Annals of Nuclear Medicine最新文献

¹⁸F-Fluciclovine was the first ¹⁸F-labeled amino acid PET tracer to be approved for clinical use in Japan, receiving regulatory approval in March 2021 and being listed for reimbursement in June 2024. In response to this development, the Japanese Society of Nuclear Medicine initiated the formulation of clinical guidelines to ensure the appropriate use of this radiopharmaceutical in clinical practice. This guideline provides a comprehensive overview of the clinical characteristics of ¹⁸F-Fluciclovine in malignant glioma, including indications for use, imaging protocols, interpretation of PET images, and considerations for radiation safety. The Japanese version of this guideline was compiled by a voluntary editorial committee and officially approved by the Japanese Society of Nuclear Medicine on August 16, 2024. The primary objective of this guideline is to consolidate the current scientific evidence on ¹⁸F-Fluciclovine and to clarify its clinical utility, appropriate usage, and imaging methodologies. By doing so, it aims to promote the proper implementation of ¹⁸F-Fluciclovine in clinical settings and to serve as a reference for future applications related to the expansion of insurance coverage and reimbursement decisions.

It is recommended that PET examinations using ¹⁸F-Fluciclovine in Japan be conducted in accordance with this guideline. Although the content is tailored to the Japanese medical system and regulatory framework, the imaging protocols, radiation safety management, and interpretation methods described herein are also expected to be internationally applicable and relevant.

Objective

To investigate the prognostic value of 18F-fluorodeoxyglucose positron emission tomography-computed tomography (FDG-PET/CT) in patients with mucosal melanoma of the head and neck (MMHN) treated with carbon ion radiotherapy (CIRT).

Methods

This single-center retrospective study included patients with MMHN who underwent CIRT and FDG-PET/CT. Correlations between pre-treatment FDG-PET/CT-derived parameters, including the maximum standardized uptake variable (SUVmax), metabolic tumor volume (MTV) with a 50% threshold, total lesion glycolysis (TLG), bone marrow/liver SUVmax and mean standardized uptake variable (SUVmean) ratios, and spleen/liver SUVmax and SUVmean ratios (SLRmax, SLRmean), with clinical parameters and prognosis were statistically analyzed.

Results

A total of 32 patients with MMHN were enrolled (median age, 72.5 years). The tumor stages were distributed as follows: T3, 17 patients; T4a, 14 patients; T4b, one patient. The median total observation period was 22.6 months, the median overall survival (OS) was 21.6 months, and the median progression-free survival (PFS) was 11.5 months. Thirteen patients (40.6%) died, 10 (31.3%) experienced local recurrence, and 19 (59.4%) had distant metastases during the observation period. The 1 and 3-year survival rates were 78.1% and 62.5%, respectively. FDG-PET/CT showed pronounced positive uptake for all tumors (median SUVmax: 13.8, range 2.7–33.0). SLRmax was high in patients with negative programmed death-ligand 1 expression in the tumor (p = 0.05). PFS was shorter in patients with a high MTV (p = 0.018). In the multivariate analysis, MTV was an independent prognostic factor for PFS (hazard ratio, 2.60; 95% confidence interval, 1.065–6.345; p = 0.036). MTV and TLG were not predictive of OS in the univariate analysis.

Conclusions

FDG-PET/CT showed a strong positive uptake for MMHN. FDG-PET/CT-derived imaging parameters may be significant prognostic biomarkers for predicting tumor progression in patients with MMHN.

Purpose

Brown adipose tissue (BAT) contributes to thermoregulation and energy expenditure. Although BAT is abundant in early childhood and declines with age, its distribution across age groups remains unclear. This study examined age-related BAT distribution using fluorodeoxyglucose positron emission tomography/computed tomography (FDG-PET/CT).

Materials and methods

A total of 8695 FDG-PET/CT scans performed for clinical purposes were retrospectively reviewed. FDG accumulation with a standardized uptake value (SUV) max > 1.5 in known BAT regions was considered positive. BAT distribution patterns were classified into T-type (positive accumulation in the supraclavicular or axillary region), I-type (positive accumulation in the cervical or paravertebral region without supraclavicular or axillary involvement), lipomatous hypertrophy of the interatrial septum (LHIS)-type (positive accumulation localized only to the LHIS), and others (cases not fitting any type).

Results

BAT accumulation was observed in 78 patients (0.9% prevalence): T-type (18), I-type (39), LHIS-type (18), and others (3). The mean ages for T-type, I-type, LHIS-type, and others were 29.8 ± 17.3, 73.6 ± 18.1, 72.9 ± 12.5, and 67.0 ± 11.5 years, respectively. Patients in the T-type group were significantly younger than those in the I-type- and LHIS-type groups (p < 0.01).

Conclusions

This study identified three BAT distribution types, with T-type occurring in mostly younger compared with the I-type and LHIS type. Recognizing these patterns may improve FDG-PET/CT diagnostic accuracy.

Recent advances in PET image reconstruction have focused on achieving high image quality and quantitative accuracy. Bayesian penalized likelihood (BPL) algorithms, such as Q.Clear and HYPER Iterative that have been integrated into commercial PET systems offer robust image noise suppression and edge preservation through regularization. In parallel, methods based on deep learning such as SubtlePET, AiCE, uAI^® HYPER DLR, and Precision DL have emerged primarily as post-processing techniques. They use trained convolutional neural networks to reduce image noise while preserving lesion contrast. These methods have reduced image acquisition times or reduced radiotracer doses while maintaining diagnostic confidence. uAI^® HYPER DPR represents a hybrid approach by embedding deep learning in iterative reconstruction. This review summarizes the technical principles and the clinical performance of BPL and deep learning-based PET reconstruction algorithms, and discusses key considerations such as image quality and quantitative accuracy of PET images. This review should deepen understanding of advanced PET image reconstruction techniques and accelerate their clinical implementation across diverse PET imaging applications.

Objective

Vision language models (VLMs) allow visual input to Large Language Models. VLMs have been developing rapidly, and their accuracy is improving rapidly. Their performance in nuclear medicine compared to state-of-the-art models, including reasoning models, is not yet clear. We evaluated state-of-the-art VLMs using problems from the past Japan Nuclear Medicine Board Examination (JNMBE) and assessed their strengths and limitations.

Methods

We collected 180 multiple-choice questions from JNMBE (2022–2024). About one-third included diagnostic images. We used eight latest VLMs. ChatGPT o1 pro, ChatGPT o1, ChatGPT o3-mini, ChatGPT-4.5, Claude 3.7, Gemini 2.0 Flash thinking, Llama 3.2, and Gemma 3 were tested. Each model answered every question three times in a deterministic setting, and the final answer was set by majority vote. Two board-certified nuclear medicine physicians independently provided reference answers, with a third expert resolving disagreements. We calculated overall accuracy with 95% confidence intervals and performed subgroup analyses by question type, content, and exam year.

Results

Overall accuracies ranged from 36.1% (Gemma 3) to 83.3% (ChatGPT o1 pro). ChatGPT o1 pro achieved the highest score (150/180, 83.3% [95% CI: 77.1–88.5%]), followed by ChatGPT o3-mini (82.8%) and ChatGPTo1 (78.9%). All models performed better on text-only questions than on image-based ones; ChatGPT o1 pro correctly answered 89.5% of text questions versus 66.0% of image questions. VLMs demonstrated limitations in handling with questions on Japanese regulations. ChatGPT 4.5 excelled in neurology-related image-based questions (76.9%). Accuracy was slightly lower from 2022 to 2024 for most models.

Conclusions

VLMs demonstrated high accuracy on the JNMBE, especially on text-based questions, but exhibited limitations with image recognition questions. These findings show that VLMs can be a good assistant for text-based questions in medical domains but have limitations when it comes to comprehensive questions that include images. Currently, VLMs cannot replace comprehensive training and expert interpretation. Because VLMs evolve rapidly and exam difficulty varies annually, these findings should be interpreted in that context.

Background and objective

In neuroinflammation, activated astrocytes, called reactive astrocytes, highly express monoamine oxidase B (MAO-B). [¹⁸F]SMBT-1 is a novel PET tracer developed for imaging neuroinflammation, with highly selective binding to MAO-B. The quantification method for [¹⁸F]SMBT-1 PET imaging has not been established, although some human studies using [¹⁸F]SMBT-1 PET imaging have already been conducted. In this study, we explored the most appropriate method for quantifying [¹⁸F]SMBT-1 PET.

Methods

Dynamic PET scanning of [¹⁸F]SMBT-1, accompanied by serial arterial blood sampling, was performed in healthy elderly subjects. With PET and blood data, the total distribution volumes (Vts) in the brain regions were calculated using a one-tissue compartment model (1TCM), a two-tissue compartment model (2TCM), and Logan graphical analysis. Standardized uptake values (SUVs) and SUV ratio-1 (SUVR-1) were determined for different time frames and reference regions.

Results

The values of the χ² criterion and Akaike's Information Criterion (AIC) in the brain regions were lower in 2TCM than in 1TCM, suggesting that 2TCM was a better model in terms of curve fitting. However, the very high coefficient of variation (%COV) for parameters such as K1, k2, k3, and k4 in 2TCM suggests that these parameters may not have been properly estimated. SUVs, especially at 50–70 and 70–90 min post-injection, were strongly correlated with Vt (r = 0.9188–0.9445, p < 0.0001). SUVR-1 at these time points, referenced to various regions, showed significant correlations with MAO-B distribution in the brain shown in a previous postmortem study (r = 0.9362–0.9399, p < 0.0001).

Conclusions

These findings suggest that SUVR-1, especially at 50–70 min and 70–90 min post-injection, reflects MAO-B distribution and is useful for quantifying [¹⁸F]SMBT-1 PET imaging, potentially enabling noninvasive assessment of neuroinflammation in the brain.

Trial registration

Japan Registry of Clinical Trials (jRCT) (jRCTs021200019). It was registered on August 25, 2020. The jRCT was approved as a member of the Primary Registry Network of the WHO ICTRP.

Backgrounds

Automated PROMISE (aPROMISE), which is an artificial intelligence-supported software for prostate-specific membrane antigen (PSMA) PET/CT based on PROMISE V2, has demonstrated diagnostic utility with better correspondence rates compared to manual diagnosis. However, previous studies have consistently utilized ¹⁸F-PSMA PET/CT. Therefore, we investigated the diagnostic utility of aPROMISE using both ¹⁸F- and ⁶⁸ Ga-PSMA PET/CT of Japanese patients with metastatic prostate cancer (mPCa).

Materials and methods

We retrospectively evaluated 21 PSMA PET/CT images (⁶⁸ Ga-PSMA PET/CT: n = 12, ¹⁸F-PSMA PET/CT: n = 9) from 21 patients with mPCa. A single, well-experienced nuclear radiologist performed manual diagnosis following PROMISE V2 and subsequently performed aPROMISE-assisted diagnosis to assess miTNM and details of metastatic sites. We compared the diagnostic time and correspondence rates of miTNM diagnosis between manual and aPROMISE-assisted diagnoses. Additionally, we investigated the differences in diagnostic performance between the two radioisotopes.

Results

aPROMISE-assisted diagnosis was significantly associated with shorter median diagnostic time compared to manual diagnosis (427 s [IQR: 370–834] vs. 1,114 s [IQR: 922–1291], p < 0.001). The time reduction with aPROMISE-assisted diagnosis was particularly notable when using ⁶⁸ Ga-PSMA PET/CT. aPROMISE had high diagnostic accuracy with 100% sensitivity for miT, M1a, and M1b stages. Notably, for M1b stages, aPROMISE achieved 100% sensitivity and specificity, regardless of the type of radioisotope used. However, aPROMISE was misinterpreted in lymph node detection in some cases and missed five visceral metastases (2 adrenal and 3 liver), resulting in lower sensitivity for miM1c stage (63%). In addition to detecting metastatic sites, aPROMISE successfully provided detailed metrics, including the number of metastatic lesions, total metastatic volume, and SUV mean.

Conclusions

Despite the preliminary nature of the study, aPROMISE-assisted diagnosis significantly reduces diagnostic time and achieves satisfactory accuracy compared to manual diagnosis. While aPROMISE is effective in detecting bone metastases, its limitations in identifying lymph node and visceral metastases must be carefully addressed. This study supports the utility of aPROMISE in Japanese patients with mPCa and underscores the need for further validation in larger cohorts.

Purpose

Accurate differentiation of Parkinsonism subtypes—including Parkinson’s disease (PD), multiple system atrophy (MSA), and progressive supranuclear palsy (PSP)—is essential for clinical prognosis and treatment planning. However, this remains a major challenge due to overlapping symptomatology and high inter-individual variability in cerebral glucose metabolism patterns observed on fluorodeoxyglucose positron emission tomography (FDG-PET).

Methods

To address these challenges, we propose PETFormer-SCL, a clinically informed deep learning framework that integrates convolutional neural networks (CNNs) with a channel-wise Transformer module, guided by supervised contrastive learning (SCL). This architecture is designed to enhance disease-specific feature learning while mitigating individual variability.

Results

Trained on 945 patients and evaluated on an independent test cohort of 330 patients (1275 in total), PETFormer-SCL achieved AUCs of 0.9830, 0.9702, and 0.9565 for MSA, PD, and PSP, respectively. In addition, class activation maps (CAMs) highlighted key disease-related brain regions—including the cerebellum, midbrain, and basal ganglia—demonstrating strong alignment with known pathophysiological findings.

Conclusions

PETFormer-SCL not only achieves high diagnostic accuracy, particularly for subtypes with overlapping phenotypes, but also enhances interpretability. These results support its potential as a reliable clinical decision-support tool for the early and differential diagnosis of Parkinsonism.

Objective

This systematic review aims to assess the diagnostic performance of FDG PET/CT and FAPi PET/CT in patients with gastric carcinoma, specifically for the evaluation of primary tumors, metastatic lymph nodes, and metastatic lesions.

Methods

Following PRISMA guidelines, relevant databases were searched until January 20, 2023. Studies reporting histopathology or surgical outcomes as the reference standard were included. Pooled estimates of diagnostic accuracy were generated using meta-analysis.

Results

Six studies with 167 patients who underwent FDG PET/CT and 169 patients who underwent FAPi PET/CT were included. For the detection of primary gastric carcinoma, FDG PET/CT demonstrated a pooled sensitivity of 0.86 (95% CI 0.47–0.98) and specificity of 0.70 (95% CI 0.39–0.90). The pooled positive likelihood ratio was 2.9 (95% CI 1.0–8.4), and the negative likelihood ratio was 0.21 (95% CI 0.03–1.30). The diagnostic odds ratio was 14 (95% CI 1–224), and the area under the SROC curve was 0.82. For FAPi PET/CT, pooled sensitivity and specificity for detecting primary gastric carcinoma were 0.90 (95% CI 0.90–0.90) and 0.50 (95% CI 0.50–0.50), respectively. The pooled positive and negative likelihood ratios were 1.8 (95% CI 1.8–1.8) and 0.20 (95% CI 0.20–0.20), respectively. The diagnostic odds ratio was 9 (95% CI 9–9), and the area under the SROC curve was 0.54.

Conclusion

FAPi PET/CT demonstrated comparable diagnostic performance to FDG PET/CT in the diagnosis of primary gastric carcinoma, lymph nodal metastases, and metastatic lesions. When compared to histopathology or surgical findings, FAPi PET/CT showed good sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio.