Next Article in Journal
Predictive Factors of Local Recurrence after Colorectal Cancer Liver Metastases Thermal Ablation
Previous Article in Journal
Autokeras Approach: A Robust Automated Deep Learning Network for Diagnosis Disease Cases in Medical Images
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
J. Imaging
Volume 9
Issue 3
10.3390/jimaging9030065
jimaging-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Communication

Comparison of Image Quality and Quantification Parameters between Q.Clear and OSEM Reconstruction Methods on FDG-PET/CT Images in Patients with Metastatic Breast Cancer

by
Mohammad Naghavi-Behzad
1,2,3,*,
Marianne Vogsen
1,2,3,4,
Oke Gerke
1,2,
Sara Elisabeth Dahlsgaard-Wallenius
1,2,
Henriette Juel Nissen
1,2,
Nick Møldrup Jakobsen
1,2,
Poul-Erik Braad
1,5,
Mie Holm Vilstrup
1,2,
Paul Deak
6,
Malene Grubbe Hildebrandt
1,2,3,7 and
Thomas Lund Andersen
1,8
1
Department of Clinical Research, University of Southern Denmark, 5000 Odense, Denmark
2
Department of Nuclear Medicine, Odense University Hospital, 5000 Odense, Denmark
3
Centre for Personalized Response Monitoring in Oncology, Odense University Hospital, 5000 Odense, Denmark
4
Department of Oncology, Odense University Hospital, 5000 Odense, Denmark
5
Department at Clinical Engineering, Region of Southern Denmark, 6200 Aabenraa, Denmark
6
Healthcare Science Technology, GE Healthcare, Chicago, IL 06828, USA
7
Centre for Innovative Medical Technology, Odense University Hospital, 5000 Odense, Denmark
8
Department of Clinical Physiology and Nuclear Medicine, Rigshospitalet, 2100 Copenhagen, Denmark
*
Author to whom correspondence should be addressed.
J. Imaging 2023, 9(3), 65; https://doi.org/10.3390/jimaging9030065
Submission received: 23 January 2023 / Revised: 1 March 2023 / Accepted: 7 March 2023 / Published: 9 March 2023
Browse Figures
Review Reports Versions Notes

Abstract

:
We compared the image quality and quantification parameters through bayesian penalized likelihood reconstruction algorithm (Q.Clear) and ordered subset expectation maximization (OSEM) algorithm for 2-[18F]FDG-PET/CT scans performed for response monitoring in patients with metastatic breast cancer in prospective setting. We included 37 metastatic breast cancer patients diagnosed and monitored with 2-[18F]FDG-PET/CT at Odense University Hospital (Denmark). A total of 100 scans were analyzed blinded toward Q.Clear and OSEM reconstruction algorithms regarding image quality parameters (noise, sharpness, contrast, diagnostic confidence, artefacts, and blotchy appearance) using a five-point scale. The hottest lesion was selected in scans with measurable disease, considering the same volume of interest in both reconstruction methods. SULpeak (g/mL) and SUVmax (g/mL) were compared for the same hottest lesion. There was no significant difference regarding noise, diagnostic confidence, and artefacts within reconstruction methods; Q.Clear had significantly better sharpness (p < 0.001) and contrast (p = 0.001) than the OSEM reconstruction, while the OSEM reconstruction had significantly less blotchy appearance compared with Q.Clear reconstruction (p < 0.001). Quantitative analysis on 75/100 scans indicated that Q.Clear reconstruction had significantly higher SULpeak (5.33 ± 2.8 vs. 4.85 ± 2.5, p < 0.001) and SUVmax (8.27 ± 4.8 vs. 6.90 ± 3.8, p < 0.001) compared with OSEM reconstruction. In conclusion, Q.Clear reconstruction revealed better sharpness, better contrast, higher SUVmax, and higher SULpeak, while OSEM reconstruction had less blotchy appearance.

1. Introduction

Positron emission tomography with integrated computed tomography (PET/CT) is broadly used in the initial diagnosis, staging, and therapeutic response evaluation of numerous malignant diseases [1]. There are continuous technical improvements in PET/CT scanners, leading to improved imaging quality from developed hardware specifications and reconstruction algorithms [2,3,4]. Novel PET/CT scanners based on digital silicon photomultiplier (SiPM) technology have become the new standard in PET by replacing the older generation of PET/CT systems based on analog photomultiplier tubes. This leads to a considerable improvement in image contrast and noise level [5,6], which could provide better diagnostic accuracy and overall image quality, compared with analog PET/CT scanners [3,4,7,8]. A new reconstruction algorithm employing the block sequential regularized expectation maximization (BSREM) technique under the commercial name Q.Clear has been introduced. This method allows for fully convergent iterative reconstruction, resulting in higher image contrast while suppressing noise compared with ordered subset expectation maximization (OSEM) [9]. Advanced reconstruction methods aim to improve not only the quality of imaging but also quantitative measures [7], as using the Q.Clear algorithm potentially increases the maximal standardized uptake (SUVmax) values within metastatic lesions compared with OSEM reconstruction [10].
2-deoxy-2-[18F]fluoro-D-glucose PET/CT (2-[18F]FDG-PET/CT) has increasingly been introduced in metastatic breast cancer based on excellent sensitivity (over 95%) with regard to detection of distant metastases [11]. Response evaluation using 2-[18F]FDG-PET/CT may improve clinical management and survival [12,13,14]. Quantitative PET/CT is becoming increasingly important for more objective evaluations of tumor response [7], with the PET Response Criteria in Solid Tumors (PERCIST) being suggested as feasible and valuable criteria in breast cancer [15,16,17].
Prior studies have indicated Q.Clear reconstruction to be superior in overall image quality, improved contrast recovery and noise suppression, better lesion detectability, and more accurate quantification compared with OSEM reconstruction [18,19,20]. Literature comparing the two reconstruction methods in a clinical setting is lacking and is warranted before introducing the Q.Clear algorithm as a reconstruction method for semi-quantitative PERCIST analyses in clinical practice [21]. Thus, we aimed to compare the two reconstruction algorithms (Q.Clear vs. OSEM) regarding the overall image quality and quantification parameters in a clinical setting of patients being response-evaluated during treatment for metastatic breast cancer.

2. Materials and Methods

2.1. Patients

A prospective comparative study was conducted at the Department of Nuclear Medicine at Odense University Hospital (Denmark), following the STROBE guideline [22]. Patients in the present study represent a subpopulation of a patient group analyzed in a larger prospective study of response monitoring in metastatic breast cancer (Clinical.Trials.gov: NCT03358589). Women referred to the Department of Oncology (Odense University Hospital) with advanced breast cancer between 2018 (September) and 2020 (September) and examined initially with 2-[18F]FDG-PET/CT were considered eligible. The Danish Ethics Committee approved the study protocol (S-20170019), and all subjects signed written informed consent. The research was conducted in accordance with the Declaration of Helsinki, and all the scans were performed following the guideline of European Association of Nuclear Medicine (EANM) [23].
Inclusion criteria for NCT03358589 were biopsy-verified relapsed or de novo metastatic breast cancer (biopsy verification of primary tumor and disseminated disease at baseline scan), imaging performed on PET/CT scanners with digital technology, and available clinico-histopathological data. Exclusion criteria were age less than 18 years and patients in treatment for other invasive cancers [24].
The baseline scans were performed prior to the treatment initiation; patients were monitored on the same PET/CT scanner and were scanned according to the standardization criteria suggested by PERCIST [17]. Patients were scanned with 9–12 weeks of imaging intervals according to Danish clinical guidelines. The baseline, first follow-up, and second follow-up scans were included in the present analysis. Scans were analyzed with both Q.Clear and OSEM reconstruction algorithms by the same group of experienced nuclear medicine physicians comparing overall image quality parameters, maximum standardized uptake value (SUVmax), and peak lean body mass corrected SUV (SULpeak).

2.2. PET/CT Imaging Protocol

PET/CT data were acquired on GE Discovery MI 4-ring PET/CT (GE Healthcare, Waukesha, WI, USA) scanners with a field of view of 25 cm. PET scans were performed 60 min after injection of 4 mBq/kg FDG (min 200 MBq, max 400 MBq) using a standard whole-body (head-to-thigh) acquisition protocol, with slice overlaps of 40% and acquisition time of 1.5 min per bed position. PET datasets were reconstructed using time-of-flight 3D OSEM (GE VPFX, 4 iterations, 17 subsets) with point-spread-function correction (GE SharpIR) and using Q.Clear (β = 250) in matrix sizes of 256 × 256 (pixel size 2.74 mm). Corrections for attenuation, scatter, randoms, deadtime, and normalization were performed inside the iterative loop. Attenuation correction was based on a dedicated helical CT attenuation correction scan acquired after the PET scan using a standard CT protocol with a scan field of view of 70 cm. Data were reconstructed with a standard filter into trans-axial slices with a field of view of 50 cm, matrix size of 512 × 512 (pixel size 0.98 mm), and a slice thickness of 3.75 mm [23].

2.3. Qualitative Image Analysis

A clarification session was held by a senior nuclear medicine specialist (M.H.V.), ensuring the use of the same approach by the interpreters before analyzing the image quality. Three experienced nuclear medicine physicians analyzed the image quality parameters on both Q.Clear and OSEM reconstruction algorithms. One physician evaluated each scan regarding the qualitative image analyses. The same physician performed the analyses for two reconstruction methods at the same time on two separate screens (side-by-side) regarding six image quality parameters, while they were blinded to the reconstruction methods. Using a five-point scale (1 = worst and 5 = best, Table 1), the following quality parameters were compared between the reconstruction methods: noise, sharpness, contrast, diagnostic confidence, artefacts, and blotchy appearance [25]. The parameters were evaluated on a subjective scale, while the physicians were experienced enough to be good at assessing noise of the scans, as well as separating the parameters with overlap such as noise, contrast, and sharpness. The interpreters evaluated the scans independently of one another, and they had knowledge of the clinical indication of PET/CT.

2.4. Quantitative Image Analysis

PET/CT scans were evaluated for quantitative measurements. The hottest lesion according to one-lesion PERCIST was selected in scans with measurable disease, considering the same image number, considering the same volume of interest (VOI), and using the PETVCAR automatic software (AW version 3.2, GE Healthcare, Chicago, IL, USA) in both reconstruction methods. SULpeak was defined as the highest possible mean value of a 1 cm3 spherical in the VOI positioned within the metastatic lesions. Patients were checked for serum glucose level, ensuring that it was in an acceptable range (less than 200 mg/dL) according to the PERCIST criteria [17]. SUVmax was defined as the maximum uptake in the VOI that reflects the maximum tissue concentration of FDG uptake in the tumor. Body weight and height were used for SUVmax normalization [26]. SULpeak (g/mL) and SUVmax (g/mL) were calculated and compared for the same hottest lesion. Quantification of FDG uptake was performed on scans with measurable disease at baseline and for comparable scans during follow-up (i.e., SULpeak and SUVmax were not measured for scans with complete metabolic response).

2.5. Outcome Measure and Statistical Analysis

Continuous data were presented using the median (range) and mean ± standard deviation. Frequencies and respective percentages were given for categorical variables. A t-test was used to compare the six parameters regarding the image quality and quantitative parameters (SULpeak and SUVmax) of the hottest lesion between the two algorithms. The statistical level of significance was set to 0.05. All statistical analyses were conducted with STATA/IC (version 16.1, StataCorp, College Station, USA).

3. Results

A total of 37 patients with 37 baseline scans and 63 follow-up scans (including first and second follow-up scans) were available for the analysis. A study flowchart is seen in Figure 1. The clinical and histopathological information of included patients is summarized in Table 2. More detailed information of included patients is available in a previous publication [24].
Comparing the parameters related to the quality of images, Q.Clear had significantly better sharpness (mean scores of 4.65 vs. 3.91) and contrast (mean scores of 4.23 vs. 4.10) compared with the OSEM reconstruction (p < 0.001 and p = 0.001, respectively), while there was no significant difference regarding noise, diagnostic confidence, and artefacts when comparing the two reconstruction methods. The OSEM reconstruction had less blotchy appearance (4.57 vs. 4.34) compared with Q.Clear reconstruction (p < 0.001). Scores related to imaging quality parameters are summarized in Table 3. An example of 2-[18F]FDG-PET/CT, comparing the sharpness and contrast using OSEM and Q.Clear reconstructions, respectively, is shown in Figure 2.
A total of 31/37 (84%) patients had measurable disease at baseline (Figure 1), for whom quantitative analysis was performed on the hottest lesion according to the PERCIST criteria. At follow-up scans, quantitative analyses were performed in 44/63 (70%) scans being comparable according to the PERCIST criteria. Q.Clear reconstruction had significantly higher SULpeak (5.33 ± 2.8 vs. 4.85 ± 2.5, p < 0.001) and SUVmax (8.27 ± 4.8 vs. 6.90 ± 3.8, p < 0.001) compared with the OSEM reconstruction (Table 4). When comparing the two reconstruction methods for change in SULpeak and SUVmax between the two following scans, there was no significant difference in the median SULpeak changes, while the median SUVmax changes were significantly higher for Q.Clear reconstruction.

4. Discussion

A prospective comparison of OSEM and Q.Clear reconstruction algorithms was conducted on image quality parameters and quantitative analysis on 2-[18F]FDG-PET/CT scans of patients with metastatic breast cancer. The Q.Clear reconstruction showed significantly improved sharpness and contrast compared with an OSEM reconstruction, while OSEM reconstruction had a less “blotchy appearance” compared with the Q.Clear reconstruction, which may be a consequence of reduced noise in OSEM reconstruction. Registered SULpeak and SUVmax values with Q.Clear reconstruction were significantly higher than with OSEM reconstruction. Changes in SULpeak values over the follow-up period stayed independent of the reconstruction method, while the changes related to SUVmax were significantly higher in Q.Clear reconstruction.
According to our study, images performed with Q.Clear reconstruction had better contrast and sharpness than images from OSEM reconstruction, which is in line with the results of other studies indicating a superiority for the reconstruction methods based on the BSREM technique over the OSEM reconstruction [27,28]. A more “blotchy appearance” on Q.Clear images could be a consequence of a missing Gaussian filter, while using a 4 mm Gaussian filter regularly in OSEM reconstruction smoothed the blotchy appearance and resulted in a better image quality [28]. We found no difference in the diagnostic confidence between the two methods, indicating that any of the reconstruction methods may be preferred for clinical application.
Similar studies have also reported better overall image quality by Q.Clear compared with OSEM in [68Ga]Ga-DOTANOC PET/CT scans [29], 18F-fluciclovine PET/CT scans [20], and [68Ga]Ga-PSMA PET/CT scans [30]. This indicates that the improved image quality provided by Q.Clear is not exclusively dedicated to 2-[18F]FDG-PET/CT scans. Furthermore, a few studies have reported that Q.Clear had better image quality than OSEM in PET/MR scans [2,18,31]. The reason could be the same as PET/CT scans, presuming the inability of OSEM reconstruction to achieve full convergence due to increased noise and the iteration times. However, more studies including phantom data are needed to ensure the preferred β value of reconstructions ensuring the optimal image quality [20,29,30]. The reason for the better image quality with the Q.Clear algorithm could be that Q.Clear reduces the noise by resembling an adaptive filter with adjustable filter width and improving the contrast by increasing quantification, effectively creating the effect of better image quality. This is in line with our results, indicating improved contrast on Q.Clear reconstructions, which allows reaching full convergence without the excessive typical noise of OSEM [29]. This effectively limits the number of iterations of the OSEM algorithm to avoid excessive noise within the image, resulting in a lack of convergence and decreasing image contrast [2]. On the other hand, Q.Clear can achieve full convergence, resulting in higher resolution and more precise quantitative measurements due to the noise regularization [19,32].
There was a significant difference between the absolute values of SUVmax and SULpeak within the OSEM and Q.Clear reconstruction algorithms, which is in line with the results of previous studies indicating that Q.Clear allows a significant increase in quantitative parameters and better reflects the true uptake [9,20]. Therefore, Q.Clear has the potential to provide an improved quantification accuracy, which could be beneficial for research purposes, as well as using quantification-based analyses (e.g., PERCIST analysis) in clinical practice.
Lundeberg et al. compared the two reconstruction modalities on lung cancer patients in a clinical setting and reached the similar results indicating that the Q.Clear reconstruction provides a higher SUVmax for suspected lymph node metastases compared to the OSEM reconstruction. However, a higher values of SUVmax did not lead to an improvement in detection of metastatic lesions [33]. In addition, our results showed that SULpeak changes over the follow-up period were not associated with the reconstruction algorithm, as opposed to SUVmax, for which changes were higher for the Q.Clear than OSEM reconstruction, indicating SULpeak to be robust for PERCIST analysis across reconstruction algorithms. This complies with the proposed upward bias for SUVmax using a single pixel, as the size of a single voxel may differ considerably among PET/CT scans resulting in different noise levels in the metric followed by various filtering, while PERCIST suggests using a larger region of interest by SULpeak [17]. Q.Clear also showed a clinically relevant recovery coefficient for different sphere diameters (10–37 mm), which is beneficial for lesion detection and is more compatible with the quantification of lesions according to PERCIST criteria [34].
It has also been reported that PERCIST has high applicability [16], has a higher level of overall interrater agreement and reliability compared with a qualitative assessment [35,36], and is superior in the detection of new lesions or unequivocal progression in nontarget lesions [37]. Therefore, clinical implementation of the PERCIST assessment may improve the prognostic stratification [15,37,38] and provide a standardized approach independent of interpreters and reconstruction methods. The difference in the absolute value of SULpeak may result in dissimilar response categories according to the PERCIST criteria and eventually affect the patients’ treatment plan. Hence, a clinical indication of the Q.Clear reconstruction algorithm may lead to a more precise treatment monitoring by PET/CT through an improved quantification accuracy [39]. However, the clinical indication of the Q.Clear algorithm in the treatment evaluation of lymphoma patients is reported to be uncertain, which could be explained by the incompatibility with the current guidelines [10]. Therefore, compatibility of Q.Clear with existing guidelines is required before introducing the algorithm into clinical practice to have a significant impact on patient management.
A strength of this study was the inclusion of clinical follow-up of metastatic patients representing daily clinical practice. Furthermore, the PERCIST criteria were used for response monitoring and quantification of FDG uptake with strict compliance with standardization criteria such as the PERCIST guidelines [17]. Accordingly, patients were followed on the same type of scanner, eliminating scanner variation effects, while comparing quantitative measures between follow-up scans. The interpreters for imaging quality parameters were blinded to the reconstruction methods, and the same physician evaluated both reconstruction methods at the same time (side-by-side). As a limitation, we only performed a visual comparison in terms of overall image quality, which could be biased by the personal preferences of the interpreter physician, as only one operator analyzed each scan. Furthermore, the number and size of the detected lesions for the two reconstruction methods were not measured, which could have strengthened the results of quantitative analyses.
Future multicenter studies on a larger number of scans evaluated by several experienced nuclear medicine physicians could verify the results of the current study. The lesion-based accuracy for the two reconstruction methods and the potential effect on response categories remain unanswered. Furthermore, the determination of the optimal penalization factor (β-value) for clinical use and phantom measurements related to Q.Clear should be considered in future studies.

5. Conclusions

Q.Clear reconstruction showed a significantly better sharpness and contrast compared with OSEM reconstruction, while the blotchy appearance was less evident in OSEM reconstruction. There was no difference in diagnostic confidence between the two reconstruction algorithms, making them equally perfect for daily clinical practice. The Q.Clear algorithm had higher quantitative measures with higher SUVmax and SULpeak than OSEM reconstruction. SULpeak changes at follow-up scans stayed independent of the reconstruction methods, indicating SULpeak to be robust for PERCIST analyses regardless of the reconstruction method.

Author Contributions

M.N.-B., M.V., P.-E.B., P.D. and M.G.H. contributed to the study conceptualization and design; M.N.-B., M.V., S.E.D.-W., H.J.N., N.M.J., P.-E.B., M.H.V., M.G.H. and T.L.A. contributed to the data collection, data analysis was performed by M.N.-B. and O.G., the first draft of the manuscript was written by M.N.-B. and all authors commented on previous versions of the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by GE Healthcare Company, Chicago, IL, USA (10224776), the University of Southern Denmark (19/1435), and Odense University Hospital, Denmark (10212315).

Institutional Review Board Statement

The study protocol was approved by the Danish Ethics Committee (S-20170019) in the Region of Southern Denmark.

Informed Consent Statement

Written informed consent was obtained from the patients to publish this paper.

Data Availability Statement

The datasets analyzed during the current study are available from the corresponding author on reasonable request.

Acknowledgments

The authors would like to thank Margit Vølund (Department of Nuclear Medicine, Odense University Hospital, Denmark) for help with anonymizing the scans.

Conflicts of Interest

P.D. is affiliated with the GE Healthcare Company (USA). The other authors declare no conflict of interest.

References

  1. Ben-Haim, S.; Ell, P. 18F-FDG PET and PET/CT in the evaluation of cancer treatment response. J. Nucl. Med. 2009, 50, 88–99. [Google Scholar] [CrossRef] [Green Version]
  2. Tian, D.; Yang, H.; Li, Y.; Cui, B.; Lu, J. The effect of Q.Clear reconstruction on quantification and spatial resolution of 18F-FDG PET in simultaneous PET/MR. EJNMMI Phys. 2022, 9, 1. [Google Scholar] [CrossRef]
  3. Fuentes-Ocampo, F.; López-Mora, D.A.; Flotats, A.; Camacho, V.; Sizova, M.; Abouzian, S.; Duch, J.; Fernández, A.; Estorch, M.; Carrió, I. Digital versus analog PET/CT in patients with known or suspected liver metastases. Nucl. Med. Commun. 2021, 42, 421–428. [Google Scholar] [CrossRef] [PubMed]
  4. López-Mora, D.A.; Flotats, A.; Fuentes-Ocampo, F.; Camacho, V.; Fernández, A.; Ruiz, A.; Duch, J.; Sizova, M.; Domènech, A.; Estorch, M.; et al. Comparison of image quality and lesion detection between digital and analog PET/CT. Eur. J. Nucl. Med. Mol. Imaging 2019, 46, 1383–1390. [Google Scholar] [CrossRef] [PubMed]
  5. Zhang, J.; Maniawski, P.; Knopp, M.V. Performance evaluation of the next generation solid-state digital photon counting PET/CT system. EJNMMI Res. 2018, 8, 97. [Google Scholar] [CrossRef]
  6. Salvadori, J.; Imbert, L.; Perrin, M.; Karcher, G.; Lamiral, Z.; Marie, P.Y.; Verger, A. Head-to-head comparison of image quality between brain (18)F-FDG images recorded with a fully digital versus a last-generation analog PET camera. EJNMMI Res. 2019, 9, 61. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  7. Nguyen, N.C.; Vercher-Conejero, J.L.; Sattar, A.; Miller, M.A.; Maniawski, P.J.; Jordan, D.W.; Muzic, R.F., Jr.; Su, K.H.; O’Donnell, J.K.; Faulhaber, P.F. Image Quality and Diagnostic Performance of a Digital PET Prototype in Patients with Oncologic Diseases: Initial Experience and Comparison with Analog PET. J. Nucl. Med. 2015, 56, 1378–1385. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  8. Fuentes-Ocampo, F.; López-Mora, D.A.; Flotats, A.; Paillahueque, G.; Camacho, V.; Duch, J.; Fernández, A.; Domènech, A.; Estorch, M.; Carrió, I. Digital vs. analog PET/CT: Intra-subject comparison of the SUVmax in target lesions and reference regions. Eur. J. Nucl. Med. Mol. Imaging 2019, 46, 1745–1750. [Google Scholar] [CrossRef]
  9. Texte, E.; Gouel, P.; Thureau, S.; Lequesne, J.; Barres, B.; Edet-Sanson, A.; Decazes, P.; Vera, P.; Hapdey, S. Impact of the Bayesian penalized likelihood algorithm (Q.Clear®) in comparison with the OSEM reconstruction on low contrast PET hypoxic images. EJNMMI Phys. 2020, 7, 28. [Google Scholar] [CrossRef]
  10. Wyrzykowski, M.; Siminiak, N.; Kaźmierczak, M.; Ruchała, M.; Czepczyński, R. Impact of the Q.Clear reconstruction algorithm on the interpretation of PET/CT images in patients with lymphoma. EJNMMI Res. 2020, 10, 99. [Google Scholar] [CrossRef]
  11. Sun, Z.; Yi, Y.L.; Liu, Y.; Xiong, J.P.; He, C.Z. Comparison of whole-body PET/PET-CT and conventional imaging procedures for distant metastasis staging in patients with breast cancer: A meta-analysis. Eur. J. Gynaecol. Oncol. 2015, 36, 672–676. [Google Scholar] [CrossRef]
  12. Hildebrandt, M.G.; Naghavi-Behzad, M.; Vogsen, M. A role of FDG-PET/CT for response evaluation in metastatic breast cancer? Semin. Nucl. Med. 2022, 52, 520–530. [Google Scholar] [CrossRef] [PubMed]
  13. Riedl, C.C.; Pinker, K.; Ulaner, G.A.; Ong, L.T.; Baltzer, P.; Jochelson, M.S.; McArthur, H.L.; Gönen, M.; Dickler, M.; Weber, W.A. Comparison of FDG-PET/CT and contrast-enhanced CT for monitoring therapy response in patients with metastatic breast cancer. Eur. J. Nucl. Med. Mol. Imaging 2017, 44, 1428–1437. [Google Scholar] [CrossRef] [PubMed]
  14. Naghavi-Behzad, M.; Vogsen, M.; Vester, R.M.; Olsen, M.M.B.; Oltmann, H.; Braad, P.E.; Asmussen, J.T.; Gerke, O.; Vach, W.; Kidholm, K.; et al. Response monitoring in metastatic breast cancer: A comparison of survival times between FDG-PET/CT and CE-CT. Br. J. Cancer 2022, 126, 1271–1279. [Google Scholar] [CrossRef]
  15. Depardon, E.; Kanoun, S.; Humbert, O.; Bertaut, A.; Riedinger, J.M.; Tal, I.; Vrigneaud, J.M.; Lasserre, M.; Toubeau, M.; Berriolo-Riedinger, A.; et al. FDG PET/CT for prognostic stratification of patients with metastatic breast cancer treated with first line systemic therapy: Comparison of EORTC criteria and PERCIST. PLoS ONE 2018, 13, e0199529. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  16. Vogsen, M.; Bülow, J.L.; Ljungstrøm, L.; Oltmann, H.R.; Alamdari, T.A.; Naghavi-Behzad, M.; Braad, P.E.; Gerke, O.; Hildebrandt, M.G. FDG-PET/CT for Response Monitoring in Metastatic Breast Cancer: The Feasibility and Benefits of Applying PERCIST. Diagnostics 2021, 11, 723. [Google Scholar] [CrossRef] [PubMed]
  17. Hyun, O.J.; Lodge, M.A.; Wahl, R.L. Practical PERCIST: A Simplified Guide to PET Response Criteria in Solid Tumors 1.0. Radiology 2016, 280, 576–584. [Google Scholar] [CrossRef] [Green Version]
  18. Ribeiro, D.; Hallett, W.; Tavares, A.A.S. Performance evaluation of the Q.Clear reconstruction framework versus conventional reconstruction algorithms for quantitative brain PET-MR studies. EJNMMI Phys. 2021, 8, 41. [Google Scholar] [CrossRef]
  19. Te Riet, J.; Rijnsdorp, S.; Roef, M.J.; Arends, A.J. Evaluation of a Bayesian penalized likelihood reconstruction algorithm for low-count clinical (18)F-FDG PET/CT. EJNMMI Phys. 2019, 6, 32. [Google Scholar] [CrossRef] [Green Version]
  20. Teoh, E.J.; McGowan, D.R.; Schuster, D.M.; Tsakok, M.T.; Gleeson, F.V.; Bradley, K.M. Bayesian penalised likelihood reconstruction (Q.Clear) of (18)F-fluciclovine PET for imaging of recurrent prostate cancer: Semi-quantitative and clinical evaluation. Br. J. Radiol. 2018, 91, 20170727. [Google Scholar] [CrossRef]
  21. Barrington, S.F.; Sulkin, T.; Forbes, A.; Johnson, P.W.M. All that glitters is not gold—New reconstruction methods using Deauville criteria for patient reporting. Eur. J. Nucl. Med. Mol. Imaging 2018, 45, 316–317. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  22. von Elm, E.; Altman, D.G.; Egger, M.; Pocock, S.J.; Gøtzsche, P.C.; Vandenbroucke, J.P. The Strengthening the Reporting of Observational studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Lancet 2007, 370, 1453–1457. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  23. Boellaard, R.; Delgado-Bolton, R.; Oyen, W.J.; Giammarile, F.; Tatsch, K.; Eschner, W.; Verzijlbergen, F.J.; Barrington, S.F.; Pike, L.C.; Weber, W.A.; et al. FDG PET/CT: EANM procedure guidelines for tumour imaging: Version 2.0. Eur. J. Nucl. Med. Mol. Imaging 2015, 42, 328–354. [Google Scholar] [CrossRef] [PubMed]
  24. Vogsen, M.; Harbo, F.; Jakobsen, N.M.; Nissen, H.J.; Dahlsgaard-Wallenius, S.E.; Gerke, O.; Jensen, J.D.; Asmussen, J.T.; Jylling, A.M.B.; Braad, P.E.; et al. Response monitoring in metastatic breast cancer—A prospective study comparing (18)F-FDG PET/CT with conventional CT. J. Nucl. Med. 2022, 64, 263358. [Google Scholar] [CrossRef]
  25. Arapakis, I.; Efstathopoulos, E.; Tsitsia, V.; Kordolaimi, S.; Economopoulos, N.; Argentos, S.; Ploussi, A.; Alexopoulou, E. Using “iDose4” iterative reconstruction algorithm in adults’ chest-abdomen-pelvis CT examinations: Effect on image quality in relation to patient radiation exposure. Br. J. Radiol. 2014, 87, 20130613. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  26. Baun, C.; Falch, K.; Gerke, O.; Hansen, J.; Nguyen, T.; Alavi, A.; Høilund-Carlsen, P.F.; Hildebrandt, M.G. Quantification of FDG-PET/CT with delayed imaging in patients with newly diagnosed recurrent breast cancer. BMC Med. Imaging 2018, 18, 11. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  27. Otani, T.; Hosono, M.; Kanagaki, M.; Onishi, Y.; Matsubara, N.; Kawabata, K.; Kimura, H. Evaluation and Optimization of a New PET Reconstruction Algorithm, Bayesian Penalized Likelihood Reconstruction, for Lung Cancer Assessment According to Lesion Size. AJR Am. J. Roentgenol. 2019, 213, W50–W56. [Google Scholar] [CrossRef] [PubMed]
  28. Yoshii, T.; Miwa, K.; Yamaguchi, M.; Shimada, K.; Wagatsuma, K.; Yamao, T.; Kamitaka, Y.; Hiratsuka, S.; Kobayashi, R.; Ichikawa, H.; et al. Optimization of a Bayesian penalized likelihood algorithm (Q.Clear) for (18)F-NaF bone PET/CT images acquired over shorter durations using a custom-designed phantom. EJNMMI Phys. 2020, 7, 56. [Google Scholar] [CrossRef] [PubMed]
  29. Zanoni, L.; Argalia, G.; Fortunati, E.; Malizia, C.; Allegri, V.; Calabrò, D.; Civollani, S.; Campana, D.; Fanti, S.; Ambrosini, V. Can Q.Clear reconstruction be used to improve [68 Ga]Ga-DOTANOC PET/CT image quality in overweight NEN patients? Eur. J. Nucl. Med. Mol. Imaging 2022, 49, 1607–1612. [Google Scholar] [CrossRef] [PubMed]
  30. Rijnsdorp, S.; Roef, M.J.; Arends, A.J. Impact of the Noise Penalty Factor on Quantification in Bayesian Penalized Likelihood (Q.Clear) Reconstructions of (68)Ga-PSMA PET/CT Scans. Diagnostics 2021, 11, 847. [Google Scholar] [CrossRef] [PubMed]
  31. Ter Voert, E.; Muehlematter, U.J.; Delso, G.; Pizzuto, D.A.; Müller, J.; Nagel, H.W.; Burger, I.A. Quantitative performance and optimal regularization parameter in block sequential regularized expectation maximization reconstructions in clinical (68)Ga-PSMA PET/MR. EJNMMI Res. 2018, 8, 70. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  32. Chilcott, A.K.; Bradley, K.M.; McGowan, D.R. Effect of a Bayesian Penalized Likelihood PET Reconstruction Compared With Ordered Subset Expectation Maximization on Clinical Image Quality Over a Wide Range of Patient Weights. AJR Am. J. Roentgenol. 2018, 210, 153–157. [Google Scholar] [CrossRef]
  33. Economou Lundeberg, J.; Oddstig, J.; Bitzén, U.; Trägårdh, E. Comparison between silicon photomultiplier-based and conventional PET/CT in patients with suspected lung cancer-a pilot study. EJNMMI Res. 2019, 9, 35. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  34. Devriese, J.; Beels, L.; Maes, A.; Van de Wiele, C.; Pottel, H. Impact of PET reconstruction protocols on quantification of lesions that fulfil the PERCIST lesion inclusion criteria. EJNMMI Phys. 2018, 5, 35. [Google Scholar] [CrossRef] [PubMed]
  35. Sørensen, J.S.; Vilstrup, M.H.; Holm, J.; Vogsen, M.; Bülow, J.L.; Ljungstrøm, L.; Braad, P.E.; Gerke, O.; Hildebrandt, M.G. Interrater Agreement and Reliability of PERCIST and Visual Assessment When Using 18F-FDG-PET/CT for Response Monitoring of Metastatic Breast Cancer. Diagnostics 2020, 10, 1001. [Google Scholar] [CrossRef]
  36. Fledelius, J.; Khalil, A.; Hjorthaug, K.; Frøkiær, J. Inter-observer agreement improves with PERCIST 1.0 as opposed to qualitative evaluation in non-small cell lung cancer patients evaluated with F-18-FDG PET/CT early in the course of chemo-radiotherapy. EJNMMI Res. 2016, 6, 71. [Google Scholar] [CrossRef] [Green Version]
  37. Hildebrandt, M.G.; Lauridsen, J.F.; Vogsen, M.; Holm, J.; Vilstrup, M.H.; Braad, P.E.; Gerke, O.; Thomassen, M.; Ewertz, M.; Høilund-Carlsen, P.F. FDG-PET/CT for Response Monitoring in Metastatic Breast Cancer: Today, Tomorrow, and Beyond. Cancers 2019, 11, 1190. [Google Scholar] [CrossRef] [Green Version]
  38. Min, S.J.; Jang, H.J.; Kim, J.H. Comparison of the RECIST and PERCIST criteria in solid tumors: A pooled analysis and review. Oncotarget 2016, 7, 27848–27854. [Google Scholar] [CrossRef]
  39. Wu, Z.; Guo, B.; Huang, B.; Hao, X.; Wu, P.; Zhao, B.; Qin, Z.; Xie, J.; Li, S. Phantom and clinical assessment of small pulmonary nodules using Q.Clear reconstruction on a silicon-photomultiplier-based time-of-flight PET/CT system. Sci. Rep. 2021, 11, 10328. [Google Scholar] [CrossRef]
Jimaging 09 00065 g001 550
Figure 1. Study flowchart for image quality and quantitative analyses (OSEM: ordered subset expectation maximization; Q.Clear: Refers to the reconstruction algorithm using block sequential regularized expectation maximization; PERCIST: PET Response Criteria in Solid Tumors).
Figure 1. Study flowchart for image quality and quantitative analyses (OSEM: ordered subset expectation maximization; Q.Clear: Refers to the reconstruction algorithm using block sequential regularized expectation maximization; PERCIST: PET Response Criteria in Solid Tumors).
Jimaging 09 00065 g001
Jimaging 09 00065 g002 550
Figure 2. 2-[18F]FDG-PET/CT scan for a patient with metastatic breast cancer, illustrating the sharpness and contrast via OSEM and Q.Clear reconstructions ((A,B) OSEM reconstruction vs. (C,D) Q.Clear reconstruction).
Figure 2. 2-[18F]FDG-PET/CT scan for a patient with metastatic breast cancer, illustrating the sharpness and contrast via OSEM and Q.Clear reconstructions ((A,B) OSEM reconstruction vs. (C,D) Q.Clear reconstruction).
Jimaging 09 00065 g002
Table
Table 1. Grading scale for subjective image quality evaluation.
Table 1. Grading scale for subjective image quality evaluation.
Parameters54321
NoiseMinimal or no noiseNo significant noiseNoisy diagnosticSignificant noise (affects diagnosis)High-level noise (nondiagnostic)
SharpnessExcellent sharpnessGood sharpnessModerate sharpnessPoor sharpness (bad visibility)Zero visibility
(nondiagnostic)
ContrastExcellent contrastVery good contrastGood contrastPoor contrast (unsatisfactory visualization)Image similar to use of no contrast (nondiagnostic)
Diagnostic confidence Completely confidenceHigh confidenceGood confidencePoor confidenceNo diagnostic confidence (unacceptable)
Artefacts No artefactsInsignificant artefactsMinor artefactsMajor artefacts (diagnosis still possible)Artefacts affecting diagnostic information
Blotchy appearance AbsentMildModerateSignificant (diagnosis still possible)Intense (affecting diagnosis)
Table
Table 2. Clinicopathological characteristics of included patients with metastatic breast cancer.
Table 2. Clinicopathological characteristics of included patients with metastatic breast cancer.
CharacteristicsResults *
Age (years)71.9 (45.9–91.1)
Primary cancer treatmentPostoperative adjuvant treatment24 (64.7)
Adjuvant and neoadjuvant treatments3 (8.1)
No treatment/unknown10 (27.0)
History of radiotherapy24 (64.7)
Primary disseminated cancer12 (32.4)
HistopathologyAdenocarcinoma28 (75.7)
Invasive ductal carcinoma5 (13.5)
Invasive lobular carcinoma4 (10.8)
Positive estrogen receptor32 (86.5)
Negative Herceptin receptor34 (91.9)
Origin of biopsy **Bone13 (35.1)
Liver7 (18.9)
Lung1 (2.7)
Lymph nodes6 (16.2)
Breast10 (27.0)
First-line treatmentEndocrine therapy5 (13.5)
Endocrine therapy + CDK4/6 inhibitor24 (64.9)
Chemotherapy4 (10.8)
Others4 (10.8)
* Data are shown as median (range) and frequency (%). ** In some cases of de novo cancer, we used information from the initial breast biopsy in case of absence of biopsy from metastases.
Table
Table 3. Scores of image quality parameters within the OSEM and Q.Clear reconstruction methods.
Table 3. Scores of image quality parameters within the OSEM and Q.Clear reconstruction methods.
CharacteristicsOSEM *Q.Clear *Mean Difference (95% CI)p-Value
Noise4.41 ± 0.554.42 ± 0.540.01 (−0.16–0.14)0.88
Sharpness3.91 ± 0.494.65 ± 0.59−0.74 (−0.83–−0.65)<0.001
Contrast4.1 ± 0.664.23 ± 0.74−0.13 (−0.22–−0.04)0.001
Diagnostic confidence4.52 ± 0.704.52 ± 0.690 (−0.28–0.28)0.99
Artifacts4.37 ± 0.684.38 ± 0.66−0.01 (−0.3–0.01)0.32
Blotchy appearance4.57 ± 0.574.34 ± 0.590.23 (0.12–0.34)<0.001
OSEM: ordered subset expectation maximization; CI: confidence interval; Q.Clear: refers to the reconstruction algorithm using block sequential regularized expectation maximization (BSREM). * Image quality scores (mean ± standard deviation) are reported using a five-scale questionnaire (1 = worst and 5 = best).
Table
Table 4. Quantitative analysis within the hottest lesion via OSEM and Q.Clear reconstruction methods.
Table 4. Quantitative analysis within the hottest lesion via OSEM and Q.Clear reconstruction methods.
CharacteristicsOSEM *Q.Clear *Mean Difference (95% CI)p-Value
SULpeakBaseline scans5.82 (1.4–12.12)6.84 (1.61–12.95)−0.6 (−0.82–−0.39)<0.001
Follow-up scans3.01 (1.65–11.01)3.47 (1.79–12.82)−0.39 (−0.52–−0.26)0.001
All scans4.3 (1.4–12.12)4.63 (1.61–12.95)−0.47 (−0.59–−0.36)<0.001
Change to 1st follow-up1.94 (0.07–5.71)1.95 (0–5.86)0.04 (−0.24–0.33)0.75
Change to 2nd follow-up0.53 (0.02–4)0.8 (0.05–4.16)0.12 (−0.23–0.26)0.1
SUVmaxBaseline scans8.12 (2.0–18.42)9.46 (2.37–24.86)−1.49 (−1.97–1.01)<0.001
Follow-up scans4.61 (2.22–18.42)5.48 (2.47–24.86)−1.25 (−1.73–−0.77)0.005
All scans6.16 (2.0–18.42)7.15 (2.37–24.86)−1.35 (−1.69–−1.01)<0.001
Change to 1st follow-up2.34 (0–8.3)2.39 (0–9.5)0.45 (0.10–0.79)0.01
Change to 2nd follow-up1.1 (0.05–4.92)1.5 (0.26–11.2)0.89 (0.20–1.58)0.04
OSEM: ordered subset expectation maximization; CI: confidence interval; Q.Clear: refers to the reconstruction algorithm using block sequential regularized expectation maximization (BSREM); SUVmax: maximum standardized uptake value; SULpeak: peak lean body mass corrected SUV. * Data are shown as the median (range).
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Naghavi-Behzad, M.; Vogsen, M.; Gerke, O.; Dahlsgaard-Wallenius, S.E.; Nissen, H.J.; Jakobsen, N.M.; Braad, P.-E.; Vilstrup, M.H.; Deak, P.; Hildebrandt, M.G.; et al. Comparison of Image Quality and Quantification Parameters between Q.Clear and OSEM Reconstruction Methods on FDG-PET/CT Images in Patients with Metastatic Breast Cancer. J. Imaging 2023, 9, 65. https://doi.org/10.3390/jimaging9030065

AMA Style

Naghavi-Behzad M, Vogsen M, Gerke O, Dahlsgaard-Wallenius SE, Nissen HJ, Jakobsen NM, Braad P-E, Vilstrup MH, Deak P, Hildebrandt MG, et al. Comparison of Image Quality and Quantification Parameters between Q.Clear and OSEM Reconstruction Methods on FDG-PET/CT Images in Patients with Metastatic Breast Cancer. Journal of Imaging. 2023; 9(3):65. https://doi.org/10.3390/jimaging9030065

Chicago/Turabian Style

Naghavi-Behzad, Mohammad, Marianne Vogsen, Oke Gerke, Sara Elisabeth Dahlsgaard-Wallenius, Henriette Juel Nissen, Nick Møldrup Jakobsen, Poul-Erik Braad, Mie Holm Vilstrup, Paul Deak, Malene Grubbe Hildebrandt, and et al. 2023. "Comparison of Image Quality and Quantification Parameters between Q.Clear and OSEM Reconstruction Methods on FDG-PET/CT Images in Patients with Metastatic Breast Cancer" Journal of Imaging 9, no. 3: 65. https://doi.org/10.3390/jimaging9030065

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop