Application of Artificial Intelligence in Oncologic Molecular PET-Imaging: A Narrative Review on Beyond [18F]F-FDG Tracers - Part I. PSMA, Choline, and DOTA Radiotracers.

Artificial intelligence (AI) has evolved significantly in the past few decades. This thriving trend has also been seen in medicine in recent years, particularly in the field of imaging. Machine learning (ML), deep learning (DL), and their methods (eg, SVM, CNN), as well as radiomics, are the terminologies that have been introduced to this field and, to some extent, become familiar to the expert clinicians. PET is one of the modalities that has been enhanced via these state-of-the-art algorithms. This robust imaging technique further merged with anatomical modalities, such as computed tomography (CT) and magnetic resonance imaging (MRI), to provide reliable hybrid modalities, PET/CT and PET/MRI. Applying AI-based algorithms on the different components (PET, CT, and MRI) has resulted in promising results, maximizing the value of PET imaging. However, [18F]F-FDG, the most commonly utilized tracer in molecular imaging, has been mainly in the spotlight. Thus, we aimed to look into the less discussed tracers in this review, moving beyond [18F]F-FDG. The novel non-[18F]F-FDG agents also showed to be valuable in various clinical tasks, including lesion detection and tumor characterization, accurate delineation, and prognostic impact. Regarding prostate patients, PSMA-based models were highly accurate in determining tumoral lesions' location and delineating them, particularly within the prostate gland. However, they also could assess whole-body images to detect extra-prostatic lesions in a patient automatically. Considering the prognostic value of prostate-specific membrane antigen (PSMA) PET using AI, it could predict response to treatment and patient survival, which are crucial in patient management. Choline imaging, another non-[18F]F-FDG tracer, similarly showed acceptable results that may be of benefit in the clinic, though the current evidence is significantly more limited than PSMA. Lastly, different subtypes of DOTA ligands were found to be valuable. They could diagnose tumoral lesions in challenging sites and even predict histopathology grade, being a highly advantageous noninvasive tool. In conclusion, the current limited investigations have shown promising results, leading us to a bright future for AI in molecular imaging beyond [18F]F-FDG.

Application of Artificial Intelligence in Oncologic Molecular PET-Imaging: A Narrative Review on Beyond [18F]F-FDG Tracers Part II. [18F]F-FLT, [18F]F-FET, [11C]C-MET and Other Less-Commonly Used Radiotracers.

Following the previous part of the narrative review on artificial intelligence (AI) applications in positron emission tomography (PET) using tracers rather than 18F-fluorodeoxyglucose ([18F]F-FDG), in this part we review the impact of PET-derived radiomics data on the diagnostic performance of other PET radiotracers, 18F-O-(2-fluoroethyl)-L-tyrosine ([18F]F-FET), 18F-Fluorothymidine ([18F]F-FLT) and 11C-Methionine ([11C]C-MET). [18F]F-FET-PET, using an artificial amino acid taken up into upregulated tumoral cells, showed potential in lesion detection and tumor characterization, especially with its ability to reflect glioma heterogeneity. [18F]F-FET-PET-derived textural features appeared to have the potential to reveal considerable information for accurate delineation for guiding biopsy and treatment, differentiate between low-grade and high-grade glioma and related wild-type genotypes, and distinguish pseudoprogression from true progression. In addition, models built using clinical parameters and [18F]F-FET-PET-derived radiomics features showed acceptable results for survival stratification of glioblastoma patients. [18F]F-FLT-PET-based characteristics also showed potential in evaluating glioma patients, correlating with Ki-67 and patient prognosis. AI-based PET-volumetry using this radiotracer as a proliferation marker also revealed promising preliminary results in terms of guide-targeting bone marrow-preserving adaptive radiation therapy. Similar to [18F]F-FET, the other amino acid tracer which reflects cellular proliferation, [11C]C-MET, has also shown acceptable performance in predicting tumor grade, distinguishing brain tumor recurrence from radiation necrosis, and treatment monitoring by PET-derived radiomics models. In addition, PET-derived radiomics features of various radiotracers such as [18F]F-DOPA, [18F]F-FACBC, [18F]F-NaF, [68Ga]Ga-CXCR-4 and [18F]F-FMISO may also provide useful information for tumor characterization and predict of disease outcome. In conclusion, AI using tracers beyond [18F]F-FDG could improve the diagnostic performance of PET-imaging for specific indications and help clinicians in their daily routine by providing features that are often not detectable by the naked eye.

Automatic detection and segmentation of lesions in 18 F-FDG PET/CT imaging of patients with Hodgkin lymphoma using 3D dense U-Net.

OBJECTIVE: The accuracy of automatic tumor segmentation in PET/computed tomography (PET/CT) images is crucial for the effective treatment and monitoring of Hodgkin lymphoma. This study aims to address the challenges faced by certain segmentation algorithms in accurately differentiating lymphoma from normal organ uptakes due to PET image resolution and tumor heterogeneity. MATERIALS AND METHODS: Variants of the encoder-decoder architectures are state-of-the-art models for image segmentation. Among these kinds of architectures, U-Net is one of the most famous and predominant for medical image segmentation. In this study, we propose a fully automatic approach for Hodgkin lymphoma segmentation that combines U-Net and DenseNet architectures to reduce network loss for very small lesions, which is trained using the Tversky loss function. The hypothesis is that the fusion of these two deep learning models can improve the accuracy and robustness of Hodgkin lymphoma segmentation. A dataset with 141 samples was used to train our proposed network. Also, to test and evaluate the proposed network, we allocated two separate datasets of 20 samples. RESULTS: We achieved 0.759 as the mean Dice similarity coefficient with a median value of 0.767, and interquartile range (0.647-0.837). A good agreement was observed between the ground truth of test images against the predicted volume with precision and recall scores of 0.798 and 0.763, respectively. CONCLUSION: This study demonstrates that the integration of U-Net and DenseNet architectures, along with the Tversky loss function, can significantly enhance the accuracy of Hodgkin lymphoma segmentation in PET/CT images compared to similar studies.

Pre-treatment 68 Ga-PSMA-11 PET/CT Prognostic Value in Predicting Response to 177Lu-PSMA-I&T Therapy and Patient Survival.

PURPOSE: To assess the prognostic value of pre-treatment [68Ga]Ga-PSMA-11 PET/CT and other baseline clinical characteristics in predicting prostate cancer (PCa) patients response to [177Lu]Lu-PSMA (PSMA-I&T), as well as patient survival. PROCEDURES: In this retrospective study, 81 patients who received [177Lu]Lu-PSMA-I&T between October 2018 and January 2023 were reviewed. Eligible patients had metastatic castration-resistant PCa, underwent pre-treatment [68Ga]Ga-PSMA-11 PET/CT, and had serum prostate-specific antigen (PSA) levels available. On PET/CT images, SUVmax, SULmax, SUVpeak, and SULpeak of the most-avid tumoral lesion, as well as SUVmean of the parotid gland (P-SUVmean) and liver (L-SUVmean), were measured. Also, whole-body PSMA tumour volume (PSMA-TV) and total lesion PSMA (TL-PSMA) were calculated. To interpret treatment response after [177Lu]Lu-PSMA-I&T, a composite of PSA values and [68Ga]Ga-PSMA-11 PET/CT findings were considered. The outcomes were dichotomised into progressive versus controlled (stable disease or partial response) disease. Then, the association of baseline parameters with patient response was evaluated. Also, survival analyses were performed to assess baseline parameters in predicting overall survival. RESULTS: Sixty patients (age:73 ± 8, PSA:185 ± 371) were included. Patients received at least one cycle of [177Lu]Lu-PSMA therapy (median = 4). Overall, half of the patients showed disease progression. In the progressive versus controlled disease evaluation, the highest SULmax, as well as SUVmax and SULmax to both backgrounds (L-SUVmean and P-SUVmean), were significantly correlated with the outcome (p-values < 0.05). In the multivariate analysis, only SULmax to the L-SUVmean remained significant (p-value = 0.038). The best cut-off was 8 (AUC = 0.71). With a median follow-up of 360 days, 11 mortal events were documented. In the multivariate survival analysis, only SULmax to P-SUVmean (cut-off = 2.4; p-value = 0.043) retained significance (hazard ratio = 4.0). CONCLUSIONS: A greater level of PSMA uptake, specifically higher tumour-to-background uptake in the hottest lesion, may hold substantial prognostic significance, considering both [177Lu]Lu-PSMA-I&T response and patient survival. These ratios may have the potential to be used for PCa patient selection for radioligand therapy.

Impact of [18F]FDG PET/CT in the Assessment of Immunotherapy-Induced Arterial Wall Inflammation in Melanoma Patients Receiving Immune Checkpoint Inhibitors.

We aimed to investigate the role of [18F]FDG positron emission tomography/computed tomography (PET/CT) in the early detection of arterial wall inflammation (AWI) in melanoma patients receiving immune checkpoint inhibitors (ICIs). Our retrospective study enrolled 95 melanoma patients who had received ICIs. Inclusion criteria were ICI therapy for at least six months and at least three [18F]FDG PET/CTs, including one pretreatment session plus two scans three and six months after treatment initiation. AWI was assessed using quantitative and qualitative methods in the subclavian artery, thoracic aorta, and abdominal aorta. We found three patients with AWI visual suspicion in the baseline scan, which increased to five in the second and twelve in the third session. Most of these patients' treatments were terminated due to either immune-related adverse events (irAEs) or disease progression. In the overall population, the ratio of arterial-wall maximum standardized uptake value (SUVmax)/liver-SUVmax was significantly higher three months after treatment than the pretreatment scan in the thoracic aorta (0.83 ± 0.12 vs. 0.79 ± 0.10; p-value = 0.01) and subclavian artery (0.67 ± 0.13 vs. 0.63 ± 0.12; p-value = 0.01), and it remained steady in the six-month follow-up. None of our patients were diagnosed with definite clinical vasculitis on the dermatology follow-up reports. To conclude, our study showed [18F]FDG PET/CT's potential to visualise immunotherapy-induced subclinical inflammation in large vessels. This may lead to more accurate prediction of irAEs and better patient management.

18F-FDG PET/CT prognostic role in predicting response to salvage therapy in relapsed/refractory Hodgkin's lymphoma.

PURPOSE: To evaluate the response predictors, both clinical and 18F-FDG PET/CT parameters, in Hodgkin's lymphoma (HL) patients diagnosed with refractory/relapsed disease who were planned to receive salvage therapy. METHODS: In this retrospective study, all HL patients referred to our center between March 2015 and July 2021 were reviewed. Patients with refractory/relapsed disease who were candidates for salvage therapy were included. 18F-FDG PET/CT measurements at the time of diagnosis were extracted as the predictors, and the lesions' response at the end of the salvage therapy was considered the outcomes. The Kaplan-Meier method and multiple Cox regression were utilized to find the significant parameters to predict the time to reach the complete response. The statistical significance level was set at a two-sided p-value <0.05. RESULTS: A total of 303 tumoral lesions from 64 patients were included. Regarding the factors associated with the response, B symptoms (p-value < 0.01), pathologic subtype (p-value < 0.001), and patient stage (p-value < 0.01) were the significant clinical parameters. In addition, SUVmax (p-value = 0.03), SUVmax/hepatic background SUVmax (p-value = 0.04), SUVmean (in all thresholds; 41% p-value = 0.02, 51% p-value = 0.04, 61% p-value = 0.01), and MTV (in all thresholds; 41% p-value = 0.04, 51% p-value = 0.04, 61% p-value = 0.05) were the significant parameters in the 18F-FDG PET/CT scans. At the median follow-up of 9 months, we found that pathologic subtype (p-value < 0.01), patient stage (p-value = 0.03), SUVmax (p-value = 0.02), SUVmax/hepatic background SUVmax (p-value = 0.03), SUVmean (in all thresholds; 41% p-value = 0.01, 51% p-value = 0.02, 61% p-value = 0.02), and MTV ≥ 41% (p-value = 0.02) were significant predictive factors. Multiple Cox regression showed the pathologic subtype (p-value = 0.02), SUVmax (p-value = 0.02), and MTV ≥ 41% (p-value = 0.04) were the most significant predictors. CONCLUSION: Our study demonstrated that by knowing the histopathology of the lesions, the pre-treatment 18F-FDG PET/CT might be able to predict response after salvage therapy in the relapsed/refractory HL.