A statistical study of the factors influencing the extent of respiratory motion blur in PET imaging

https://doi.org/10.1016/j.compbiomed.2011.10.002Get rights and content

Abstract

Respiratory motion results in significant motion blur in thoracic and abdomen PET imaging. The extent of respiratory motion blur is mainly correlated with breathing amplitude, tumor size and location. In this paper we introduce a statistical study to quantitatively show the factors influencing the extent of respiratory motion blur in thoracic PET images. The study is centered on two regression models, one is linked with motion blur induced loss of mean intensity(LMI), tumor motion magnitude and tumor size, and another is to investigate the influence of tumor location, patient gender and patient height on tumor motion magnitude. We use the blur identification and image restoration technique to estimate the tumor motion and compute the LMI. The regression model was validated by simulation and phantom data before extended to 39 cases of clinical lung tumor PET images corrupted with blurring artifact. Results show that the motion magnitude of lung tumor during breathing is 10.9±3.7 mm in transaxial plane, and it is significantly greater in lower lung lobes than in upper lobes. The LMI is 7.1±2.4% in the region of interest (ROI) above 40% of the image's maximum intensity. The least-square estimate of regression equations demonstrates that LMI is proportional to tumor motion magnitude and is inversely proportional to tumor size; the two factors play the same role in determining the extent of respiratory motion blur in thoraco-abdominal PET imaging. The location of tumor was shown as the major factor determining its motion magnitude, while the influencing of patient gender and height on tumor motion was not shown significant.

Introduction

Respiratory motion is a main source of degradation in thoracic positron emission tomography (PET) and single photon emission computed tomography (SPECT) imaging due to the long scanning time of imaging system. The respiratory motion blur in PET or SPECT images results from the dispersed radiotracer activity over an area proportional to the breathing motion magnitude [1]. Studies have shown that the respiratory motion blur is harmful both to diagnosis and PET/CT guided radiotherapy: it not only degrades PET images by reducing spatial resolution, tumor-to-background contrast and signal-to-noise ratio (SNR) level, distorting the shape and location of the tumor [2], [3], but also affects the accuracy of quantitation, reduces the measured standard uptake value (SUV), causes the tumor size to be overestimated and thus increases the planned target volume [4], [5].

The breathing magnitude is known as one of the major factors to affect the extent of the respiratory motion blur. The effective spatial resolution of PET imaging is depended on the intrinsic resolution full width at half maximum (FWHM) and the amplitude of breathing and can be computed as FWHMsystem2+amplitude2 [6]. It was found that the thoraco-abdominal tumors present displacements in the range of 6–23 mm mainly in the super-inferior and posterior–anterior direction, depending on their locations [2]. Respiratory motion blur observed at the lung/diaphragm interface is normally more serious than other locations since the diaphragm moves with the largest breathing magnitude among all organs. In [7] the average magnitude of lung tumor motion in the lower lobes was found to be 12 mm. Studies in [8] showed that a 20 mm motion magnitude can result in a loss of average activity between 25% and 45%. Besides motion magnitude, lesion size is another major factor influencing the severity of respiratory motion blur, small lung tumors or pulmonary nodules are inclined to suffer from serious blurring artifacts thus is hard to be detected in PET imaging [9], [10].

Previous studies mostly revealed the factors influencing respiratory motion blur by visual comparing the original blurred images with motion corrected or deblurred images [4]. Respiration gated PET/CT acquisition in conjunction with reconstruction or registration based motion correction algorithms are normally used to correct the motion blur [3], [10], [11], [12]. In gated acquisition the breathing magnitude of internal lesions can be estimated from the displacement of external markers recorded by tracking devices [7], [13], [14]. A post-reconstruction deconvolution was also proposed in [15], [16] to reduce the motion blur directly in ungated PET image domain, where the respiratory motion and corresponding point spread function (PSF) was estimated from gated CT.

The purpose of this study is to statistically show the factors influencing the extent of respiratory motion blur in PET images by regression analysis. The main usage of obtained regression equations is to estimate the effect of breathing on thoracic PET or SPECT imaging. The loss of mean intensity (LMI) in the region-of-interest is used as an index to represent the extent of motion blur. The rest of this paper is organized as follows: in Section 2 we first propose the regression models linked with LMI, tumor motion magnitude, tumor size and location and patient gender and height, then we introduce the method of blur identification and image restoration, which is used to estimate the respiration motion in image domain and derive the LMI. Section 3 is the results and discussions; the regression equations obtained from simulation data was tested by phantom SPECT data degraded with known motion, then we extend the work to clinical PET data to obtain meaningful regression equations. Section 4 gives conclusions.

Section snippets

Regression models

In general, there exists certain quantitative correlation between the image motion blur, tumor motion magnitude and tumor size. In nuclear medical imaging, PET or SPECT images corrupted with blurring artifact represent a decreased intensity level (number of counts), spatial resolution, contrast and SNR. Since the variation of image intensity is easier to compute than other parameters, we adopt the percentage loss of mean intensity (LMI) in the region of interest (ROI) to describe the extent of

Simulation study

The simulation study is designed to theoretically validate the regression models (3), (4). The static synthetic image size is 128×128, the ROI is a centered square object with uniform intensity and the background is zero intensity (square object is selected since the LMI caused by simulated motion blur can be more accurately computed than ROI in other shape). To investigate the impact of blur extent L and ROI size on motion blur, the square size is changed from 1×1 to 24×24, and L varies from

Conclusions

We presented a statistical study to quantitatively show the factors influencing the extent of respiratory motion blur in PET imaging. We proposed regression models linked with the motion blur induced loss of mean intensity (LMI), tumor motion magnitude, tumor size, patient gender and patient height. The blur identification and image restoration technique was employed to estimate the tumor motion and compute the LMI. Regression results were shown consistent in simulation, phantom and clinical

Conflict of interest statement

We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work; there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be constructed as influencing the position presented.

Acknowledgment

This work was supported by the National Natural Science Foundation of China (Grant no. 30730036).

References (26)

  • L. Boucher et al.

    Respiratory gating for 3-dimensional PET of the thorax: feasibility and initial results

    J. Nucl. Med.

    (2004)
  • G. Chang et al.

    Implementation of an automated respiratory amplitude gating technique for PET/CT: clinical evaluation

    J. Nucl. Med.

    (2010)
  • J. Wang et al.

    Computer-assisted quantification of lung tumors in respiratory gated PET/CT images: phantom study

    Med. Biol. Eng. Comput.

    (2008)
  • Cited by (0)

    1

    Contributed equally to this work.

    View full text