Artificial intelligence for tumor [18F]FDG-PET imaging: Advancement and future trends—part I

Seminars in Nuclear Medicine, Journal Year: 2025, Volume and Issue: unknown

Published: March 1, 2025

The advent of sophisticated image analysis techniques has facilitated the extraction increasingly complex data, such as radiomic features, from various imaging modalities, including [18F]FDG PET/CT, a well-established cornerstone oncological imaging. Furthermore, use artificial intelligence (AI) algorithms shown considerable promise in enhancing interpretation these quantitative parameters. Additionally, AI-driven models enable integration parameters multiple modalities along with clinical facilitating development comprehensive significant impact. However, challenges remain regarding standardization and validation AI-powered models, well their implementation real-world practice. variability acquisition protocols, segmentation methods, feature approaches across different institutions necessitates robust harmonization efforts to ensure reproducibility utility. Moreover, successful translation AI into practice requires prospective large cohorts, seamless existing workflows assess ability enhance clinicians' performance. This review aims provide an overview literature highlight three key applications: diagnostic impact, prediction treatment response, long-term patient prognostication. In first part, we will focus on head neck, lung, breast, gastroesophageal, colorectal, gynecological malignancies.

Language: Английский

Synergic value of 3D CT-derived body composition and triglyceride glucose body mass for survival prognostic modeling in unresectable pancreatic cancer

Frontiers in Nutrition, Journal Year: 2025, Volume and Issue: 12

Published: March 19, 2025

Background Personalized and accurate survival risk prognostication remains a significant challenge in advanced pancreatic ductal adenocarcinoma (PDAC), despite extensive research on prognostic predictive markers. Patients with PDAC are prone to muscle loss, fat consumption, malnutrition, which is associated inferior outcomes. This study investigated the use of three-dimensional (3D) anthropometric parameters derived from computed tomography (CT) scans triglyceride glucose-body mass index (TyG-BMI) relation overall (OS) outcomes patients. Additionally, model for 1 year OS was developed based body components hematological indicators. Methods A retrospective analysis conducted 303 patients locally or synchronous metastases undergoing first-line chemotherapy, all whom had undergone pretreatment abdomen-pelvis CT scans. Automatic 3D measurements subcutaneous visceral volume, skeletal density (SMD) were assessed at L3 vertebral level by an artificial intelligence assisted diagnosis system (HY Medical). Various indicators including TyG-BMI, nutritional [geriatric (GNRI) prealbumin], inflammation [(C-reactive protein (CRP) neutrophil lymphocyte ratio (NLR)] also recorded. All underwent follow-up least dynamic nomogram personalized prediction constructed. Results We included 211 [mean (standard deviation) age, 63.4 ± 11.2 years; 89 women (42.2) %)]. Factors such as low (SMI) ( P = 0.011), high adipose tissue area (VSR) < 0.001), (VFI) TyG-BMI 0.004), prealbumin 0.001) identified independent factors OS. The under curve established 0.846 calibration showed good consistency. High-risk (> 211.9 points calculated using nomogram) significantly reduced rates. Conclusion In this study, proposed (with web-based tool) enabled individualized could help guide risk-adapted treatment unresectable metastases.

Language: Английский

18F-FDG PET/CT-Based Deep Radiomic Models for Enhancing Chemotherapy Response Prediction in Breast Cancer

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: May 15, 2025

Background Enhancing the accuracy of tumor response predictions enables development tailored therapeutic strategies for patients with breast cancer. In this study, we developed deep-radiomic models to enhance prediction chemotherapy after first treatment cycle. Methods 18F- Fludeoxyglucose (FDG) PET/CT imaging data and clinical record from 60 cancer were retrospectively obtained Cancer Imaging Archive (QIN-BREAST). scans conducted at three distinct stages treatment; prior initiation (T1), following cycle (T2), full regimen (T3). The patient's primary gross volume (GTV) was delineated on PET images using a 40% threshold maximum standardized uptake value (SUVmax). Radiomic features extracted GTV based images. addition, Squeeze-and-Excitation Network (SENet) deep learning model employed generate additional combined analysis. A XGBoost machine compared conventional (ML) algorithm (random forest [RF], logistic regression [LR] support vector [SVM]). performance each assessed receiver operating characteristics area under curve (ROC AUC) analysis, in validation cohort. Model evaluated through 5-fold cross-validation entire cohort, splits stratified by categories ensure balanced representation. Results AUC values only radiomic 0.85(XGBoost), 0.76 (RF), 0.80 (LR), 0.59 (SVM), showing best performance. After incorporating learning-derived SENet, increased 0.92, 0.88, 0.90, 0.61, respectively, demonstrating significant improvements predictive accuracy. Predictions pre-treatment (T1) post-first-cycle (T2) data, enabling early assessment initial Conclusion Integrating significantly enhanced patients. This study demonstrated superior capability model, emphasizing its potential optimize personalized accurately identifying unlikely respond

Language: Английский