Strategies to Improve CT Dose Optimization for Hybrid PET/CT Imaging

Background: To evaluate whether current dose reduction strategies for the CT component of hybrid Positron Emission Tomography—Computed Tomography (PET/CT) systems could reduce patient dose with maintaining adequate image quality for PET/CT studies. Materials and Methods: Literature survey was initially based on the selection of keywords and years of publication to identify potentially relevant articles, then the further search was conducted on the authors and references from these articles. The abstract of each article was first appraised to decide whether the content was relevant to this research question. The articles were classified into five groups: studies on dosimetry, studies on radiation-induced diseases, studies on dose reduction methods for CT-based attenuation correction (CTAC), studies on dose reduction methods for CT localization, and studies on reducing the need for a full-dose diagnostic CT in PET/CT imaging. 58 peer-reviewed articles were selected and appraised and 29 articles were used to compose this literature review. Results: The published nuclear medicine and medical physics literature were reviewed. CT dose contributed 47% 81% of the total effective dose of a standard PET/CT study and was associated with radiation-induced diseases. The dose reduction techniques were extracted and divided into three categories: reducing the CT dose for attenuation correction (AC) and localization, selectively localizing CT use, and reducing the need for a full-dose diagnostic CT. Conclusion: Three strategies have been demonstrated, with high potential for reducing patient dose while maintaining an adequate CT image quality, used for CTCA localization and diagnosis, respectively.


Introduction
Over the past 30 years, PET imaging has improved from a research tool to a strong imaging tool. The continues development of hardware, software, imaging techniques and radiopharmaceuticals, lead to expanded the PET role in many areas, including, but not limited to cardiology, neurology, endocrinology, image-guided therapy planning, biopsy, delineation of pathologic volume, treatment management and pharmacology research [1] [2].
PET provides functional information according to the bio-distribution of an administered radiopharmaceutical. However, precise localization of functional information with anatomical reference is the challenging because of the relatively low spatial resolution. The theoretical spatial resolution of FDG 18 PET imaging is approximately 1 mm, but the spatial resolution of clinical PET imaging is between 3.3 and 7 mm [3]. The anatomical landmarks that are demonstrated on CT or magnetic resonance image (MRI) images are often blurry or absent on PET images. In addition to that, the uptake of PET radiopharmaceuticals may be non-specific [3].
The first hybrid PET/CT system for clinical usage was made commercially available in 2001 [4]. Nowadays, as the significant advantage of hybrid PET/CT compared with stand-alone PET has grown, the major manufacturers do not produce stand-alone PET anymore [3] [5] [6] [7].
Although hybrid PET/CT has made a considerable clinical impact, but the frequent usage will be generating growing concern about the increase of ionizing radiation that exposure patients and lead to the possibility of radiation-induced diseases [8].

Dosimetry of PET/CT
The PET effective dose is typically calculated based on the Medical Internal Radiation Dose (MIRD) schema, that developed by the MIRD Committee of the Society of Nuclear Medicine. The CT dose of PET/CT is determined by the same way as a stand-alone CT, typically characterized by dosimetry metrics such as computed tomography (CT) dose index (CTDIvol) and dose-length product (DLP) [9].
CT dose depends on many parameters such as current, voltage, gantry rotation speed, pitch, and patient characteristics. The radiation doses of 18F-FDG PET/CT scans are different between countries, clinics, equipment, protocols and patients. The overall dose of 18F-FDG PET/CT ranges from 8 mSv to 29.8 mSv.
The CT component appears to account at least 54% from the total dose of a whole-body, with a maximum of 81% [8].
A retrospective study conducted by Huang et al. (2009) reported that an average of 18F-FDG PET effective dose was 6.2 mSv, and CT effective dose was 13 -24.8 mSv (accounting for 54% -81% from the total effective dose) [10]. Wu et al. (2004) found that an average PET effective dose was 10.7 mSv, and for CT effective dose was 19.0 mSv (accounting for 64% of the total effective Open Journal of Medical Imaging dose) [9]. A study on oncology patients conducted by Khamwan et al. (2010) reported an average PET dose of 4.4 mSv and a CT dose of 14.5 mSv (contributing 77% of the total effective dose) [11]. A recent study on pediatric patients by Kornerup et al. (2015) found that an average of CT effective dose of 35.0 mSv, accounting for 75% of the total effective dose from PET/CT scans that have high-resolution diagnostic quality [2].

Radiation-Induced Health Risk
The relationship between ionizing radiation exposure and radiation-induced disease is becoming increasingly noteworthy. The health risks of ionizing radiation exposure from medical imaging studies have become an important issue, drawing attention from both medical professionals and the public during the past 30 years [2] [12] [13].
The research conducted by Huang et al. (2009) is the first studied published to investigate the rate of cancer incidence induced by ionizing radiation from hybrid PET/CT imaging, though the study covered only the U.S. and Hong Kong populations [10].
The estimation of the effective dose was based on dose coefficients and tissue weighting factors recommended by ICRP publication 80 and ICRP publication 103. The average 18F-FDG PET effective dose was 6.23 mSv. The CT effective doses varied between 7.22 mSv and 27.3 mSv [12] [13] [14].
The sum of the effective doses from PET and CT ranged from 13.45 mSv to 32.18 mSv. As a result, the CT effective dose accounted for 54% -81% of the total effective dose. The lifetime attributable risk (LAR) of cancer was calculated by the method introduced in the BEIR VII report. Huang et al. (2009) [10] demonstrated that a person who received an effective dose of 32.18 mSv from an 18F-FDG PET/CT scan at the age of 20 was expected to have LAR of cancer of 0.514% (for the U.S. population) or 0.622% (for the Hong Kong population); LAR of cancer increases as age decreases ("Biological Effects of Ionizing Radiation (BEIR) VII report", Committee on the Biological Effects of Ionizing, 2006

Materials and Methods
The literature survey was initially based on the selection of keywords (PET/CT; dose; reduction; CTAC and/or localization and/or diagnostic CT) and years of publication (2010-2020). Then from these results, further research was conducted on the authors and references of these articles.

Results
58 peer-reviewed articles were selected and reviewed, and 31 articles were used to compose this literature review. The published nuclear medicine and medical physics literature indicate that CT dose contributed 47% -81% of the total effective dose of a routine PET/CT study, and was associated with radiation-induced diseases. The dose reduction techniques were extracted and grouped into three categories: reducing the CT dose for attenuation correction (AC) and localization; selectively using localization CT; and reducing the need for full-dose diagnostic CT scans. Study on automatic tube current modulation (ATCM) and reconstruction algorithms were not intentionally reviewed, as they are not dosed reduction methods specific to the CT component of PET/CT [22].

CT-Based Attenuation Correction (AC) Strategy
The CT dose for AC can be 10 -100 times lower than that of a diagnostic CT, as the tube current of ultra-low-dose CTAC can be reduced to as low as 5 -10 mA  [20].
The result showed the reduction in the current (from 80 to 10 mA) at 80 keV resulted in a 400% increase in CT noise, but only a 2% increase in the PET image Open Journal of Medical Imaging noise in all phantoms. With the pediatric phantoms, the minimum current was reduced to 5 mA at 80 kVp; with adult phantoms, the minimum current was 10 mA. An increase in CT noise did not correspond to a significant increase in PET noise; the PET images reconstructed by CTAC with different current settings (5 -80 mA) were indistinguishable [25].
The reduction in the current (from 160 to 10 mA) and voltage (from 140 to 80 kVp) resulted in an approximately 100-fold effective dose reduction. The result of this study was consistent with the earlier study conducted by Kamel et al.
(2002) [23], which demonstrated that the variation in tube current did not lead to a substantial difference in the quantification of 18F-FDG uptake [20] [23] ifications, including tube current and voltage reduction, spectra filtration, reconstruction algorithms and post-process method. The study used a computer-assisted tomography simulator (CatSim) to simulate CT dose exposure with different parameters and imaging conditions, including tube mA and kVp, radiation dose, beam hardening, filtration, and scattering. The study simulated CT tube current at 0.5, 1, 5, 10, 20, 25, 50 and 100 mA; CT voltage at 80, 100, 120 and 160 kVp; and copper filtration at a thickness of 0, 0.25, 0.5 and 1.0 mm. Then, each combination of these three parameters was post-processed at different sinogram smoothing levels: no sinogram smoothing, 20 boxcar smoothing, and 20 3 × 3 boxcar plus adaptive trimmed mean (ATM) filtering. 1080 sets of CTAC data were simulated with different noise levels to assess PET bias. The study concluded that within: 5% of PET bias, increased filtration substantially narrowed the spectrum, resulting in increased mean energy at all kVp settings. In fact, with a tube current of 13 mA, the radiation exposure at the narrowest energy spectra of 140 kVp with 1 mm Cu filtration showed a 66% dose reduction compared to that of unfiltered 80 kVp spectra, reducing the dose from 0.42 mGy to 0.14 mGy. Also, by applying 20 3 × 3 boxcars and ATM filtering, the current could be reduced from 13 mA to less than 4 mA at 140 kVp, while the CT data still exhibited a low noise level and the PET image bias was still below 5%.
Very recently, based on the studies conducted by Fahey et al. (2007) [20], Brady and Shulkin (2015 [25] conducted a study on ultra-low-dose CTAC for pediatric PET/CT scans. The study evaluated an aggressive CT dose reduction method while applying 100% adaptive statistical iterative reconstruction (ASIR). They used an anthropomorphic phantom to evaluate the change in CT noise and Hounsfield units value at increasingly lower CT tube currents. The tube current modulation software controlled the tube current. For example (AutomA. and SmartmA) and gradually decreased to 10% of baseline computed tomography (CT) attenuation correction (CTAC) images was reconstructed with 100% ASIR Open Journal of Medical Imaging to reduce CT noise. Then the PET data were reconstructed using CTAC through an iterative algorithm.
The result indicated that, with A 90% reduction from the baseline CTDivol, there was no statistically significant change to the The use of standardized uptake values (SUV), background uniformity or spatial resolution of the PET images. Based on the basic phantom study, Brady and Shukl (2015) [25] imple-  [25] suggested that, if diagnostic CT and detailed anatomical localization are not required, low-dose/ultra-low-dose CT is feasible for CTAC. The CTAC dose can be reduced from a range of 3.2 -8.3 mSv to 0.9 -3.2 mSv. This is especially beneficial for radiation-sensitive patients such as children, young adults and pregnant women [2] [27] [28] [29]. Another study conducted evaluated a new CT dose reduction method for PET/CT imaging, applying the "selective CT" technique for localization in follow-up studies of pediatric patients with Langerhans cell histiocytosis. The hypothesis was that the total dose from the combination of the low-dose CTAC and the regional localization CT was expected to be considerably lower than that of a full-range CT scan for both CTAC and localization purposes [30] [31] [32].
A group of 34 patients was imaged with the "selective CT" protocol, consisting of a low-dose CTAC scan and a regional localization CT scan. The low-dose CTAC was performed at the low-dose parameters (80 kVp; 10 mA) described by Fahey et al. (2007) [20], with an average effective dose of 1.65 mSv, moderately higher than that of the ultra-low dose CT protocol described by Brady et al.
(2015) [25]. Then, a selective localization CT scan of a higher dose (100 kVp; 25 mA) was immediately repeated at the bed position of pathologic or equivocal PET findings. Another group of 17 patients was imaged using full-range CT (100 kVp; 25 mA) for both CTAC and whole-body localization.
With the selective CT technique, the average effective dose from low-dose  [18] demonstrated that there was no significant difference between low-dose CT and full-dose enhanced diagnostic CT used in PET imaging for initial staging of lymphomas. As the above studies indicate, the use of additional full-dose diagnostic CT in PET studies should be carefully justified and reviewed by a nuclear medicine professional, as evidence is accumulating that the benefit of performing a full-dose diagnostic CT may not outweigh the radiation risks.

Conclusion
As the International Committee on Radiological Protection (ICRP, 2007) stated, firstly, a referral for a PET/CT should be justified in a patient-specific manner to avoid unnecessary exposure; secondly, the radiation exposure from a PET/CT Open Journal of Medical Imaging study should be just sufficient to generate reasonably achievable high-quality diagnostic information, while applying the As Low As Reasonably Achievable (ALARA) principle as best as possible. The three CT dose reduction strategies discussed above have a high potential for reducing patient dose while maintaining an adequate CT image quality used for CTAC, localization and diagnostic purposes, achieving better image optimization. With ongoing development in CT hardware, software and modified protocols, we can forecast a further reduction in dose from the component, and hence a further improvement in PET/CT imaging optimization.