Data-Driven Motion Detection and Event-by-Event Correction for Brain PET: Comparison with Vicra

Abstract

Head motion degrades image quality and causes erroneous parameter estimates in tracer kinetic modeling in brain PET studies. Existing motion correction methods include frame-based image registration (FIR) and correction using real-time hardware-based motion tracking (HMT) information. However, FIR cannot correct for motion within 1 predefined scan period, and HMT is not readily available in the clinic since it typically requires attaching a tracking device to the patient. In this study, we propose a motion correction framework with a data-driven algorithm, that is, using the PET raw data itself, to address these limitations. Methods: We propose a data-driven algorithm, centroid of distribution (COD), to detect head motion. In COD, the central coordinates of the line of response of all events are averaged over 1-s intervals to generate a COD trace. A point-to-point change in the COD trace in 1 direction that exceeded a user-defined threshold was defined as a time point of head motion, which was followed by manually adding additional motion time points. All the frames defined by such time points were reconstructed without attenuation correction and rigidly registered to a reference frame. The resulting transformation matrices were then used to perform the final motion-compensated reconstruction. We applied the new COD framework to 23 human dynamic datasets, all containing large head motion, with 18 F-FDG (n 5 13) and 11 C-UCB-J ((R)-1-((3-( 11 C-methyl-11 C)pyridin-4-yl)methyl)-4-(3,4,5-trifluorophenyl)pyrrolidin-2-one) (n 5 10) and compared its performance with FIR and with HMT using Vicra (an optical HMT device), which can be considered the gold standard. Results: The COD method yielded a 1.0% ± 3.2% (mean ± SD across all subjects and 12 gray matter regions) SUV difference for 18 F-FDG (3.7% ± 5.4% for 11 C-UCB-J) compared with HMT, whereas no motion correction (NMC) and FIR yielded −15.7% ± 12.2% (−20.5% ± 15.8%) and −4.7% ± 6.9% (−6.2% ± 11.0%), respectively. For 18 F-FDG dynamic studies, COD yielded differences of 3.6% ± 10.9% in K i value as compared with HMT, whereas NMC and FIR yielded −18.0% ± 39.2% and −2.6% ± 19.8%, respectively. For 11 C-UCB-J, COD yielded 3.7% ± 5.2% differences in V T compared with HMT, whereas NMC and FIR yielded −20.0% ± 12.5% and −5.3% ± 9.4%, respectively. Conclusion: The proposed COD-based data-driven motion correction method outperformed FIR and achieved comparable or even better performance than the Vicra HMT method in both static and dynamic studies.

Optimizing the frame duration for data‐driven rigid motion estimation in brain PET imaging

“…Data-driven techniques generally fall into two groups, both of which split the data up into many frames. The first approach uses some format of the data, for example, downsampled sinograms, 3 centroid-of-density calculations, 4 or full reconstructions, 5,6 to identify time points where motion occurred. The data are reframed according to these time points, reconstructions of each new frame are performed, and the subsequent reconstructions are aligned and combined.…”

Conditional Generative Adversarial Networks Aided Motion Correction of Dynamic 18F-FDG PET Brain Studies

“…The problem with motion artefacts is even more severe when an IDIF is extracted from the dynamic PET sequence, given that the magnitude of random displacements and the system resolution are typically larger than the small size of the arteries. Accordingly, accurate motion correction is an important prerequisite for absolute quantification in PET imaging (17).…”

“…general approaches: data-driven approaches (11)(12)(13)(14)(15)(16)(17), frame-based image-registration (FIR) 87 (18,19) and real-time hardware motion tracking (HMT) (20). Real-time hardware-based motion tracking detects subject motion with excellent temporal resolution (20), but is typically not used 89 in clinical routine due to its complexity and the necessity to integrate external data (motion 90 tracking) with the imaging system (applying the motion vector to images).…”

Deep Learning Based Joint PET Image Reconstruction and Motion Estimation

“…One motion correction method is the event-by-event correction based on list-mode data processing. Motion is measured using a special motion detection hardware, such as a POLARIS [4] or Anzai AZ-733V system (Anzai Medical Co, Ltd., Tokyo, Japan), or using a data-driven approach [5], [6]. The event-driven motion compensation is performed by transforming the line-of-response (LOR) along which the event is measured to the position it would have been measured if the object had not moved [7].…”