First results on kinetic modelling and parametric imaging of dynamic 18F-FDG datasets from a long axial FOV PET scanner in oncological patients

Mingels, Clemens; Alberts, Ian; Hu, Jicun; Buesser, Dorothee; Shah, Vijay; Schepers, Robin; Caluori, Patrik; Panin, Vladimir; Conti, Maurizio; Afshar‐Oromieh, Ali; Shi, Kuangyu; Eriksson, Lars; Rominger, Axel; Cumming, Paul

doi:10.1007/s00259-021-05623-6

Cited by 74 publications

(108 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fat-and water-based soft tissue segmentations were obtained by thresholding voxels with LAC values outside 0.080-0.090 cm −1 range and 0.090-0.105 cm −1 range respectively. Furthermore, three dimensional segmentations of liver, lungs, kidneys, spleen, grey and white matter of the brain were obtained using a semi-automatic (1) rME(%) = 100 method [36,37]. Hypermetabolic tumour lesions (n = 24) were delineated by a qualified nuclear medicine physician using an isocontour tool (PMOD 4.1, threshold set to 50% of max value).…”

Section: Discussionmentioning

confidence: 99%

Quantitative evaluation of a deep learning-based framework to generate whole-body attenuation maps using LSO background radiation in long axial FOV PET scanners

Teimoorisichani

Mingels

Alberts

et al. 2022

Eur J Nucl Med Mol Imaging

Self Cite

View full text Add to dashboard Cite

Purpose Attenuation correction is a critically important step in data correction in positron emission tomography (PET) image formation. The current standard method involves conversion of Hounsfield units from a computed tomography (CT) image to construct attenuation maps (µ-maps) at 511 keV. In this work, the increased sensitivity of long axial field-of-view (LAFOV) PET scanners was exploited to develop and evaluate a deep learning (DL) and joint reconstruction-based method to generate µ-maps utilizing background radiation from lutetium-based (LSO) scintillators. Methods Data from 18 subjects were used to train convolutional neural networks to enhance initial µ-maps generated using joint activity and attenuation reconstruction algorithm (MLACF) with transmission data from LSO background radiation acquired before and after the administration of 18F-fluorodeoxyglucose (18F-FDG) (µ-mapMLACF-PRE and µ-mapMLACF-POST respectively). The deep learning-enhanced µ-maps (µ-mapDL-MLACF-PRE and µ-mapDL-MLACF-POST) were compared against MLACF-derived and CT-based maps (µ-mapCT). The performance of the method was also evaluated by assessing PET images reconstructed using each µ-map and computing volume-of-interest based standard uptake value measurements and percentage relative mean error (rME) and relative mean absolute error (rMAE) relative to CT-based method. Results No statistically significant difference was observed in rME values for µ-mapDL-MLACF-PRE and µ-mapDL-MLACF-POST both in fat-based and water-based soft tissue as well as bones, suggesting that presence of the radiopharmaceutical activity in the body had negligible effects on the resulting µ-maps. The rMAE values µ-mapDL-MLACF-POST were reduced by a factor of 3.3 in average compared to the rMAE of µ-mapMLACF-POST. Similarly, the average rMAE values of PET images reconstructed using µ-mapDL-MLACF-POST (PETDL-MLACF-POST) were 2.6 times smaller than the average rMAE values of PET images reconstructed using µ-mapMLACF-POST. The mean absolute errors in SUV values of PETDL-MLACF-POST compared to PETCT were less than 5% in healthy organs, less than 7% in brain grey matter and 4.3% for all tumours combined. Conclusion We describe a deep learning-based method to accurately generate µ-maps from PET emission data and LSO background radiation, enabling CT-free attenuation and scatter correction in LAFOV PET scanners.

show abstract

Section: Discussionmentioning

confidence: 99%

Quantitative evaluation of a deep learning-based framework to generate whole-body attenuation maps using LSO background radiation in long axial FOV PET scanners

Teimoorisichani

Mingels

Alberts

et al. 2022

Eur J Nucl Med Mol Imaging

Self Cite

View full text Add to dashboard Cite

show abstract

“…Parametric images were reconstructed using the direct Patlak method implemented in a dedicated parametric imaging software prototype (Siemens Healthineers) which employs a nested expectation maximization algorithm [24]. Parametric images were reconstructed using the PSF + TOF method with 8 iterations and 5 subsets, 30 nested loops and were smoothed using a 2 mm FWHM Gaussian lter [9]. Quantitative evaluation of images was performed by computing non-absolute and absolute relative change (%RC), the structural similarity index measure (SSIM), and peak signal-to-noise ratio (PSNR) relative to corresponding images obtained with the IDIF and t*=35 min.…”

Section: Methodsmentioning

confidence: 99%

“…With the introduction of long axial FOV (LAFOV) PET/CT scanners, IDIFs can be derived from various large vascular structures or blood pools (i.e. aorta, left ventricle), minimizing the partial volume effects [9,10]. Furthermore, the increased sensitivity of these systems enables use of short frame durations in the reconstruction of early PET frames [11][12][13], allowing a more detailed capture of the IDIF curve peaks.…”

Section: Introductionmentioning

confidence: 99%

Feasibility of using abbreviated scans protocols with population-based input functions for accurate kinetic modelling of 18F-FDG datasets from a long-axial FOV PET scanner

Eriksson

Mingels

Alberts

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Background: Accurate kinetic modelling of 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) data requires accurate knowledge of the available tracer concentration in the plasma during the scan time, known as the arterial input function (AIF). The gold standard method to derive the AIF requires collection of serial arterial blood samples but the introduction of long axial field of view (LAFOV) PET systems enables use of non-invasive image derived input functions (IDIF) from large blood pools such as the aorta without any need for bed movement. However, such protocols require a prolonged dynamic PET acquisition which is impractical in a busy clinical setting. Population-based input functions (PBIF) have previously shown potential in accurate Patlak analysis of 18F-FDG datasets and can enable the use of shortened dynamic imaging protocols. We not exploit the high sensitivity and temporal resolution of a LAFOV PET system and explore use of PBIF with abbreviated protocols in 18F-FDG total body kinetic modelling. Methods: Dynamic PET data were acquired in 24 oncological subjects for 65 minutes following the administration of 18F-FDG. IDIFs were extracted from the descending thoracic aorta and a PBIF was generated from 16 datasets. Five different scaled PBIFs (sPBIF) were generated by scaling the PBIF with AUC of IDIF curve tails using various portions of image data (35-65, 40-65, 45-65, 50-65 and 55-65 min post injection). The sPBIFs were compared with the IDIFs using the AUCs and Patlak Ki estimates in tumour lesions and cerebral grey matter. Patlak plot start time (t*) was also varied to evaluate the performance of shorter acquisitions on accuracy of Patlak Ki estimates. Patlak Ki estimates with IDIF and t*=35 min was used as reference and mean bias and precision (standard deviation of bias) were calculated to assess relative performance of different sPBIFs. Comparison of parametric images generated using IDIF and sPBIFs was also performed. Results: There was no statistically significant difference between AUCs of the IDIF and sPBIFs (Wilcoxon test: P>0.05). The sPBIF55-65 showed the best performance with 1.5% bias and %6.8 precision in tumour lesions. Using the sPBIF55-65 with Patlak model, 20 minutes of PET data (i.e. 45 to 65 min post injection) achieved <15% precision error in Ki estimates in tumour lesions compared to the estimates with the IDIF. Parametric images reconstructed using the IDIF and sPBIFs with and without an abbreviated protocol were visually comparable. Using Patlak Ki generated with an IDIF and 30 mins of PET data as reference, Patlak Ki images generated using sPBIF55-65 with 20 minutes of PET data (t*=45 min) provided excellent image quality with structural similarity index measure > 0.99 and peak signal-to-noise ratio > 55 dB. Conclusion: We demonstrate the feasibility of performing accurate 18F-FDG Patlak analysis using sPBIFs with only 20 minutes of PET data from a LAFOV PET scanner.

show abstract

“…First, the whole-body coverage allows for the noninvasive input function from the aortic arteries to be obtained, which has already been shown to be close to the arterial sampling, at least for FDG [ 35 ]. Zhang et al [ 36 ] and Sari et al [ 37 ] investigated the feasibility of extracting an input function from dynamic images of the left atrium, left ventricle, aortic artery, and carotid artery, for example. This capability can improve the dependability of IF estimation methods such as PBIF approximation and the hybrid method.…”

Section: Introductionmentioning

confidence: 99%

Comparison between a dual-time-window protocol and other simplified protocols for dynamic total-body 18F-FDG PET imaging

Wang

et al. 2022

EJNMMI Phys

View full text Add to dashboard Cite

Purpose Efforts have been made both to avoid invasive blood sampling and to shorten the scan duration for dynamic positron emission tomography (PET) imaging. A total-body scanner, such as the uEXPLORER PET/CT, can relieve these challenges through the following features: First, the whole-body coverage allows for noninvasive input function from the aortic arteries; second, with a dramatic increase in sensitivity, image quality can still be maintained at a high level even with a shorter scan duration than usual. We implemented a dual-time-window (DTW) protocol for a dynamic total-body 18F-FDG PET scan to obtain multiple kinetic parameters. The DTW protocol was then compared to several other simplified quantification methods for total-body FDG imaging that were proposed for conventional setup. Methods The research included 28 patient scans performed on an uEXPLORER PET/CT. By discarding the corresponding data in the middle of the existing full 60-min dynamic scan, the DTW protocol was simulated. Nonlinear fitting was used to estimate the missing data in the interval. The full input function was obtained from 15 subjects using a hybrid approach with a population-based image-derived input function. Quantification was carried out in three areas: the cerebral cortex, muscle, and tumor lesion. Micro- and macro-kinetic parameters for different scan durations were estimated by assuming an irreversible two-tissue compartment model. The visual performance of parametric images and region of interest-based quantification in several parameters were evaluated. Furthermore, simplified quantification methods (DTW, Patlak, fractional uptake ratio [FUR], and standardized uptake value [SUV]) were compared for similarity to the reference net influx rate Ki. Results Ki and K1 derived from the DTW protocol showed overall good consistency (P < 0.01) with the reference from the 60-min dynamic scan with 10-min early scan and 5-min late scan (Ki correlation: 0.971, 0.990, and 0.990; K1 correlation: 0.820, 0.940, and 0.975 in the cerebral cortex, muscle, and tumor lesion, respectively). Similar correlationss were found for other micro-parameters. The DTW protocol had the lowest bias relative to standard Ki than any of the quantification methods, followed by FUR and Patlak. SUV had the weakest correlation with Ki. The whole-body Ki and K1 images generated by the DTW protocol were consistent with the reference parametric images. Conclusions Using the DTW protocol, the dynamic total-body FDG scan time can be reduced to 15 min while obtaining accurate Ki and K1 quantification and acceptable visual performance in parametric images. However, the trade-off between quantification accuracy and protocol implementation feasibility must be considered in practice. We recommend that the DTW protocol be used when the clinical task requires reliable visual assessment or quantifying multiple micro-parameters; FUR with a hybrid input function may be a more feasible approach to quantifying regional metabolic rate with a known lesion position or organs of interest.

show abstract

First results on kinetic modelling and parametric imaging of dynamic 18F-FDG datasets from a long axial FOV PET scanner in oncological patients

Cited by 74 publications

References 41 publications

Quantitative evaluation of a deep learning-based framework to generate whole-body attenuation maps using LSO background radiation in long axial FOV PET scanners

Quantitative evaluation of a deep learning-based framework to generate whole-body attenuation maps using LSO background radiation in long axial FOV PET scanners

Feasibility of using abbreviated scans protocols with population-based input functions for accurate kinetic modelling of 18F-FDG datasets from a long-axial FOV PET scanner

Comparison between a dual-time-window protocol and other simplified protocols for dynamic total-body 18F-FDG PET imaging

Contact Info

Product

Resources

About