Multiparametric MRI for Assessment of the Biological Invasiveness and Prognosis of Pancreatic Ductal Adenocarcinoma in the Era of Artificial Intelligence

Journal of Magnetic Resonance Imaging, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 9, 2025

Pancreatic ductal adenocarcinoma (PDAC) is the deadliest malignant tumor, with a grim 5‐year overall survival rate of about 12%. As its incidence and mortality rates rise, it likely to become second‐leading cause cancer‐related death. The radiological assessment determined stage management PDAC. However, highly heterogeneous disease complexity tumor microenvironment, challenging adequately reflect biological aggressiveness prognosis accurately through morphological evaluation alone. With dramatic development artificial intelligence (AI), multiparametric magnetic resonance imaging (mpMRI) using specific contrast media special techniques can provide functional information high image quality powerful tool in quantifying intratumor characteristics. Besides, AI has been widespread field medical analysis. Radiomics high‐throughput mining quantitative features from that enables data be extracted applied for better decision support. Deep learning subset neural network algorithms automatically learn feature representations data. AI‐enabled biomarkers mpMRI have enormous promise bridge gap between personalized medicine demonstrate huge advantages predicting characteristics current AI‐based models PDAC operate mainly realm single modality relatively small sample size, technical reproducibility interpretation present barrage new potential challenges. In future, integration multi‐omics data, such as radiomics genomics, alongside establishment standardized analytical frameworks will opportunities increase robustness interpretability bring these closer clinical practice. Evidence Level 3 Technical Efficacy Stage 4

Language: Английский

---

Role of Multiphasic Computed Tomography in the Evaluation of Neoplastic Pancreatic Masses: A Single-Center Observational Study

Cureus, Journal Year: 2025, Volume and Issue: unknown

Published: May 6, 2025

Language: Английский

---

Detection and characterization of pancreatic lesion with artificial intelligence: The SFR 2023 artificial intelligence data challenge

Diagnostic and Interventional Imaging, Journal Year: 2024, Volume and Issue: 105(10), P. 395 - 399

Published: July 23, 2024

Language: Английский

---

Optimized Spatial Transformer for Segmenting Pancreas Abnormalities

Deleted Journal, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 4, 2024

The precise delineation of the pancreas from clinical images poses a substantial obstacle in realm medical image analysis and surgical procedures. Challenges arise complexities complications practice related to pancreas. To tackle these challenges, novel approach called Spatial Horned Lizard Attention Approach (SHLAM) has been developed. As result, preprocessing function developed examine eliminate noise barriers trained MRI data. Furthermore, an assessment current attributes is conducted, followed by identification essential elements for forecasting impacted region. Once affected region identified, undergo segmentation. it crucial emphasize that present study assigns 80% data training 20% testing purposes. optimal parameters were assessed based on precision, accuracy, recall, F-measure, error rate, Dice, Jaccard. performance improvement demonstrated validating method various existing models. SHLAM proposed accuracy rate 99.6%, surpassing all alternative methods.

Language: Английский

---

Deep Machine Learning for Automatic Segmentation of the Pancreatic Parenchyma and its hypo- and hypervascular lesions on CT Images

Medical Visualization, Journal Year: 2024, Volume and Issue: 28(3), P. 12 - 21

Published: Aug. 7, 2024

Objective of the study. To develop and evaluate effectiveness a technology for segmenting pancreatic parenchyma its hyper- hypovascular lesions on abdominal computed tomography (CT) scans using deep machine learning. Materials methods. CT from database A.V. Vishnevsky National Medical Research Center Surgery were used training testing algorithms – total number approximately 150 studies (arterial venous phases). A test dataset 46 anonymized phases) was prepared validation obtained algorithms, independently assessed by expert physicians. The primary segmentation neural network is nn-UNet (M. Antonelli et al., 2022). Results . average accuracy model determining masks pancreas images had an AUC 0.8 phase 0.85 arterial phase. formations 0.6. Conclusion. Automated structure learning technologies demonstrated high accuracy. However, hypo- hypervascular requires improvement. overlap showed rather low result, but in all cases, location pathological formation correctly identified algorithm. Enhancing algorithm could increase No false negative results when detecting formations; INS detected “suspicious” areas parenchyma. This can help reduce omission pathologies scans, their further assessment be carried out radiologist himself.

Language: Английский

---

Accuracy of machine learning models for pre-diagnosis and diagnosis of pancreatic ductal adenocarcinoma in contrast-CT images: a systematic review and meta-analysis

Abdominal Radiology, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 25, 2024

Language: Английский

---