Deep Learning-aided Respiratory Motion Compensation in PET/CT: Addressing Motion Induced Resolution Loss, Attenuation Correction Artifacts and PET-CT Misalignment

Overview

Journal Eur J Nucl Med Mol Imaging

Specialties Biophysics
Nuclear Medicine
Radiology

Date 2024 Aug 13

PMID 39136740

Authors

Yihuan Lu

Fei Kang

Duo Zhang

Yue Li

Hao Liu

Chen Sun

Hao Zeng

Lei Shi

Yumo Zhao

Jing Wang

Affiliations

Soon will be listed here.

Abstract

Purpose: Respiratory motion (RM) significantly impacts image quality in thoracoabdominal PET/CT imaging. This study introduces a unified data-driven respiratory motion correction (uRMC) method, utilizing deep learning neural networks, to solve all the major issues caused by RM, i.e., PET resolution loss, attenuation correction artifacts, and PET-CT misalignment.

Methods: In a retrospective study, 737 patients underwent [F]FDG PET/CT scans using the uMI Panorama PET/CT scanner. Ninety-nine patients, who also had respiration monitoring device (VSM), formed the validation set. The remaining data of the 638 patients were used to train neural networks used in the uRMC. The uRMC primarily consists of three key components: (1) data-driven respiratory signal extraction, (2) attenuation map generation, and (3) PET-CT alignment. SUV metrics were calculated within 906 lesions for three approaches, i.e., data-driven uRMC (proposed), VSM-based uRMC, and OSEM without motion correction (NMC). RM magnitude of major organs were estimated.

Results: uRMC enhanced diagnostic capabilities by revealing previously undetected lesions, sharpening lesion contours, increasing SUV values, and improving PET-CT alignment. Compared to NMC, uRMC showed increases of 10% and 17% in SUV and SUV across 906 lesions. Sub-group analysis showed significant SUV increases in small and medium-sized lesions with uRMC. Minor differences were found between VSM-based and data-driven uRMC methods, with the SUV was found statistically marginal significant or insignificant between the two methods. The study observed varied motion amplitudes in major organs, typically ranging from 10 to 20 mm.

Conclusion: A data-driven solution for respiratory motion in PET/CT has been developed, validated and evaluated. To the best of our knowledge, this is the first unified solution that compensates for the motion blur within PET, the attenuation mismatch artifacts caused by PET-CT misalignment, and the misalignment between PET and CT.

Citing Articles

Quantitative analysis of patient motion in walk-through PET scanner and standard axial field of view pet scanner using infrared-based tracking.

Aziz R, Maebe J, Muller F, DAsseler Y, Vandenberghe S EJNMMI Phys. 2024; 11(1):99.

PMID: 39581945 PMC: 11586328. DOI: 10.1186/s40658-024-00704-5.

References

Ren S, Jin X, Chan C, Jian Y, Mulnix T, Liu C . Data-driven event-by-event respiratory motion correction using TOF PET list-mode centroid of distribution. Phys Med Biol. 2017; 62(12):4741-4755. PMC: 6048592. DOI: 10.1088/1361-6560/aa700c. View

Allen-Auerbach M, Yeom K, Park J, Phelps M, Czernin J . Standard PET/CT of the chest during shallow breathing is inadequate for comprehensive staging of lung cancer. J Nucl Med. 2006; 47(2):298-301. View

Hwang D, Kang S, Kim K, Choi H, Seo S, Lee J . Data-driven respiratory phase-matched PET attenuation correction without CT. Phys Med Biol. 2021; 66(11). DOI: 10.1088/1361-6560/abfc8f. View

Hashimoto F, Onishi Y, Ote K, Tashima H, Reader A, Yamaya T . Deep learning-based PET image denoising and reconstruction: a review. Radiol Phys Technol. 2024; 17(1):24-46. PMC: 10902118. DOI: 10.1007/s12194-024-00780-3. View

Chan C, Onofrey J, Jian Y, Germino M, Papademetris X, Carson R . Non-Rigid Event-by-Event Continuous Respiratory Motion Compensated List-Mode Reconstruction for PET. IEEE Trans Med Imaging. 2017; 37(2):504-515. PMC: 7304524. DOI: 10.1109/TMI.2017.2761756. View

Lee H, Son H, Yun M, Moon J, Kim Y, Woo J . Prone position [F]FDG PET/CT to reduce respiratory motion artefacts in the evaluation of lung nodules. Eur Radiol. 2021; 31(7):4606-4614. DOI: 10.1007/s00330-021-07894-x. View

Schaefferkoetter J, Shah V, Hayden C, Prior J, Zuehlsdorff S . Deep learning for improving PET/CT attenuation correction by elastic registration of anatomical data. Eur J Nucl Med Mol Imaging. 2023; 50(8):2292-2304. DOI: 10.1007/s00259-023-06181-9. View

Tiss A, Chemli Y, Guehl N, Marin T, Johnson K, El Fakhri G . Effects of List-Mode-Based Intraframe Motion Correction in Dynamic Brain PET Imaging. IEEE Trans Radiat Plasma Med Sci. 2024; 8(8):950-958. PMC: 11540417. DOI: 10.1109/trpms.2024.3432322. View

Gatidis S, Hepp T, Fruh M, Fougere C, Nikolaou K, Pfannenberg C . A whole-body FDG-PET/CT Dataset with manually annotated Tumor Lesions. Sci Data. 2022; 9(1):601. PMC: 9532417. DOI: 10.1038/s41597-022-01718-3. View

10.

Quirk S, Becker N, Smith W . External respiratory motion analysis and statistics for patients and volunteers. J Appl Clin Med Phys. 2013; 14(2):4051. PMC: 5714366. DOI: 10.1120/jacmp.v14i2.4051. View

11.

Changlai S, Huang C, Luzhbin D, Lin F, Wu J . Using Cine-Averaged CT With the Shallow Breathing Pattern to Reduce Respiration-Induced Artifacts for Thoracic Cavity PET/CT Scans. AJR Am J Roentgenol. 2019; 213(1):140-146. DOI: 10.2214/AJR.18.20606. View

12.

Wasserthal J, Breit H, Meyer M, Pradella M, Hinck D, Sauter A . TotalSegmentator: Robust Segmentation of 104 Anatomic Structures in CT Images. Radiol Artif Intell. 2023; 5(5):e230024. PMC: 10546353. DOI: 10.1148/ryai.230024. View

13.

Pan T, Thomas M, Luo D . Data-driven gated CT: An automated respiratory gating method to enable data-driven gated PET/CT. Med Phys. 2022; 49(6):3597-3611. PMC: 9187617. DOI: 10.1002/mp.15620. View

14.

Dikaios N, Fryer T . Registration-weighted motion correction for PET. Med Phys. 2012; 39(3):1253-64. DOI: 10.1118/1.3675922. View

15.

Meng Q, Yang R, Wu R, Xu L, Liu H, Yang G . Evaluation of a respiratory motion-corrected image reconstruction algorithm in 2-[F]FDG and [Ga]Ga-DOTA-NOC PET/CT: impacts on image quality and tumor quantification. Quant Imaging Med Surg. 2023; 13(1):370-383. PMC: 9816722. DOI: 10.21037/qims-22-557. View

16.

Ren S, Lu Y, Bertolli O, Thielemans K, Carson R . Event-by-event non-rigid data-driven PET respiratory motion correction methods: comparison of principal component analysis and centroid of distribution. Phys Med Biol. 2019; 64(16):165014. DOI: 10.1088/1361-6560/ab0bc9. View

17.

Hu H, Huang Y, Sun H, Zhou K, Jiang L, Zhong J . A proper protocol for routine F-FDG uEXPLORER total-body PET/CT scans. EJNMMI Phys. 2023; 10(1):51. PMC: 10495295. DOI: 10.1186/s40658-023-00573-4. View

18.

Shi L, Zhang J, Toyonaga T, Shao D, Onofrey J, Lu Y . Deep learning-based attenuation map generation with simultaneously reconstructed PET activity and attenuation and low-dose application. Phys Med Biol. 2022; 68(3). DOI: 10.1088/1361-6560/acaf49. View

19.

Siva S, Pham D, Gill S, Bressel M, Dang K, Devereux T . An analysis of respiratory induced kidney motion on four-dimensional computed tomography and its implications for stereotactic kidney radiotherapy. Radiat Oncol. 2013; 8:248. PMC: 3829388. DOI: 10.1186/1748-717X-8-248. View

20.

Balamoutoff N, Serrano B, Hugonnet F, Garnier N, Paulmier B, Faraggi M . Added Value of a Single Fast 20-second Deep-Inspiration Breath-hold Acquisition in FDG PET/CT in the Assessment of Lung Nodules. Radiology. 2017; 286(1):260-270. DOI: 10.1148/radiol.2017160534. View