This study aimed to develop and validate deep learning models using [¹⁸F]FDG PET/CT to predict PD-L1 status in esophageal cancer (EC) patients. Additionally, we assessed the potential of derived deep learning model scores (DLS) for survival stratification in immunotherapy. In this retrospective study, we included 331 EC patients from two centers, dividing them into training, internal validation, and external validation cohorts. Fifty patients who received immunotherapy were followed up. We developed four 3D ResNet10-based models-PET + CT + clinical factors (CPC), PET + CT (PC), PET (P), and CT (C)-using pre-treatment [¹⁸F]FDG PET/CT scans. For comparison, we also constructed a logistic model incorporating clinical factors (clinical model). The DLS were evaluated as radiological markers for survival stratification, and nomograms for predicting survival were constructed. The models demonstrated accurate prediction of PD-L1 status. The areas under the curve (AUCs) for predicting PD-L1 status were as follows: CPC (0.927), PC (0.904), P (0.886), C (0.934), and the clinical model (0.603) in the training cohort; CPC (0.882), PC (0.848), P (0.770), C (0.745), and the clinical model (0.524) in the internal validation cohort; and CPC (0.843), PC (0.806), P (0.759), C (0.667), and the clinical model (0.671) in the external validation cohort. The CPC and PC models exhibited superior predictive performance. Survival analysis revealed that the DLS from most models effectively stratified overall survival and progression-free survival at appropriate cut-off points (P < 0.05), outperforming stratification based on PD-L1 status (combined positive score ≥ 10). Furthermore, incorporating model scores with clinical factors in nomograms enhanced the predictive probability of survival after immunotherapy. Deep learning models based on [¹⁸F]FDG PET/CT can accurately predict PD-L1 status in esophageal cancer patients. The derived DLS can effectively stratify survival outcomes following immunotherapy, particularly when combined with clinical factors.

Tumor-infiltrating lymphocytes (TILs) are key immune biomarkers associated with prognosis and treatment response in early-stage breast cancer (BC), particularly in the triple-negative subtype. This study aimed to evaluate whether [18F]FDG PET/CT imaging and routine cell blood count (CBC)-derived biomarkers can serve as non-invasive surrogates for TILs, using machine-learning models. We retrospectively analyzed 358 patients with biopsy-proven early-stage invasive BC who underwent pre-treatment [18F]FDG PET/CT imaging. PET-derived biomarkers were extracted from the primary tumor, lymph nodes, and lymphoid organs (spleen and bone marrow). CBC-derived biomarkers included neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR). TILs were assessed histologically and categorized as low (0-10%), intermediate (11-59%), or high (≥ 60%). Correlations were assessed using Spearman's rank coefficient, and classification and regression models were built using several machine-learning algorithms. Tumor SUVmax and tumor SUVmean showed the highest correlation with TIL levels (ρ = 0.29 and 0.30 respectively, p < 0.001 for both), but overall associations between TILs and PET or CBC-derived biomarkers were weak. No CBC-derived biomarker showed significant correlation or discriminative performance. Machine-learning models failed to predict TIL levels with satisfactory accuracy (maximum balanced accuracy = 0.66). Lymphoid organ metrics (SLR, BLR) and CBC-derived parameters did not significantly enhance predictive value. In this study, neither [18F]FDG PET/CT nor routine CBC-derived biomarkers reliably predict TILs levels in early-stage BC. This observation was made in presence of potential scanner-related variability and for a restricted set of usual PET metrics. Future models should incorporate more targeted imaging approaches, such as immunoPET, to non-invasively assess immune infiltration with higher specificity and improve personalized treatment strategies.

Positron emission tomography-computed tomography (PET-CT) is recommended for response evaluation in aggressive large B-cell lymphoma (LBCL) but cannot detect minimal residual disease (MRD). Circulating tumor DNA (ctDNA) has emerged as a promising biomarker for real-time disease monitoring. This study evaluated longitudinal ctDNA monitoring as an MRD marker in LBCL. In this prospective, single-center study, 14 newly diagnosed LBCL patients receiving first-line immunochemotherapy underwent frequent longitudinal blood sampling. A 53-gene targeted sequencing panel quantified ctDNA and evaluated its kinetics, correlating it with clinical parameters and PET-CT, including total metabolic tumor volume (TMTV) calculated using AI-based analysis via RECOMIA. Baseline ctDNA was detected in 11 out of 14 patients (79%), with a median variant allele frequency of 6.88% (interquartile range: 1.19-10.20%). ctDNA levels correlated significantly with TMTV (ρ = 0.91, p < 0.0001) and lactate dehydrogenase. Circulating tumor DNA kinetics, including after one treatment cycle, mirrored PET-CT metabolic changes and identified relapsing or refractory cases. This study demonstrates ctDNA-based MRD monitoring in LBCL using a fixed targeted assay with an analytical sensitivity of at least 10-3. The kinetics of ctDNA reflects the clinical course and PET-CT findings, underscoring its complementary potential to PET-CT.

In the last decade, immunotherapy, particularly immune checkpoint inhibitors, has revolutionized cancer treatment and improved prognosis. However, severe checkpoint inhibitor-induced liver injury (CHILI), which can lead to treatment discontinuation or death, occurs in up to 18% of the patients. The aim of this study is to evaluate the value of PET/CT radiomics analysis for the detection of CHILI. Patients with CHILI grade 2 or higher who underwent liver function tests and liver biopsy were retrospectively included. Minors, patients with cognitive impairments, and patients with viral infections were excluded from the study. The patients' liver and spleen were contoured on the anonymized PET/CT imaging data, followed by radiomics feature extraction. Principal component analysis (PCA) and Bonferroni corrections were used for statistical analysis and exploration of radiomics features related to CHILI. Sixteen patients were included and 110 radiomics features were extracted from PET images. Liver PCA-5 showed significance as well as one associated feature but did not remain significant after Bonferroni correction. Spleen PCA-5 differed significantly between CHILI and non-CHILI patients even after Bonferroni correction, possibly linked to the higher metabolic function of the spleen in autoimmune diseases due to the recruitment of immune cells. This pilot study identified statistically significant differences in PET-derived radiomics features of the spleen and observable changes in the liver on PET/CT scans before and after the onset of CHILI. Identifying these features could aid in diagnosing or predicting CHILI, potentially enabling personalized treatment. Larger multicenter prospective studies are needed to confirm these findings and develop automated detection methods.

Though ultra-low dose (ULD) imaging offers notable benefits, its widespread clinical adoption faces challenges. Long-axial field-of-view (LAFOV) PET/CT systems are expensive and scarce, while artificial intelligence (AI) shows great potential but remains largely limited to specific systems and is not yet widely used in clinical practice. However, integrating AI techniques and technological advancements into ULD imaging is helping bridge the gap between standard axial field-of-view (SAFOV) and LAFOV PET/CT systems. This paper offers an initial evaluation of ULD capabilities using one of the latest SAFOV PET/CT device. A patient injected with 16.4 MBq ¹⁸F-FDG underwent a local protocol consisting of a dynamic acquisition (first 30 min) of the abdominal section and a static whole body 74 min post-injection on a GE Omni PET/CT. From the acquired images we computed the dosimetry and compared clinical output from kidney function and brain uptake to kidney model and normal databases, respectively. The effective PET dose for this patient was 0.27 ± 0.01 mSv and the absorbed doses were 0.56 mGy, 0.89 mGy and 0.20 mGy, respectively to the brain, heart, and kidneys. The recorded kidney concentration closely followed the kidney model, matching the increase and decrease in activity concentration over time. Normal values for the z-score were observed for the brain uptake, indicating typical brain function and activity patterns consistent with healthy individuals. The signal to noise ration obtained in this study (13.1) was comparable to the LAFOV reported values. This study shows promising capabilities of ultra-low-dose imaging in SAFOV PET devices, previously deemed unattainable with SAFOV PET imaging.

The integration of artificial intelligence (AI) into [¹⁸F]FDG PET/CT imaging continues to expand, offering new opportunities for more precise, consistent, and personalized oncologic evaluations. Building on the foundation established in Part I, this second part explores AI-driven innovations across a broader range of malignancies, including hematological, genitourinary, melanoma, and central nervous system tumors as well applications of AI in pediatric oncology. Radiomics and machine learning algorithms are being explored for their ability to enhance diagnostic accuracy, reduce interobserver variability, and inform complex clinical decision-making, such as identifying patients with refractory lymphoma, assessing pseudoprogression in melanoma, or predicting brain metastases in extracranial malignancies. Additionally, AI-assisted lesion segmentation, quantitative feature extraction, and heterogeneity analysis are contributing to improved prediction of treatment response and long-term survival outcomes. Despite encouraging results, variability in imaging protocols, segmentation methods, and validation strategies across studies continues to challenge reproducibility and remains a barrier to clinical translation. This review evaluates recent advancements of AI, its current clinical applications, and emphasizes the need for robust standardization and prospective validation to ensure the reproducibility and generalizability of AI tools in PET imaging and clinical practice.

Glioma, the most common primary intracranial tumor, poses significant challenges to precision diagnosis and treatment due to its heterogeneity and invasiveness. With the introduction of the 2021 WHO classification standard based on molecular biomarkers, the role of imaging in non-invasive subtyping and therapeutic monitoring of gliomas has become increasingly crucial. While conventional MRI shows limitations in assessing metabolic status and differentiating tumor recurrence, positron emission tomography (PET) combined with radiomics and artificial intelligence technologies offers a novel paradigm for precise diagnosis and treatment monitoring through quantitative extraction of multimodal imaging features (e.g., intensity, texture, dynamic parameters). This review systematically summarizes the technical workflow of PET radiomics (including tracer selection, image segmentation, feature extraction, and model construction) and its applications in predicting molecular subtypes (such as IDH mutation and MGMT methylation), distinguishing recurrence from treatment-related changes, and prognostic stratification. Studies demonstrate that amino acid tracers (e.g., ¹⁸F-FET, ¹¹C-MET) combined with multimodal radiomics models significantly outperform traditional parametric analysis in diagnostic efficacy. Nevertheless, current research still faces challenges including data heterogeneity, insufficient model interpretability, and lack of clinical validation. Future advancements require multicenter standardized protocols, open-source algorithm frameworks, and multi-omics integration to facilitate the transformative clinical translation of PET radiomics from research to practice.

Brain tumor segmentation plays a crucial role in clinical diagnostics and treatment planning, yet accurate and efficient segmentation remains a significant challenge due to complex tumor structures and variations in imaging modalities. Multi-feature selection and region classification depend on continuous homogeneous features to improve the precision of tumor detection. This classification is required to suppress the discreteness across various extraction rates to consent to the smallest segmentation region that is infected. This study proposes a Finite Segmentation Model (FSM) with Improved Classifier Learning (ICL) to enhance segmentation accuracy in Positron Emission Tomography (PET) images. The FSM-ICL framework integrates advanced textural feature extraction, deep learning-based classification, and an adaptive segmentation approach to differentiate between tumor and non-tumor regions with high precision. Our model is trained and validated on the Synthetic Whole-Head Brain Tumor Segmentation Dataset, consisting of 1000 training and 426 testing images, achieving a segmentation accuracy of 92.57%, significantly outperforming existing approaches such as NRAN (62.16%), DSSE-V-Net (71.47%), and DenseUNet+ (83.93%). Furthermore, FSM-ICL enhances classification precision to 95.59%, reduces classification error to 5.67%, and minimizes classification time to 572.39 ms, demonstrating a 10.09% improvement in precision and a 10.96% boost in classification rates over state-of-the-art methods. The hybrid classifier learning approach effectively addresses segmentation discreteness, ensuring continuous and discrete tumor region detection with superior feature differentiation. This work has significant implications for automated tumor detection, personalized treatment strategies, and AI-driven medical imaging advancements. Future directions include incorporating micro-segmentation and pre-classification techniques to further optimize performance in dense pixel-packed datasets.

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). 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. 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. 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.

Accurate identification of bone marrow invasion (BMI) is critical for determining the prognosis of and treatment strategies for lymphoma. Although bone marrow biopsy (BMB) is the current gold standard, its invasive nature and sampling errors highlight the necessity for noninvasive alternatives. We aimed to develop and validate an interpretable machine learning model that integrates clinical data, ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT) parameters, radiomic features, and deep learning features to predict BMI in lymphoma patients. We included 159 newly diagnosed lymphoma patients (118 from Center I and 41 from Center II), excluding those with prior treatments, incomplete data, or under 18 years of age. Data from Center I were randomly allocated to training (n = 94) and internal test (n = 24) sets; Center II served as an external validation set (n = 41). Clinical parameters, PET/CT features, radiomic characteristics, and deep learning features were comprehensively analyzed and integrated into machine learning models. Model interpretability was elucidated via Shapley Additive exPlanations (SHAPs). Additionally, a comparative diagnostic study evaluated reader performance with and without model assistance. BMI was confirmed in 70 (44%) patients. The key clinical predictors included B symptoms and platelet count. Among the tested models, the ExtraTrees classifier achieved the best performance. For external validation, the combined model (clinical + PET/CT + radiomics + deep learning) achieved an area under the receiver operating characteristic curve (AUC) of 0.886, outperforming models that use only clinical (AUC 0.798), radiomic (AUC 0.708), or deep learning features (AUC 0.662). SHAP analysis revealed that PET radiomic features (especially PET_lbp_3D_m1_glcm_DependenceEntropy), platelet count, and B symptoms were significant predictors of BMI. Model assistance significantly enhanced junior reader performance (AUC improved from 0.663 to 0.818, p = 0.03) and improved senior reader accuracy, although not significantly (AUC 0.768 to 0.867, p = 0.10). Our interpretable machine learning model, which integrates clinical, imaging, radiomic, and deep learning features, demonstrated robust BMI prediction performance and notably enhanced physician diagnostic accuracy. These findings underscore the clinical potential of interpretable AI to complement medical expertise and potentially reduce the reliance on invasive BMB for lymphoma staging.