Frontiers in nuclear medicine (Lausanne, Switzerland)最新文献

Background: Animal models of nerve compression have revealed neuroinflammation not only at the entrapment site, but also remotely at the spinal cord. However, there is limited information on the presence of neuroinflammation in human compression neuropathies. The objectives of this study were to: (1) assess which tracer kinetic model most optimally quantified [¹¹C]DPA713 uptake in the spinal cord and neuroforamina in patients with painful cervical radiculopathy, (2) evaluate the performance of linearized methods (e.g., Logan) and simplified (e.g., standardized uptake value - SUV) methods, and (3) assess the test-retest reliability of these methods. Microglia activation associated with neuroinflammation was quantified using positron emission tomography (PET) with the radiotracer [¹¹C]DPA713, targeting the 18 kDa translocator protein (TSPO). The Akaike information criterion, visual inspection of the fits and number of outliers were used to select the optimal kinetic model. As unaffected tissue, the spinal cord and neuroforamina three cervical levels above the affected target tissue was used.

Results: The single tissue (1T2k) compartment model was the preferred model to describe [¹¹C]DPA713 kinetics at the spinal cord and neuroforamina. Higher levels of 1T2k V _T were observed in the affected neuroforamina and spinal cord compared with corresponding unaffected tissues. Logan V _T (≥0.73) showed high correlation with 1T2k V _T at both locations. Of the simplified methods, neuroforamina and spinal cord SUV normalized for the metabolite corrected plasma (TBR-PP) exhibited high correlations with 1T2k V _T (r ≥ 0.84). Test-retest reliability varied between fair to excellent.

Conclusions: These results indicate that a 1T2k model with metabolite corrected image derived input function can be used to describe the kinetics of [¹¹C]DPA713 in the spinal cord and neuroforamina in humans. 1T2k V _T or Logan V _T can be used as binding metric, while TBR-PP is the recommended choice among simplified models.

[This corrects the article DOI: 10.3389/fnume.2025.1632112.].

Neuroendocrine tumors (NETs) are a heterogeneous group of neoplasms characterized by their overexpression of somatostatin receptors (SSTRs), which can be utilized for peptide receptor radionuclide therapy. This review provides a comprehensive update on the clinical trials of radiolabeled SSTR-targeting radiopharmaceuticals since 2020, with a focus on somatostatin receptor agonists and antagonists radiolabeled with ⁶⁸Ga, ¹⁸F, ^99mTc, ¹⁷⁷Lu, ¹⁶¹Tb, ²¹²Pb, ⁶⁷Cu, and ²²⁵Ac. Head-to-head clinical trials demonstrate that radiolabeled SSTR antagonists such as [⁶⁸Ga]Ga-DOTA-JR11 and [⁶⁸Ga]Ga-DOTA-LM3 offer improved lesion detection and tumor-to-background ratios (particularly in liver metastases) compared to radiolabeled agonists like [⁶⁸Ga]Ga-DOTA-TOC and [⁶⁴Cu]Cu-DOTA-TATE. Additionally, ¹⁸F-labeled agents offer logistical and dosimetric advantages over ⁶⁸Ga, due to ¹⁸F's longer half-life and cyclotron production, allowing for delayed imaging and increased availability to a wider range of patients. Emerging targeted alpha therapy agents, including [²²⁵Ac]Ac-DOTA-TATE, show promising results in treating disease resistant to conventional therapies due to the high linear energy transfer of alpha particles, which leads to improved localized cytotoxicity. Collectively, these developments support a shift toward more precise, receptor-specific theragnostics, emphasizing the need for further head-to-head clinical trials and integration of dosimetry-driven, personalized treatment planning in the management of NETs.

Radiocobalt-based theranostics has emerged as a promising platform in nuclear medicine that offers dual capabilities for both diagnostic imaging and targeted radionuclide therapy. ⁵⁵Co (t_1/2 = 17.53 h, β⁺ = 77%, E _γ  = 931.1 keV, I _γ  = 75%) and ^58mCo (t_1/2 = 9.10 h, IC = 100%) serve as an elementally matched pair for positron emission tomography and targeted Auger electron therapy, respectively, that enable a more personalized approach to cancer management, where imaging with ⁵⁵Co can help to guide and predict therapeutic outcomes for ^58mCo therapy. The unique coordination chemistry of cobalt allows for stable complexation with various chelators, enhancing in vivo stability and targeting efficacy when conjugated to biomolecules such as peptides, antibodies, and small molecules. Recent developments in radiolabeling techniques, chelator design, and preclinical evaluations have significantly improved the pharmacokinetic profiles and tumor specificity of radiocobalt-based radiopharmaceuticals. The aim of this mini review is to provide an overview of the recent advancements and applications of radiocobalt isotopes with a particular focus on the production, chelation chemistry, and in vivo targeting of ⁵⁵Co- and ^58mCo-labelled radiopharmaceuticals over the last 5 years. While challenges still exist in production scalability, dosimetry optimization, and clinical translations, the current trajectory suggests a growing role for radiocobalt-based theranostics in precision oncology.

Purpose: Long axial field-of-view (LAFOV) PET systems like the Siemens Biograph Vision Quadra offer unprecedented sensitivity and imaging capabilities, but compliance with EARL standards across all acquisition modes remains unexplored. This study aimed to identify reconstruction parameters meeting EARL 1 and 2 compliance for static and continuous bed motion (CBM) acquisitions in High Sensitivity (HS) and Ultra-High Sensitivity (UHS) modes on the Quadra. The research focused on optimising image quality while maintaining compliance with quantitative standards.

Methods: The International Electrotechnical Commission (IEC) body phantom was filled with ¹⁸F-FDG in a 10:1 sphere-to-background activity ratio and scanned at five positions across the field of view (FOV) using static and CBM acquisitions in HS and UHS modes. Reconstructions used standard clinical parameters, varied with Gaussian filters (1-7 mm) and matrix sizes (440, 220, 128). EARL compliance was assessed with the EARL tool to evaluate SUV recovery coefficients (RCSUVmean, RCSUVmax, RCSUVpeak). Patient images were reconstructed using standard and EARL-compliant parameters for comparison.

Results: Reconstruction parameters achieving EARL compliance were identified for all acquisition modes, with no differences between static and CBM reconstructions. Achieving EARL compliance required significant image quality reductions, especially for EARL 1, with greater degradation in UHS mode. Patient images reconstructed with EARL-compliant parameters appeared smoother and had reduced contrast compared to clinical reconstructions.

Conclusion: While EARL compliance ensures quantitative standardisation, it significantly reduces image quality, especially on advanced LAFOV PET systems. An updated "EARL 3" standard is needed to reflect the capabilities of modern systems.

Introduction: Ventilation-perfusion (V/Q) nuclear scintigraphy remains a vital diagnostic tool for assessing pulmonary embolism (PE) and other lung conditions. Interpretation of these images requires specific expertise which may benefit from recent advances in artificial intelligence (AI) to improve diagnostic accuracy and confidence in reporting. Our study aims to develop a multi-center dataset combining imaging and clinical reports to aid in creating AI models for PE diagnosis.

Methods: We established a comprehensive imaging registry encompassing patient-level V/Q image data along with relevant clinical reports, CTPA images, DVT ultrasound impressions, D-dimer lab tests, and thrombosis unit records. Data extraction was performed at two hospitals in Canada and at multiple sites in the United States, followed by a rigorous de-identification process. We utilized the V7 Darwin platform for crowdsourced annotation of V/Q images including segmentation of V/Q mismatched vascular defects. The annotated data was then ingested into Deep Lake, a SQL-based database, for AI model training. Quality assurance involved manual inspections and algorithmic validation.

Results: A query of The Ottawa Hospital's data warehouse followed by initial data screening yielded 2,137 V/Q studies with 2,238 successfully retrieved as DICOM studies. Additional contributions included 600 studies from University Health Toronto, and 385 studies by private company Segmed Inc. resulting in a total of 3,122 V/Q planar and SPECT images. The majority of studies were acquired using Siemens, Philips, and GE scanners, adhering to standardized local imaging protocols. After annotating 1,500 studies from The Ottawa Hospital, the analysis identified 138 high-probability, 168 intermediate-probability, 266 low-probability, 244 very low-probability, and 669 normal, and 15 normal perfusion with reversed mismatched ventilation defect studies. In 1,500 patients were 3,511 segmented vascular perfusion defects.

Conclusion: The VQ4PEDB comprised 8 unique ventilation agents and 11 unique scanners. The VQ4PEDB database is unique in its depth and breadth in the domain of V/Q nuclear scintigraphy for PE, comprising clinical reports, imaging studies, and annotations. We share our experience in addressing challenges associated with data retrieval, de-identification, and annotation. VQ4PEDB will be a valuable resource to development and validate AI models for diagnosing PE and other pulmonary diseases.

[This corrects the article DOI: 10.3389/fnume.2025.1575026.].

Background: The characterization of solitary pulmonary nodules (SPNs) as malignant or benign remains a diagnostic challenge using conventional imaging parameters. The literature suggests using combined Positron Emission Tomography (PET) and Computed Tomography (CT) to characterise a SPN. Radiomics and machine learning are other promising technologies which can be utilised to characterise the SPN.

Purpose: This study explores the potential of PET radiomics signatures and machine learning algorithms to characterise the SPN.

Methods: This retrospective study aimed to characterize solitary pulmonary nodules (SPNs) using PET radiomics. A total of 163 patients who underwent PET/CT imaging were included in this study. A total of 1,098 features were extracted from PET images using PyRadiomics. To optimize model performance two strategies i.e., (a) feature selection and (b) feature reduction techniques were employed, including hierarchical clustering, RFE in feature selection, and PCA in feature reduction. To address outcome class imbalance, the dataset was statistically resampled (SMOTE). A random forest models was developed using original training set (RF-Model-O & RF-PCA-Model-O) and balanced training dataset (RF-Model-B & RF-PCA-Model-B) and validated on the test datasets. Additionally, 5-fold cross-validation and bootstrap validation was also performed. The model's performance was assessed using various metrics, such as accuracy, AUC, precision, recall, and F1-score.

Results: Of the 163 patients (aged 36-76 years, mean age 58 ± 7), 117 had malignant disease and 46 had granulomatous or benign conditions. In Strategy (a), five radiomic features were identified as optimal using hierarchical clustering and RFE. In Strategy (b), five principal components were deemed optimal using PCA. The model accuracy of RF-Model-O and RF-Model-B in the train-test validation, 5-fold cross-validation and bootstrap validation were found to be 0.8, 0.80 ± 0.07, 0.84 ± 1.11 and 0.8, 0.83 ± 0.10, 0.80 ± 0.07 in Strategy (a). Similarly, the model accuracy of RF-PCA-Model-O and RF-PCA-Model-B in the train-test validation, 5-fold cross-validation and bootstrap validation were found to be 0.84, 0.80 ± 0.07, 0.84 ± 07 and 0.74, 0.80 ± 0.08, 0.75 ± 0.08 in Strategy (b).

Conclusion: The PET radiomics demonstrated excellent performance in characterizing SPNs as benign or malignant.

Introduction: The COVID-19 pandemic has profoundly affected mental health, with lockdown periods particularly exacerbating negative emotions such as fear, sadness, and uncertainty. This study examines brain metabolic changes associated with the psychological context of the first French COVID-19 lockdown in vulnerable individuals.

Methods: As a proxy measure of the psychological context, we used a composite negative-emotion score derived from an open-source X/Twitter dataset ("The First French COVID-19 Lockdown Twitter Dataset"), designed to capture public sentiment over the 55-day lockdown. This score was day-by-day correlated with whole-brain voxel-based [18F]FDG PET imaging in 95 patients with neurological conditions, using statistical parametric mapping (SPM) (p-voxel < 0.001, k > 108).

Results: A significant negative correlation was found between daily negative-emotion scores and metabolism in the right ventromedial prefrontal cortex (vmPFC) and anterior cingulate cortex (ACC), key regions of the brain's fear circuit. Inter-regional correlation analysis (IRCA) of metabolic connectivity from the right vmPFC/ACC further revealed a right limbic-dominant network including the amygdala, hippocampus, thalamus, and basal ganglia.

Discussion: These findings highlight the sensitivity of the right vmPFC/ACC to societal emotional stressors, suggesting a potential cerebral substrate for the increase in psychological and psychiatric disorders observed during the pandemic. Further research is needed to validate these results in larger populations and to explore their longitudinal implications, to better understand the neurological impact of collective stress.

Kinetic modelling of brain PET data is crucial for estimating quantitative biological parameters, traditionally requiring arterial sampling. This study evaluated whether arterial samples could be omitted to estimate the image-derived input function (IDIF) using a long axial field-of-view PET scanner. The use of internal carotid arteries (ICA) for IDIF estimation, along with venous samples for plasma-to-whole blood ratios and plasma parent fractions, was also assessed. Six healthy volunteers underwent [¹⁸F]MC225 scans with manual arterial sampling. IDIFs were derived from the aortic arch (IDIF_AA) and calibrated using manual arterial samples (IDIF_{AA_CAL}). ICA-derived IDIF was also calibrated (IDIF_{CA_CAL}) and compared to IDIF_{AA_CAL}. In a separate group of six volunteers, venous and arterial samples were collected to evaluate plasma-to-whole blood ratios, plasma parent fractions, and IDIF calibration (IDIF_{CA_CAL_VEN}). Volume of distribution (V_T) of different brain regions was estimated for all IDIFs techniques, corrected for plasma-to-whole blood ratio and plasma parent fraction (IDIF_AA,P, IDIF_{AA_CAL,P}, IDIF_{ICA_CAL,P} and IDIF_{ICA_CAL_VEN_P}). Our findings revealed discrepancies between IDIF_AA and arterial samples, highlighting the importance of calibration. The differences between IDIF_AA,P and IDIF_{AA_CAL,P} were 9.2% for area under the curve and 4.0% for brain V_T. IDIF_{ICA_CAL,P} showed strong agreement with IDIF_{A_CAL,P}, with 1.2% V_T difference. Venous sampling showed consistent agreement with arterial sampling for plasma parameters but was unreliable for IDIF calibration, leading to 39% V_T differences. This study emphasises that arterial samples are still required for IDIF calibration and reliable V_T estimation for [¹⁸F]MC225 PET tracer. ICA-derived IDIF, when calibrated, provides reliable V_T estimates. Venous sampling is a potential alternative for estimating plasma parameters, but it is unsuitable for IDIF calibration.

Trial registry: NCT05618119 (clinicaltrials.gov/study/NCT05618119).