PURPOSE: Patient motion during dynamic PET studies is a well-documented source of errors. The purpose of this study was to investigate the incidence of frame-to-frame motion in dynamic ( 15)O-water myocardial perfusion PET studies, to test the efficacy of motion correction methods and to study whether implementation of motion correction would have an impact on the perfusion results. METHODS: We developed a motion detection procedure using external radioactive skin markers and frame-to-frame alignment. To evaluate motion, marker coordinates inside the field of view were determined in each frame for each study. The highest number of frames with identical spatial coordinates during the study were defined as "non-moved". Movement was considered present if even one marker changed position, by one pixel/frame compared with reference, in one axis, and such frames were defined as "moved". We tested manual, in-house-developed motion correction software and an automatic motion correction using a rigid body point model implemented in MIPAV (Medical Image Processing, Analysis and Visualisation) software. After motion correction, remaining motion was re-analysed. Myocardial blood flow (MBF) values were calculated for both non-corrected and motion-corrected datasets. RESULTS: At rest, patient motion was found in 18% of the frames, but during pharmacological stress the fraction increased to 45% and during physical exercise it rose to 80%. Both motion correction algorithms significantly decreased (p<0.006) the number of moved frames and the amplitude of motion (p<0.04). Motion correction significantly increased MBF results during bicycle exercise (p<0.02). At rest or during adenosine infusion, the motion correction had no significant effects on MBF values. CONCLUSION: Significant motion is a common phenomenon in dynamic cardiac studies during adenosine infusion but especially during exercise. Applying motion correction for the data acquired during exercise clearly changed the MBF results, indicating that motion correction is required for these studies.
PURPOSE:Patient motion during dynamic PET studies is a well-documented source of errors. The purpose of this study was to investigate the incidence of frame-to-frame motion in dynamic ( 15)O-water myocardial perfusion PET studies, to test the efficacy of motion correction methods and to study whether implementation of motion correction would have an impact on the perfusion results. METHODS: We developed a motion detection procedure using external radioactive skin markers and frame-to-frame alignment. To evaluate motion, marker coordinates inside the field of view were determined in each frame for each study. The highest number of frames with identical spatial coordinates during the study were defined as "non-moved". Movement was considered present if even one marker changed position, by one pixel/frame compared with reference, in one axis, and such frames were defined as "moved". We tested manual, in-house-developed motion correction software and an automatic motion correction using a rigid body point model implemented in MIPAV (Medical Image Processing, Analysis and Visualisation) software. After motion correction, remaining motion was re-analysed. Myocardial blood flow (MBF) values were calculated for both non-corrected and motion-corrected datasets. RESULTS: At rest, patient motion was found in 18% of the frames, but during pharmacological stress the fraction increased to 45% and during physical exercise it rose to 80%. Both motion correction algorithms significantly decreased (p<0.006) the number of moved frames and the amplitude of motion (p<0.04). Motion correction significantly increased MBF results during bicycle exercise (p<0.02). At rest or during adenosine infusion, the motion correction had no significant effects on MBF values. CONCLUSION: Significant motion is a common phenomenon in dynamic cardiac studies during adenosine infusion but especially during exercise. Applying motion correction for the data acquired during exercise clearly changed the MBF results, indicating that motion correction is required for these studies.
Authors: Anu Juslin; Jyrki Lötjönen; Sergey V Nesterov; Kari Kalliokoski; Juhani Knuuti; Ulla Ruotsalainen Journal: J Nucl Cardiol Date: 2007-01 Impact factor: 5.952
Authors: Edward P Ficaro; Venkatesh L Murthy; Benjamin C Lee; Jonathan B Moody; Alexis Poitrasson-Rivière; Amanda C Melvin; Richard L Weinberg; James R Corbett Journal: J Nucl Cardiol Date: 2018-03-23 Impact factor: 5.952
Authors: Jonghye Woo; Balaji Tamarappoo; Damini Dey; Ryo Nakazato; Ludovic Le Meunier; Amit Ramesh; Joel Lazewatsky; Guido Germano; Daniel S Berman; Piotr J Slomka Journal: Med Phys Date: 2011-11 Impact factor: 4.071
Authors: Yuka Otaki; Serge D Van Kriekinge; Chih-Chun Wei; Paul Kavanagh; Ananya Singh; Tejas Parekh; Marcelo Di Carli; Jamshid Maddahi; Arkadiusz Sitek; Christopher Buckley; Daniel S Berman; Piotr J Slomka Journal: Eur J Nucl Med Mol Imaging Date: 2021-12-30 Impact factor: 10.057