclusionsIn the current study we proposed an image-based motion compensation framework dedicated to MRg-HIFU thermal ablations in the abdomen, providing the possibility to re-optimize the therapy plan on-the-fly with the patient on the interventional table. Moreover, we have demonstrated that even under clinical conditions, the proposed approach is fully capable of continuously ensuring the spatial fidelity between the different phases of the therapeutic work-flow.

In vivo tibiofemoral cartilage strain mapping under static mechanical loading using continuous GRASP‐MRI

“…A user‐defined weighting factor alpha is set that reflects the elasticity of the tracked object. For the 3D motion tracking for this application, the value of alpha was set to 0.4 as a good compromise between performance and accuracy . The regions of interest (ROIs) were manually segmented for each slice and binary masks were created to represent femoral and tibial articular cartilage.…”

“…The regions of interest (ROIs) were manually segmented for each slice and binary masks were created to represent femoral and tibial articular cartilage. The binary masks and the motion‐corrected 3D‐time‐series were used as input to the optical flow measurement algorithm . The 3D‐flow tracking algorithm uses the image intensity levels to estimate the 3D‐motion deformation maps between each timepoint with reference to the first image in the time series.…”

Accelerating multi-modal image registration using a supervoxel-based variational framework

“…(iii) Study #3: The benefit of the algorithm in a clinical context was initially investigated by employing it for the estimation slow physiological drifts which frequently occur during lengthy MRg-HIFU therapies. This implied the registration of 3D T1 to T2-weighted images acquired in the scope of a motion compensation protocol, specifically dedicated to such interventions (described in detail within (Zachiu, Denis de Senneville, Ries & Moonen 2015) and (Zachiu et al 2017a)).…”

“…Over the duration of lengthy MRg-HIFU therapies, abdominal organs may undergo long-term physiological drifts due to several factors such as muscle relaxation and/or digestive and metabolic activity. Such displacements were demonstrated to become significant over time intervals of several minutes and, if not taken into consideration, may impair attaining the therapeutic endpoint , Zachiu et al 2017a. In order to capture the aforementioned slow physiological drifts, 3D MR-images were periodically acquired on the abdomen four healthy volunteers.…”

“…For each volunteer, two pairs of 3D T1/T2-weighted images, were acquired ≃ 60 min apart and selected for the role of reference image and respectively image to register. During those 60 min, each volonteer underwent the acquisition protocol described in (Zachiu et al 2017a). The T1-weighted sequence employed the following parameters: TE=2ms, TR= 4.3ms, image matrix 192 × 192 × 75, with an isotropic voxel size of 2 × 2 × 2 mm 3 , 10 • flip angle, resulting in an acquisition time of 60-90s, depending on the volunteer's breathing cycle reproducibility.…”