Automated extraction of the arterial input function from brain images for parametric PET studies.

BACKGROUND: Accurate measurement of the arterial input function (AIF) is crucial for parametric PET studies, but the AIF is commonly derived from invasive arterial blood sampling. It is possible to use an image-derived input function (IDIF) obtained by imaging a large blood pool, but IDIF measurement in PET brain studies performed on standard field of view scanners is challenging due to lack of a large blood pool in the field-of-view. Here we describe a novel automated approach to estimate the AIF from brain images. RESULTS: Total body 18F-FDG PET data from 12 subjects were split into a model adjustment group (n = 6) and a validation group (n = 6). We developed an AIF estimation framework using wavelet-based methods and unsupervised machine learning to distinguish arterial and venous activity curves, compared to the IDIF from the descending aorta. All of the automatically extracted AIFs in the validation group had similar shape to the IDIF derived from the descending aorta IDIF. The average area under the curve error and normalised root mean square error across validation data were - 1.59 ± 2.93% and 0.17 ± 0.07. CONCLUSIONS: Our automated AIF framework accurately estimates the AIF from brain images. It reduces operator-dependence, and could facilitate the clinical adoption of parametric PET.

Mapping 18F-FDG Kinetics Together with Patient-Specific Bootstrap Assessment of Uncertainties: An Illustration with Data from a PET/CT Scanner with a Long Axial Field of View.

The purpose of this study was to examine a nonparametric approach to mapping kinetic parameters and their uncertainties with data from the emerging generation of dynamic whole-body PET/CT scanners. Methods: Dynamic PET 18F-FDG data from a set of 24 cancer patients studied on a long-axial-field-of-view PET/CT scanner were considered. Kinetics were mapped using a nonparametric residue mapping (NPRM) technique. Uncertainties were evaluated using an image-based bootstrapping methodology. Kinetics and bootstrap-derived uncertainties are reported for voxels, maximum-intensity projections, and volumes of interest (VOIs) corresponding to several key organs and lesions. Comparisons between NPRM and standard 2-compartment (2C) modeling of VOI kinetics are carefully examined. Results: NPRM-generated kinetic maps were of good quality and well aligned with vascular and metabolic 18F-FDG patterns, reasonable for the range of VOIs considered. On a single 3.2-GHz processor, the specification of the bootstrapping model took 140 min; individual bootstrap replicates required 80 min each. VOI time-course data were much more accurately represented, particularly in the early time course, by NPRM than by 2C modeling constructs, and improvements in fit were statistically highly significant. Although 18F-FDG flux values evaluated by NPRM and 2C modeling were generally similar, significant deviations between vascular blood and distribution volume estimates were found. The bootstrap enables the assessment of quite complex summaries of mapped kinetics. This is illustrated with maximum-intensity maps of kinetics and their uncertainties. Conclusion: NPRM kinetics combined with image-domain bootstrapping is practical with large whole-body dynamic 18F-FDG datasets. The information provided by bootstrapping could support more sophisticated uses of PET biomarkers used in clinical decision-making for the individual patient.

Relationship Between Reactive Astrocytes, by [18F]SMBT-1 Imaging, with Amyloid-Beta, Tau, Glucose Metabolism, and TSPO in Mouse Models of Alzheimer's Disease.

Reactive astrocytes play an important role in the development of Alzheimer's disease (AD). Here, we aimed to investigate the temporospatial relationships among monoamine oxidase-B, tau and amyloid-ß (Aß), translocator protein, and glucose metabolism by using multitracer imaging in AD transgenic mouse models. Positron emission tomography (PET) imaging with [18F]SMBT-1 (monoamine oxidase-B), [18F]florbetapir (Aß), [18F]PM-PBB3 (tau), [18F]fluorodeoxyglucose (FDG), and [18F]DPA-714 (translocator protein) was carried out in 5- and 10-month-old APP/PS1, 11-month-old 3×Tg mice, and aged-matched wild-type mice. The brain regional referenced standard uptake value (SUVR) was computed with the cerebellum as the reference region. Immunofluorescence staining was performed on mouse brain tissue slices. [18F]SMBT-1 and [18F]florbetapir SUVRs were greater in the cortex and hippocampus of 10-month-old APP/PS1 mice than in those of 5-month-old APP/PS1 mice and wild-type mice. No significant difference in the regional [18F]FDG or [18F]DPA-714 SUVRs was observed in the brains of 5- or 10-month-old APP/PS1 mice or wild-type mice. No significant difference in the SUVRs of any tracer was observed between 11-month-old 3×Tg mice and age-matched wild-type mice. A positive correlation between the SUVRs of [18F]florbetapir and [18F]DPA-714 in the cortex and hippocampus was observed among the transgenic mice. Immunostaining validated the distribution of MAO-B and limited Aß and tau pathology in 11-month-old 3×Tg mice; and Aß deposits in brain tissue from 10-month-old APP/PS1 mice. In summary, these findings provide in vivo evidence that an increase in astrocyte [18F]SMBT-1 accompanies Aß accumulation in APP/PS1 models of AD amyloidosis.

The Value of Dual Time Point 18F-FDG PET/CT Imaging in Differentiating Lymph Node Metastasis From Reactive Hyperplasia in Bladder Urothelial Carcinoma.

RATIONALE AND OBJECTIVES: This study explored the clinical value of dual time-point 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography (PET) imaging for differentiating lymph node metastasis from lymph nodes with reactive hyperplasia. METHODS: 250 lymph nodes from 153 bladder cancer patients who underwent 18F-FDG PET/computed tomography (CT) delayed diuretic imaging were analyzed. The maximum and mean standardized uptake values (SUVmax and SUVmean, respectively), metabolic tumor volume (MTV), and related delay indices before and after PET delayed imaging were obtained. Relationships with outcomes were analyzed using nonparametric and multivariate analyses. Receiver operating characteristic curves and nomograms were drawn to predict lymph node metastasis. RESULTS: Delayed PET/CT imaging showed better detection of hyperplasia and metastatic lymph nodes. Delayed imaging with a cutoff SUVmax of 2.0 or 2.5 increased the detection rate of metastatic lymph nodes by 4.1%, and 6.9%, respectively. Delayed imaging often showed speckle-like radioactive foci in lymph nodes with reactive hyperplasia and increased FDG uptake throughout the nodes in metastatic lymph nodes. The lymph node short-axis diameter, SUVmean, and delayed index of MTV (DIMTV) were independent predictors for differentiating metastatic lymph nodes from reactive hyperplasia, and their combination showed better differentiation performance than the individual predictors. In high-risk patients, the probability of lymph node metastasis was as high as 97.6%. CONCLUSION: Dual time-point imaging can detect more metastatic lymph nodes. Some lymph nodes with hyperplasia show speckle-like radioactive foci on delayed imaging. The lymph node short-axis diameter, SUVmean, and DIMTV are three important parameters for predicting lymph node metastasis.

Detection of individual brain tau deposition in Alzheimer's disease based on latent feature-enhanced generative adversarial network.

OBJECTIVE: The conventional methods for interpreting tau PET imaging in Alzheimer's disease (AD), including visual assessment and semi-quantitative analysis of fixed hallmark regions, are insensitive to detect individual small lesions because of the spatiotemporal neuropathology's heterogeneity. In this study, we proposed a latent feature-enhanced generative adversarial network model for the automatic extraction of individual brain tau deposition regions. METHODS: The latent feature-enhanced generative adversarial network we propose can learn the distribution characteristics of tau PET images of cognitively normal individuals and output the abnormal distribution regions of patients. This model was trained and validated using 1131 tau PET images from multiple centres (with distinct races, i.e., Caucasian and Mongoloid) with different tau PET ligands. The overall quality of synthetic imaging was evaluated using structural similarity (SSIM), peak signal to noise ratio (PSNR), and mean square error (MSE). The model was compared to the fixed templates method for diagnosing and predicting AD. RESULTS: The reconstructed images archived good quality, with SSIM = 0.967 ± 0.008, PSNR = 31.377 ± 3.633, and MSE = 0.0011 ± 0.0007 in the independent test set. The model showed higher classification accuracy (AUC = 0.843, 95 % CI = 0.796-0.890) and stronger correlation with clinical scales (r = 0.508, P < 0.0001). The model also achieved superior predictive performance in the survival analysis of cognitive decline, with a higher hazard ratio: 3.662, P < 0.001. INTERPRETATION: The LFGAN4Tau model presents a promising new approach for more accurate detection of individualized tau deposition. Its robustness across tracers and races makes it a potentially reliable diagnostic tool for AD in practice.

Towards quality management of artificial intelligence systems for medical applications.

The use of artificial intelligence systems in clinical routine is still hampered by the necessity of a medical device certification and/or by the difficulty of implementing these systems in a clinic's quality management system. In this context, the key questions for a user are how to ensure robust model predictions and how to appraise the quality of a model's results on a regular basis. In this paper we discuss some conceptual foundation for a clinical implementation of a machine learning system and argue that both vendors and users should take certain responsibilities, as is already common practice for high-risk medical equipment. We propose the methodology from AAPM Task Group 100 report No. 283 as a conceptual framework for developing risk-driven a quality management program for a clinical process that encompasses a machine learning system. This is illustrated with an example of a clinical workflow. Our analysis shows how the risk evaluation in this framework can accommodate artificial intelligence based systems independently of their robustness evaluation or the user's in-house expertise. In particular, we highlight how the degree of interpretability of a machine learning system can be systematically accounted for within the risk evaluation and in the development of a quality management system.

Reduced SV2A and GABAA receptor levels in the brains of type 2 diabetic rats revealed by [18F]SDM-8 and [18F]flumazenil PET.

PURPOSE: Type 2 diabetes mellitus (T2DM) is associated with a greater risk of Alzheimer's disease. Synaptic impairment and protein aggregates have been reported in the brains of T2DM models. Here, we assessed whether neurodegenerative changes in synaptic vesicle 2 A (SV2A), γ-aminobutyric acid type A (GABAA) receptor, amyloid-ß, tau and receptor for advanced glycosylation end product (RAGE) can be detected in vivo in T2DM rats. METHODS: Positron emission tomography (PET) using [18F]SDM-8 (SV2A), [18F]flumazenil (GABAA receptor), [18F]florbetapir (amyloid-ß), [18F]PM-PBB3 (tau), and [18F]FPS-ZM1 (RAGE) was carried out in 12-month-old diabetic Zucker diabetic fatty (ZDF) and SpragueDawley (SD) rats. Immunofluorescence staining, Thioflavin S staining, proteomic profiling and pathway analysis were performed on the brain tissues of ZDF and SD rats. RESULTS: Reduced cortical [18F]SDM-8 uptake and cortical and hippocampal [18F]flumazenil uptake were observed in 12-month-old ZDF rats compared to SD rats. The regional uptake of [18F]florbetapir and [18F]PM-PBB3 was comparable in the brains of 12-month-old ZDF and SD rats. Immunofluorescence staining revealed Thioflavin S-negative, phospho-tau-positive inclusions in the cortex and hypothalamus in the brains of ZDF rats and the absence of amyloid-beta deposits. The level of GABAA receptors was lower in the cortex of ZDF rats than SD rats. Proteomic analysis further demonstrated that, compared with SD rats, synaptic-related proteins and pathways were downregulated in the hippocampus of ZDF rats. CONCLUSION: These findings provide in vivo evidence for regional reductions in SV2A and GABAA receptor levels in the brains of aged T2DM ZDF rats.

Comparative analysis of [18F]F-FAPI PET/CT, [18F]F-FDG PET/CT and magnetization transfer MR imaging to detect intestinal fibrosis in Crohn's disease: A prospective animal model and human cohort study.

PURPOSE: Accurately and early detection of intestinal fibrosis in Crohn's disease (CD) is crucial for clinical management yet remains an unmet need. Fibroblast activation protein inhibitor (FAPI) PET/CT has emerged as a promising tool to assess fibrosis. We aimed to investigate the diagnostic capability of [18F]F-FAPI PET/CT in detecting intestinal fibrosis and compared it with[18F]F-FDG PET/CT and magnetization transfer MR imaging (MTI). METHODS: Twenty-two rats underwent TNBS treatment to simulate fibrosis development, followed by three quantitative imaging sessions within one week. Mean and maximum standardized uptake values (SUVmean and SUVmax) were calculated on[18F]F-FAPI and [18F]F-FDG PET/CT, along with normalized magnetization transfer ratio on MTI. Intestinal fibrosis was assessed pathologically, with MTI serving as imaging standard for fibrosis. The diagnostic efficacy of imaging parameters in fibrosis was compared using pathological and imaging standards. Ten patients with 34 bowel strictures were prospectively recruited to validate their diagnostic performance, using the identical imaging protocol. RESULTS: In CD patients, the accuracy of FAPI uptake (both AUCs = 0.87, both P ≤ 0.01) in distinguishing non-to-mild from moderate-to-severe fibrosis was higher than FDG uptake (both AUCs = 0.82, P ≤ 0.01) and comparable to MTI (AUCs = 0.90, P ≤ 0.001). In rats, FAPI uptake responded earlier to fibrosis development than FDG and MTI; consistently, during early phase, FAPI uptake showed a stronger correlation (SUVmean: R = 0.69) with pathological fibrosis than FDG (SUVmean: R = 0.17) and MTI (R = 0.52). CONCLUSION: The diagnostic efficacy of [18F]F-FAPI PET/CT in detecting CD fibrosis is superior to [18F]F-FDG PET/CT and comparable to MTI, exhibiting great potential for early detection of intestinal fibrosis.

A Review on Low-Dose Emission Tomography Post-Reconstruction Denoising with Neural Network Approaches.

Low-dose emission tomography (ET) plays a crucial role in medical imaging, enabling the acquisition of functional information for various biological processes while minimizing the patient dose. However, the inherent randomness in the photon counting process is a source of noise which is amplified in low-dose ET. This review article provides an overview of existing post-processing techniques, with an emphasis on deep neural network (NN) approaches. Furthermore, we explore future directions in the field of NN-based low-dose ET. This comprehensive examination sheds light on the potential of deep learning in enhancing the quality and resolution of low-dose ET images, ultimately advancing the field of medical imaging.

Intelligent ultrafast total-body PET for sedation-free pediatric [18F]FDG imaging.

PURPOSE: This study aims to develop deep learning techniques on total-body PET to bolster the feasibility of sedation-free pediatric PET imaging. METHODS: A deformable 3D U-Net was developed based on 245 adult subjects with standard total-body PET imaging for the quality enhancement of simulated rapid imaging. The developed method was first tested on 16 children receiving total-body [18F]FDG PET scans with standard 300-s acquisition time with sedation. Sixteen rapid scans (acquisition time about 3 s, 6 s, 15 s, 30 s, and 75 s) were retrospectively simulated by selecting the reconstruction time window. In the end, the developed methodology was prospectively tested on five children without sedation to prove the routine feasibility. RESULTS: The approach significantly improved the subjective image quality and lesion conspicuity in abdominal and pelvic regions of the generated 6-s data. In the first test set, the proposed method enhanced the objective image quality metrics of 6-s data, such as PSNR (from 29.13 to 37.09, p < 0.01) and SSIM (from 0.906 to 0.921, p < 0.01). Furthermore, the errors of mean standardized uptake values (SUVmean) for lesions between 300-s data and 6-s data were reduced from 12.9 to 4.1% (p < 0.01), and the errors of max SUV (SUVmax) were reduced from 17.4 to 6.2% (p < 0.01). In the prospective test, radiologists reached a high degree of consistency on the clinical feasibility of the enhanced PET images. CONCLUSION: The proposed method can effectively enhance the image quality of total-body PET scanning with ultrafast acquisition time, leading to meeting clinical diagnostic requirements of lesion detectability and quantification in abdominal and pelvic regions. It has much potential to solve the dilemma of the use of sedation and long acquisition time that influence the health of pediatric patients.

Surgical radioguidance with beta-emitting radionuclides; challenges and possibilities: A position paper by the EANM.

Radioguidance that makes use of ß-emitting radionuclides is gaining in popularity and could have potential to strengthen the range of existing radioguidance techniques. While there is a strong tendency to develop new PET radiotracers, due to favorable imaging characteristics and the success of theranostics research, there are practical challenges that need to be overcome when considering use of ß-emitters for surgical radioguidance. In this position paper, the EANM identifies the possibilities and challenges that relate to the successful implementation of ß-emitters in surgical guidance, covering aspects related to instrumentation, radiation protection, and modes of implementation.

The D-SPECT SH reconstruction protocol: improved quantification of small left ventricle volumes.

BACKGROUND: Due to spatial resolution limitations, conventional NaI-SPECT typically overestimates the left ventricular (LV) ejection fraction (EF) in patients with small LV volumes. The purpose of this study was to explore the clinical application value of the small heart (SH) reconstruction protocol embedded in the postprocessing procedure of D-SPECT. METHODS: We retrospectively analyzed patients who undergo both D-SPECT and echocardiography (Echo) within one week. Patients with small LV volume were defined as those with a rest end-systolic volume (rESV) ≤ 25 mL and underwent reconstruction using the standard (SD) reconstruction protocol. The SH protocol was deemed successful in correcting the LVEF value if it decreased by 5% or more compared to the SD protocol. The ROC curve was used to calculate the optimal cutoff value of the SH protocol. LVEF, ESV and EDV were computed with SD and SH, respectively. Echo was performed as a reference, and Echo-LVEF, ESV, and EDV were calculated using the Teichholz formula. One-way ANOVA was used to compare these parameters among the three groups. RESULTS: The final study included 209 patients (73.21% female, age 67.34 ± 7.85 years). Compared with the SD protocol, the SH protocol significantly decreased LVEF (67.43 ± 7.38% vs. 71.30 ± 7.61%, p < 0.001). The optimal cutoff value for using the SH protocol was rESV > 17 mL (AUC = 0.651, sensitivity = 78.43%, specificity = 45.57%, p = 0.001). In the subgroup of rESV > 17 mL, there was no significant difference in LVEF (61.84 ± 4.67% vs. 62.83 ± 2.85%, p = 0.481) between the SH protocol and Echo, and no significant difference was observed in rESV (26.92 ± 3.25 mL vs. 27.94 ± 7.96 mL, p = 0.60) between the SH protocol and Echo. CONCLUSION: This pilot study demonstrated that the SH reconstruction protocol was able to effectively correct the overestimation of LVEF in patients with small LV volumes. Particularly, in the rESV > 17 mL subgroup, the time and computing power waste could be reduced while still ensuring the accuracy of the LVEF value and image quality.

ParaPET: non-invasive deep learning method for direct parametric brain PET reconstruction using histoimages.

BACKGROUND: The indirect method for generating parametric images in positron emission tomography (PET) involves the acquisition and reconstruction of dynamic images and temporal modelling of tissue activity given a measured arterial input function. This approach is not robust, as noise in each dynamic image leads to a degradation in parameter estimation. Direct methods incorporate into the image reconstruction step both the kinetic and noise models, leading to improved parametric images. These methods require extensive computational time and large computing resources. Machine learning methods have demonstrated significant potential in overcoming these challenges. But they are limited by the requirement of a paired training dataset. A further challenge within the existing framework is the use of state-of-the-art arterial input function estimation via temporal arterial blood sampling, which is an invasive procedure, or an additional magnetic resonance imaging (MRI) scan for selecting a region where arterial blood signal can be measured from the PET image. We propose a novel machine learning approach for reconstructing high-quality parametric brain images from histoimages produced from time-of-flight PET data without requiring invasive arterial sampling, an MRI scan, or paired training data from standard field-of-view scanners. RESULT: The proposed is tested on a simulated phantom and five oncological subjects undergoing an 18F-FDG-PET scan of the brain using Siemens Biograph Vision Quadra. Kinetic parameters set in the brain phantom correlated strongly with the estimated parameters (K1, k2 and k3, Pearson correlation coefficient of 0.91, 0.92 and 0.93) and a mean squared error of less than 0.0004. In addition, our method significantly outperforms (p < 0.05, paired t-test) the conventional nonlinear least squares method in terms of contrast-to-noise ratio. At last, the proposed method was found to be 37% faster than the conventional method. CONCLUSION: We proposed a direct non-invasive DL-based reconstruction method and produced high-quality parametric maps of the brain. The use of histoimages holds promising potential for enhancing the estimation of parametric images, an area that has not been extensively explored thus far. The proposed method can be applied to subject-specific dynamic PET data alone.

Similarity between structural and proxy estimates of brain connectivity.

Functional magnetic resonance and diffusion weighted imaging have so far made a major contribution to delineation of the brain connectome at the macroscale. While functional connectivity (FC) was shown to be related to structural connectivity (SC) to a certain degree, their spatial overlap is unknown. Even less clear are relations of SC with estimates of connectivity from inter-subject covariance of regional F18-fluorodeoxyglucose uptake (FDGcov) and grey matter volume (GMVcov). Here, we asked to what extent SC underlies three proxy estimates of brain connectivity: FC, FDGcov and GMVcov. Simultaneous PET/MR acquisitions were performed in 56 healthy middle-aged individuals. Similarity between four networks was assessed using Spearman correlation and convergence ratio (CR), a measure of spatial overlap. Spearman correlation coefficient was 0.27 for SC-FC, 0.40 for SC-FDGcov, and 0.15 for SC-GMVcov. Mean CRs were 51% for SC-FC, 48% for SC-FDGcov, and 37% for SC-GMVcov. These results proved to be reproducible and robust against image processing steps. In sum, we found a relevant similarity of SC with FC and FDGcov, while GMVcov consistently showed the weakest similarity. These findings indicate that white matter tracts underlie FDGcov to a similar degree as FC, supporting FDGcov as estimate of functional brain connectivity.

Distinct subtypes of spatial brain metabolism patterns in Alzheimer's disease identified by deep learning-based FDG PET clusters.

PURPOSE: Alzheimer's disease (AD) is a heterogeneous disease that presents a broad spectrum of clinicopathologic profiles. To date, objective subtyping of AD independent of disease progression using brain imaging has been required. Our study aimed to extract representations of unique brain metabolism patterns different from disease progression to identify objective subtypes of AD. METHODS: A total of 3620 FDG brain PET images with AD, mild cognitive impairment (MCI), and cognitively normal (CN) were obtained from the ADNI database from 1607 participants at enrollment and follow-up visits. A conditional variational autoencoder model was trained on FDG brain PET images of AD patients with the corresponding condition of AD severity score. The k-means algorithm was applied to generate clusters from the encoded representations. The trained deep learning-based cluster model was also transferred to FDG PET of MCI patients and predicted the prognosis of subtypes for conversion from MCI to AD. Spatial metabolism patterns, clinical and biological characteristics, and conversion rate from MCI to AD were compared across the subtypes. RESULTS: Four distinct subtypes of spatial metabolism patterns in AD with different brain pathologies and clinical profiles were identified: (i) angular, (ii) occipital, (iii) orbitofrontal, and (iv) minimal hypometabolic patterns. The deep learning model was also successfully transferred for subtyping MCI, and significant differences in frequency (P < 0.001) and risk of conversion (log-rank P < 0.0001) from MCI to AD were observed across the subtypes, highest in S2 (35.7%) followed by S1 (23.4%). CONCLUSION: We identified distinct subtypes of AD with different clinicopathologic features. The deep learning-based approach to distinguish AD subtypes on FDG PET could have implications for predicting individual outcomes and provide a clue to understanding the heterogeneous pathophysiology of AD.

Integrating TSPO PET imaging and transcriptomics to unveil the role of neuroinflammation and amyloid-ß deposition in Alzheimer's disease.

PURPOSE: Despite the revealed role of immunological dysfunctions in the development and progression of Alzheimer's disease (AD) through animal and postmortem investigations, direct evidence regarding the impact of genetic factors on microglia response and amyloid-ß (Aß) deposition in AD individuals is lacking. This study aims to elucidate this mechanism by integrating transcriptomics and TSPO, Aß PET imaging in clinical AD cohort. METHODS: We analyzed 85 patients with PET/MR imaging for microglial activation (TSPO, [18F]DPA-714) and Aß ([18F]AV-45) within the prospective Alzheimer's Disease Immunization and Microbiota Initiative Study Cohort (ADIMIC). Immune-related differentially expressed genes (IREDGs), identified based on AlzData, were screened and verified using blood samples from ADIMIC. Correlation and mediation analyses were applied to investigate the relationships between immune-related genes expression, TSPO and Aß PET imaging. RESULTS: TSPO uptake increased significantly both in aMCI (P < 0.05) and AD participants (P < 0.01) and showed a positive correlation with Aß deposition (r = 0.42, P < 0.001). Decreased expression of TGFBR3, FABP3, CXCR4 and CD200 was observed in AD group. CD200 expression was significantly negatively associated with TSPO PET uptake (r =-0.33, P = 0.013). Mediation analysis indicated that CD200 acted as a significant mediator between TSPO uptake and Aß deposition (total effect B = 1.92, P = 0.004) and MMSE score (total effect B =-54.01, P = 0.003). CONCLUSION: By integrating transcriptomics and TSPO PET imaging in the same clinical AD cohort, this study revealed CD200 played an important role in regulating neuroinflammation, Aß deposition and cognitive dysfunction.

Methodological evaluation of original articles on radiomics and machine learning for outcome prediction based on positron emission tomography (PET).

AIM: Despite a vast number of articles on radiomics and machine learning in positron emission tomography (PET) imaging, clinical applicability remains limited, partly owing to poor methodological quality. We therefore systematically investigated the methodology described in publications on radiomics and machine learning for PET-based outcome prediction. METHODS: A systematic search for original articles was run on PubMed. All articles were rated according to 17 criteria proposed by the authors. Criteria with >2 rating categories were binarized into "adequate" or "inadequate". The association between the number of "adequate" criteria per article and the date of publication was examined. RESULTS: One hundred articles were identified (published between 07/2017 and 09/2023). The median proportion of articles per criterion that were rated "adequate" was 65% (range: 23-98%). Nineteen articles (19%) mentioned neither a test cohort nor cross-validation to separate training from testing. The median number of criteria with an "adequate" rating per article was 12.5 out of 17 (range, 4-17), and this did not increase with later dates of publication (Spearman's rho, 0.094; p = 0.35). In 22 articles (22%), less than half of the items were rated "adequate". Only 8% of articles published the source code, and 10% made the dataset openly available. CONCLUSION: Among the articles investigated, methodological weaknesses have been identified, and the degree of compliance with recommendations on methodological quality and reporting shows potential for improvement. Better adherence to established guidelines could increase the clinical significance of radiomics and machine learning for PET-based outcome prediction and finally lead to the widespread use in routine clinical practice.

Lu-177-PSMA dosimetry for kidneys and tumors based on SPECT images at two imaging time points.

Background: Personalized dosimetry for Lu-177-PSMA treatment requires multiple-time-point SPECT/CT scans to calculate time-integrated activity (TIA). This study evaluates two-time-point (TTP) methods for TIA calculation for kidneys and tumors. Methods: A total of 18 patients treated with 3.7-7.4 GBq Lu-177 PSMA-617 were analyzed retrospectively, including 18 sets of left and right kidneys, as well as 45 tumors. Four quantitative SPECT/CT (4TP) were acquired at 2 h, 20 h, 40 h, 60 h (n = 11), or 200 h (n = 7) after treatment, and they were fit bi-exponentially as reference. The TTP method was fitted by a mono-exponential washout function using two selected imaging time points for kidneys. For tumors, one uptake and one washout phase were modeled, assuming linear (type I) and same (type II) uptake phase between 0 h to the first time point and mono-exponential washout thereafter. Two single-time-point (STP) methods were also implemented for comparison. TIA calculated by TTP and STP methods were compared with reference to the 4TP TIA. Results: For the kidneys, the TTP methods using 20 h-60 h and 40 h-200 h had smaller mean absolute errors of 8.05 ± 6.05% and 4.95 ± 3.98%, respectively, as compared to other combinations of time points and STP methods. For tumors, the type I and type II TTP methods using 20h-60 h and 40-200 h had smaller mean absolute errors of 6.14 ± 5.19% and 12.22 ± 4.44%, and 8.31 ± 7.16% and 4.48 ± 7.10%, respectively, as compared to other TTP and STP methods. Conclusion: The TTP methods based on later imaging time demonstrated fewer errors than the STP methods in kidney and tumor TIA. Imaging at 20 h-60 h and 40 h-200 h could simplify the dosimetry procedures with fewer TIA estimation errors.

Machine learning methods for tracer kinetic modelling.

Tracer kinetic modelling based on dynamic PET is an important field of Nuclear Medicine for quantitative functional imaging. Yet, its implementation in clinical routine has been constrained by its complexity and computational costs. Machine learning poses an opportunity to improve modelling processes in terms of arterial input function prediction, the prediction of kinetic modelling parameters and model selection in both clinical and preclinical studies while reducing processing time. Moreover, it can help improving kinetic modelling data used in downstream tasks such as tumor detection. In this review, we introduce the basics of tracer kinetic modelling and present a literature review of original works and conference papers using machine learning methods in this field.