HSHasan Sari

Hasan Sari

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Hasan Sari has a total of 122 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Impact of scan duration and scatter correction on quantitative accuracy in Yttrium-90 PET/CT imaging using LAFOV and SAFOV systems: a phantom and clinical evaluation

    by Konstantinos G. Zeimpekis, Magdalena Kulgawczuk, Carola Bregenzer, Marc Schindewolf, Christos Sachpekidis, Hasan Sari, Kalyani Pandya, Ali Afshar-Oromieh, Axel Rominger, Kuangyu Shi, et al.

    This study compared quantitative accuracy and image quality of post-radioembolization selective internal radiation therapy (SIRT) 90Y PET/CT using a long a

  • CT-Less TOF PET: challenges and innovations for quantitative imaging

    by Mohammadreza Teimoorisichani, Vladimir Panin, Hasan Sari

    More than a decade has passed since the seminal work of Defrise et al (2012 Phys. Med. Biol. 57 885–99), which demonstrated that time-of-flight (TOF) PET can, under appropriate conditions, be intrinsically self-correcting for several physical degradation factors, most notably attenuation, which in contemporary clinical practice is typically derived from CT in PET/CT systems. The work of Defrise et al (2012 Phys. Med. Biol. 57 885–99) popularized the concept of maximum-likelihood reconstruction of attenuation and activity (MLAA) and stimulated extensive subsequent methodological and clinical investigations. The many studies that were done on this subject, led to early clinical evaluations of joint activity–attenuation reconstruction algorithms based on advanced optimization strategies. In recent years, however, self-attenuation correction has received comparatively less attention, as the field has increasingly shifted toward the synthesis of attenuation maps using deep learning (DL) often in combination with physics-based constraints. Many of the proposed DL-based methods worked with non–attenuation-corrected PET images, sometimes supplemented by additional sources of information. Within this paradigm, attenuation estimation has effectively become an image-to-image translation problem, well suited to convolutional neural networks, generative adversarial networks, and more recently emerging diffusion-based models. These approaches benefit from the widespread availability of PET/CT datasets, which enable the construction of large training cohorts, although the inherently sequential nature of PET and CT acquisitions can introduce misregistration and associated biases. Against this backdrop, the DL evolution raises a fundamental question: what level of attenuation information can truly be recovered from suboptimal, incomplete or noisy PET emission data alone? Addressing this question challenges not only the expressive power of data-driven methods, but also their physical interpretability and robustness. More broadly, it invites reflection on whether decades of physics-based modeling, analytical reconstruction theory, and detector-driven innovation will be complemented, or reinterpreted within increasingly more powerful machine learning frameworks. In this review, we examine these competing and often complementary approaches to CT-less quantitative PET imaging, with a particular emphasis on TOF-enabled methods, while recognizing that definitive answers to these questions may only emerge through continued theoretical, technological, and clinical investigation.

  • Open Access

    Kinetics of the amino acid uptake tracer O-(2-[18F]fluoroethyl)-L-tyrosine (FET) in human brain

    by Korbinian Krieger, Thomas Pyka, Clemens Mingels, Hasan Sari, Albert Gjedde, Axel Rominger, Paul Cumming

    The large neutral amino acid [18F]fluoroethoxy-L-tyrosine ([18F]FET) is a popular tracer for detection and staging of intracranial tumors by positron emission tomography (PET). While its high tumoral uptake reflects over-expression of the L-type amino acid transporter (LAT1), there is little knowledge about the kinetics of [18F]FET uptake in healthy brain tissue, owing to the limited PET data in healthy volunteers, and to the requirement of an arterial input function for compartmental analysis. To address this, we used long axial field-of-view (LAFOV) dynamic 40-min recordings of 28 post-operative patients with intracranial tumors to undertake parametric brain mapping relative to an image-derived arterial input function (IDIF) obtained from the aorta. We averaged the individual parametric maps to obtain estimates of the physiological uptake relatively unaffected by individual residual lesions and resections, and tested simplified single-frame methods for quantitation. The analyses yielded estimates of regional unidirectional blood–brain clearance K1 (0.00825–0.0244 ml g−1 min−1), net blood–brain clearance Kin (0.00448–0.00913 ml g−1 min−1), and equilibrium distribution volume VT (0.126–0.495 ml g−1), where the lowest values depict white matter, and the highest values cerebellum. In our test of a simplified quantitation of [18F]FET uptake from single frame recordings, i.e., Gjedde–Patlak multilinear graphic analyses of K1 at five min post-injection and Kin at 40 min post injection, results were in good agreement with the analyses from the dynamic recordings (< 10% error). Compartmental analysis results for [18F]FET uptake in extra-tumoral human brain regions are in accord with the few prior reports, mainly obtained in experimental animals, and support the use of single frame quantitation. Present findings in relatively healthy brain should inform the interpretation of pathological [18F]FET uptake in tumors.


  • Open Access

    Evaluation of long axial field-of-view (LAFOV) PET/CT for post-treatment dosimetry in Yttrium-90 radioembolization of liver tumors: a comparative study with conventional SPECT imaging

    by Konstantinos G. Zeimpekis, Hasan Sari, Nasir Gözlügöl, Ngwe Rawlings Achangwa, Kuangyu Shi, Marc Schindewolf, Ali Afshar-Oromieh, Axel Rominger, Robert Seifert

    Purpose: Long axial field-of-view (LAFOV) positron emission tomography/computed tomography (PET/CT) scanners enable high sensitivity and wide anatomical coverage. Therefore, they seem ideal to perform post-selective internal radiation therapy (SIRT) 90Y scans, which are needed, to confirm that the dose is delivered to the tumors and that healthy organs are spared. However, it is unclear to what extent the use of LAFOV PET is feasible and which dosimetry approaches results in accurate measurements. Methods: In this retrospective analysis, a total number of 32 patients was included (median age 71, IQR 14), which had hepatocellular carcinoma, cholangiocarcinoma, or liver metastases. All patients underwent SIRT, and the post-therapy scan was acquired on a single photon emission computed tomography/computed tomography (SPECT/CT) and a LAFOV Biograph Quadra PET/CT with a 20-minute acquisition time. Post-treatment dosimetry, regarding the tumor, whole-liver and lung (LMD) absorbed dose was done using an organ-wise (Simplicit90Y) and a voxel-wise approach (HERMIA Dosimetry) which used a semi-Monte Carlo algorithm. The lung shunt fraction (LSF) was also measured using the voxel-wise approach and compared to the planned. Results: The planning, post-treatment SPECT and PET (SPECTpre, SPECTpost, PETpost) median tumor doses based on the organ-wise dosimetry were 276.0 Gy (200.0–330.0 Gy), 232.0 Gy (158.5–303.5 Gy) and 267.5 Gy (182.5–370.8 Gy). In contrast, the median voxel-wise PETpost dose was significantly smaller than the planned SPECTpre (152.5 Gy (94.8–223.8 Gy); p < 0.00001). Moreover, the median tumor absorbed dose at 90% (D90) of the tumor volume was significantly higher in SPECTpost compared with PETpost (123.5 Gy; 81.5–180.0 vs. 30.5 Gy; 11.3-106.3; p < 0.00001). The PETpost measured LSF was significantly lower compared to the planned SPECTpre (0.89%; 0.4–1.3% vs. 2.3%; 1.5–3.6%; p < 0.0001). Similarly, the measured PETpost median LMD was considerably lower to the planned SPECTpre (1.2 Gy; 0.6–2.3 vs. 2.5 Gy; 1.4–4.7; p < 0.0001). Conclusion: LAFOV PET enabled the direct measurement of post therapy lung dose and tumor doses that correlated well with the planned treatment doses. However, current voxel-wise-based tumor dosimetry seems to be inaccurate for LAFOV PET. In addition, dose volume histogram-based metrics also significantly underestimate the delivered dose. Therefore, improved dosimetry tools are needed for reliable voxel-wise 90Y dosimetry to leverage the sensitivity and spatial resolution of LAFOV PET scanners.


  • Total Body PET/CT: Clinical Value and Future Aspects of Quantification in Static and Dynamic Imaging

    by Narendra Rathod, Warissara Jutidamrongphan, Wolfram Andreas Bosbach, Yizhou Chen, Jan Luca Penner, Hasan Sari, Konstantinos Zeimpekis, Alejandro López Montes, Pawel Moskal, Ewa Stepien, et al.


  • Open Access

    90Y post-radioembolization clinical assessment with whole-body Biograph Vision Quadra PET/CT: image quality, tumor, liver and lung dosimetry

    by Konstantinos G. Zeimpekis, Lorenzo Mercolli, Maurizio Conti, Hasan Sari, Axel Rominger, Hendrik Rathke

    Purpose: Evaluation of 90Y liver radioembolization post-treatment clinical data using a whole-body Biograph Vision Quadra PET/CT to investigate the potential of protocol optimization in terms of scan time and dosimetry. Methods: 17 patients with hepatocellular carcinoma with median (IQR) injected activity 2393 (1348–3298) MBq were included. Pre-treatment dosimetry plan was based on 99mTc-MAA SPECT/CT with Simplicit90Y™ and post-treatment validation with Quadra using Simplicit90Y™ and HERMIA independently. Regarding the image analysis, mean and peak SNR, the coefficient of variation (COV) and lesion-to-background ratio (LBR) were evaluated. For the post-treatment dosimetry validation, the mean tumor, whole liver and lung absorbed dose evaluation was performed using Simplicit90Y and HERMES. Images were reconstructed with 20-, 15-, 10-, 5- and 1- min sinograms with 2, 4, 6 and 8 iterations. Wilcoxon signed rank test was used to show statistical significance (p < 0.05). Results: There was no difference of statistical significance between 20- and 5- min reconstructed times for the peak SNR, COV and LBR. In addition, there was no difference of statistical significance between 20- and 1- min reconstructed times for all dosimetry metrics. Lung dosimetry showed consistently lower values than the expected. Tumor absorbed dose based on Simplicit90Y™ was similar to the expected while HERMES consistently underestimated significantly the measured tumor absorbed dose. Finally, there was no difference of statistical significance between expected and measured tumor, whole liver and lung dose for all reconstruction times. Conclusion: In this study we evaluated, in terms of image quality and dosimetry, whole-body PET clinical images of patients after having been treated with 90Y microspheres radioembolization for liver cancer. Compared to the 20-min standard scan, the simulated 5-min reconstructed images provided equal image peak SNR and noise behavior, while performing also similarly for post-treatment dosimetry of tumor, whole liver and lung absorbed doses.


  • Self-supervised Learning for Physiologically-Based Pharmacokinetic Modeling in Dynamic PET

    by Francesca De Benetti, Walter Simson, Magdalini Paschali, Hasan Sari, Axel Rominger, Kuangyu Shi, Nassir Navab, Thomas Wendler


  • Open Access

    Investigating the influence of long-axial versus short-axial field of view PET/CT on stage migration in lymphoma and non-small cell lung cancer

    by Ian Alberts, Sigrid Seibel, Song Xue, Marco Viscione, Clemens Mingels, Hasan Sari, Ali Afshar-Oromieh, Andreas Limacher, Axel Rominger

    Objectives: The objective of this study was to evaluate the influence of a long-axial field-of-view (LAFOV) on stage migration using a large single-centre retrospective cohort in lymphoma and non-small cell lung cancer (NSCLC). Methods: A retrospective study is performed for patients undergoing PET/computed tomography (CT) on either a short-axial field-of-view (SAFOV) or LAFOV PET/CT system for the staging of known or suspected NSCLC or for therapeutic response in lymphoma. The primary endpoint was the Deauville therapy response score for patients with lymphoma for the two systems. Secondary endpoints were the American Joint Committee on Cancer stage for NSCLC, the frequency of cN3 and cM1 findings, the probability for a positive nodal staging (cN1-3) for NSCLC and the diagnostic accuracy for nodal staging in NSCLC. Results: One thousand two hundred eighteen records were screened and 597 patients were included for analysis (N = 367 for lymphoma and N = 291 for NSCLC). For lymphoma, no significant differences were found in the proportion of patients with complete metabolic response versus non-complete metabolic response Deauville response scores (P = 0.66). For NSCLC no significant differences were observed between the two scanners for the frequency of cN3 and cM1 findings, for positive nodal staging, neither the sensitivity nor the specificity. Conclusions: In this study use of a LAFOV system was neither associated with upstaging in lymphoma nor NSCLC compared to a digital SAFOV system. Diagnostic accuracy was comparable between the two systems in NSCLC despite shorter acquisition times for LAFOV.


  • Open Access

    Forecasting the Pharmacokinetics With Limited Early Frames in Dynamic Brain PET Imaging Using Neural Ordinary Differential Equation

    by Jimin Hong, Matthias Brendel, Kjell Erlandsson, Hasan Sari, Jiaying Lu, Christoph Clement, Ngoc Vinh Bui, Maria Meindl, Sibylle I. Ziegler, Henryk Barthel, et al.

    In dynamic brain positron emission tomography (PET) studies, acquiring a time series of images, typically lasting more than an hour, is necessary to derive pharmacokinetic parameters. Analytically, these parameters are estimated by establishing kinetic models such as compartment models that consist of sets of ordinary differential equations (ODE), and by fitting the sparse time-activity curve (TAC) of the tracer. Yet, these models are simplified approximations of highly complex underlying processes, and sufficient samples of TAC are required throughout the entire acquisition, which is not only impractical but also hindered by patient involuntary motion and intrinsic noise. Therefore, recovering samples in missing timeframes is often required, which, in practice, is achieved by interpolation or extrapolation. Here, we introduce a novel deep-learning-based method that utilizes neural ODE (N-ODE) to predict TAC in the extended timeframes by mimicking analytical method in a data-driven manner. By training N-ODE to solve and fit sets of ODE such that the solution replicates the observed TAC, the N-ODE converges to the functional shapes that best describe the underlying pharmacokinetic processes. We customized N-ODE to predict the full-dynamic images (12 frames, 60min), hence pharmacokinetic parameters, given limited early-frame images (7 to 9 frames, 20 to 30min). For proof of concept, the proposed N-ODE was applied to simulated and clinical 18F-PI-2620 brain PET. We demonstrated that the proposed N-ODE delivered promising performance, indicated by bias, variance, and mean absolute error as well as pharmacokinetic parameters such as rate constants, standardized uptake value ratio (SUVr), and binding potential (BPND).


  • Open Access

    Phantom-based evaluation of yttrium-90 datasets using biograph vision quadra

    by Konstantinos G. Zeimpekis, Lorenzo Mercolli, Maurizio Conti, Hasan Sari, George Prenosil, Kuangyu Shi, Axel Rominger

    Purpose: The image quality characteristics of two NEMA phantoms with yttrium-90 (90Y) were evaluated on a long axial field-of-view (AFOV) PET/CT. The purpose was to identify the optimized reconstruction setup for the imaging of patients with hepatocellular carcinoma after 90Y radioembolization. Methods: Two NEMA phantoms were used, where one had a 1:10 sphere to background activity concentration ratio and the second had cold background. Reconstruction parameters used are as follows: iterations 2 to 8, Gaussian filter 2- to 6-mm full-width-at-half-maximum, reconstruction matrices 440 × 440 and 220 × 220, high sensitivity (HS), and ultra-high sensitivity (UHS) modes. 50-, 40-, 30-, 20-, 10-, and 5-min acquisitions were reconstructed. The measurements included recovery coefficients (RC), signal-to-noise ratio (SNR), background variability, and lung error which measures the residual error in the corrections. Patient data were reconstructed with 20-, 10-, 5-, and 1-min time frames and evaluated in terms of SNR. Results: The RC for the hot phantom was 0.36, 0.45, 0.53, 0.63, 0.68, and 0.84 for the spheres with diameters of 10, 13, 17, 22, 28, and 37 mm, respectively, for UHS 2 iterations, a 220 × 220 matrix, and 50-min acquisition. The RC values did not differ with acquisition times down to 20 min. The SNR was the highest for 2 iterations, measured 11.7, 16.6, 17.6, 19.4, 21.9, and 27.7 while the background variability was the lowest (27.59, 27.08, 27.36, 26.44, 30.11, and 33.51%). The lung error was 18%. For the patient dataset, the SNR was 19%, 20%, 24%, and 31% higher for 2 iterations compared to 4 iterations for 20-, 10-, 5-, and 1-min time frames, respectively. Conclusions: This study evaluates the NEMA image quality of a long AFOV PET/CT scanner with 90Y. It provides high RC for the smallest sphere compared to other standard AFOV scanners at shorter scan times. The maximum patient SNR was for 2 iterations, 20 min, while 5 min delivers images with acceptable SNR.