A learning-based method for online adjustment of C-arm Cone-beam CT source trajectories for artifact avoidance

Abstract

Purpose During spinal fusion surgery, screws are placed close to critical nerves suggesting the need for highly accurate screw placement. Verifying screw placement on high-quality tomographic imaging is essential. C-arm cone-beam CT (CBCT) provides intraoperative 3D tomographic imaging which would allow for immediate verification and, if needed, revision. However, the reconstruction quality attainable with commercial CBCT devices is insufficient, predominantly due to severe metal artifacts in the presence of pedicle screws. These artifacts arise from a mismatch between the true physics of image formation and an idealized model thereof assumed during reconstruction. Prospectively acquiring views onto anatomy that are least affected by this mismatch can, therefore, improve reconstruction quality. Methods We propose to adjust the C-arm CBCT source trajectory during the scan to optimize reconstruction quality with respect to a certain task, i.e., verification of screw placement. Adjustments are performed on-the-fly using a convolutional neural network that regresses a quality index over all possible next views given the current X-ray image. Adjusting the CBCT trajectory to acquire the recommended views results in non-circular source orbits that avoid poor images, and thus, data inconsistencies. Results We demonstrate that convolutional neural networks trained on realistically simulated data are capable of predicting quality metrics that enable scene-specific adjustments of the CBCT source trajectory. Using both realistically simulated data as well as real CBCT acquisitions of a semianthropomorphic phantom, we show that tomographic reconstructions of the resulting scene-specific CBCT acquisitions exhibit improved image quality particularly in terms of metal artifacts. Conclusion The proposed method is a step toward online patient-specific C-arm CBCT source trajectories that enable high-quality tomographic imaging in the operating room. Since the optimization objective is implicitly encoded in a neural network trained on large amounts of well-annotated projection images, the proposed approach overcomes the need for 3D information at run-time.

Accuracy Assessment of Percutaneous Pedicle Screw Placement Using Cone Beam Computed Tomography with Metal Artifact Reduction

“…Wu et al [15] compared MAR with different C-arm trajectories in reducing blooming artifacts on CBCT reconstructions and demonstrated that non-circular orbits reduced metal artifacts by 46% compared to circular orbits. Thies et al [16] introduced a MAR technique that uses non-circular C-arm orbits with intraoperative adjustments of X-ray source trajectory to optimize the image reconstruction quality. These adjustments are based on a machine-learning algorithm using convolutional neural networks, which can predict quality metrics that enable scene-specific adjustments of the CBCT source trajectory, thus improving the image quality and reducing metal artifacts.…”

“…This method can also be extended by known-component image reconstruction [13,14]. Another technique uses optimized C-arm orbits in order to avoid collinearity between the X-ray source and screws [15,16]. Figures 1 and 2 demonstrate ARSN images for the baseline technology without MAR (NoMAR) in comparison to images using MAR, which is based on the interpolation of projection data neighboring the metal shadow around the pedicle screw [10].…”

The Impact of Machine Learning on 2D/3D Registration for Image-guided Interventions: A Systematic Review and Perspective

“…In many other -from a research perspective perhaps more exciting -cases, this retrospective data collection paradigm is infeasible because the task to be performed with a machine learning algorithm is not currently performed in clinical practice. The more obvious examples are visual servoing of novel robotic surgery platforms [88] or robotic imaging paradigms that alter how data is acquired [93,94]. Despite the fact that most studies included in this review address use-cases that fall under the "opportunity cost" category, we found that only 16 out of the 48 studies used real clinical or cadaveric data to train the machine learning algorithms.…”

The Impact of Machine Learning on 2D/3D Registration for Image-Guided Interventions: A Systematic Review and Perspective

“…In many other—from a research perspective perhaps more exciting—cases, this retrospective data collection paradigm is infeasible because the task to be performed with a machine learning algorithm is not currently performed in clinical practice. The more obvious examples are visual servoing of novel robotic surgery platforms ( Gao et al, 2019 ) or robotic imaging paradigms that alter how data is acquired ( Zaech et al, 2019 ; Thies et al, 2020 ). Despite the fact that most studies included in this review address use-cases that fall under the “opportunity cost” category, we found that only 16 out of the 48 studies used real clinical or cadaveric data to train the machine learning algorithms.…”