Radiogenomic analysis for predicting lymph node metastasis and molecular annotation of radiomic features in pancreatic cancer DOI Creative Commons
Yi Tang,

Yi-xi Su,

Jin-mei Zheng

et al.

Journal of Translational Medicine, Journal Year: 2024, Volume and Issue: 22(1)

Published: July 29, 2024

Abstract Background To provide a preoperative prediction model for lymph node metastasis in pancreatic cancer patients and molecular information of key radiomic features. Methods Two cohorts comprising 151 54 were included the analysis. Radiomic features from tumor region interests extracted by using PyRadiomics software. We used framework that incorporated 10 machine learning algorithms generated 77 combinations to construct radiomics-based models prediction. Weighted gene coexpression network analysis (WGCNA) was subsequently performed determine relationships between expression levels Molecular pathways enrichment uncover underlying Results Patients in-house cohort (mean age, 61.3 years ± 9.6 [SD]; 91 men [60%]) separated into training ( n = 105, 70%) validation 46, 30%) cohorts. A total 1,239 subjected algorithms. The showed moderate performance predicting metastasis, combination StepGBM Enet had best (AUC 0.84, 95% CI 0.77–0.91) 0.85, 0.73–0.98) determined 15 core variables metastasis. Proliferation-related processes may respond main alterations these Conclusions Machine learning-based radiomics could predict status cancer, which is associated with proliferation-related alterations.

Language: Английский

Application of radiomics for preoperative prediction of lymph node metastasis in colorectal cancer: A systematic review and Meta-analysis DOI Creative Commons
Elahe Abbaspour, Sahand Karimzadhagh, Abbas Monsef

et al.

International Journal of Surgery, Journal Year: 2024, Volume and Issue: unknown

Published: March 11, 2024

Background: Colorectal cancer (CRC) stands as the third most prevalent globally, projecting 3.2 million new cases and 1.6 deaths by 2040. Accurate lymph node metastasis (LNM) detection is critical for determining optimal surgical approaches, including preoperative neoadjuvant chemoradiotherapy surgery, which significantly influence CRC prognosis. However, conventional imaging lacks adequate precision, prompting exploration into radiomics, addresses this shortfall converting medical images reproducible, quantitative data. Methods: Following PRISMA, Supplemental Digital Content 1, http://links.lww.com/JS9/C77, 2, http://links.lww.com/JS9/C78 AMSTAR-2 guidelines, 3, http://links.lww.com/JS9/C79, we systematically searched PubMed, Web of Science, Embase, Cochrane Library, Google Scholar databases until January 11, 2024, to evaluate radiomics models’ diagnostic precision in predicting LNM patients. The quality bias risk included studies were assessed using Radiomics Quality Score (RQS) modified Assessment Diagnostic Accuracy Studies (QUADAS-2) tool. Subsequently, statistical analyses conducted. Results: Thirty-six encompassing 8,039 patients included, with a significant concentration 2022-2023 (20/36). models demonstrated pooled area under curve (AUC) 0.814 (95% CI: 0.78-0.85), featuring sensitivity specificity 0.77 0.69, 0.84) 0.73 0.67, 0.78), respectively. Subgroup revealed similar AUCs CT MRI-based models, rectal outperformed colon colorectal cancers. Additionally, utilizing cross-validation, 2D segmentation, internal validation, manual prospective design, single-center populations tended have higher AUCs. these differences not statistically significant. Radiologists collectively achieved AUC 0.659 0.627, 0.691), differing from performance ( P < 0.001). Conclusion: Artificial intelligence-based shows promise staging CRC, exhibiting predictive performance. These findings support integration clinical practice enhance strategies management.

Language: Английский

Citations

20
Pancreatic Adenocarcinoma: Imaging Modalities and the Role of Artificial Intelligence in Analyzing CT and MRI Images DOI Creative Commons

Cristian Anghel,

Mugur Grasu,

D. Anghel

et al.

Diagnostics, Journal Year: 2024, Volume and Issue: 14(4), P. 438 - 438

Published: Feb. 16, 2024

Pancreatic ductal adenocarcinoma (PDAC) stands out as the predominant malignant neoplasm affecting pancreas, characterized by a poor prognosis, in most cases patients being diagnosed nonresectable stage. Image-based artificial intelligence (AI) models implemented tumor detection, segmentation, and classification could improve diagnosis with better treatment options increased survival. This review included papers published last five years describes current trends AI algorithms used PDAC. We analyzed applications of detection PDAC, segmentation lesion, differential diagnosis, histopathological genomic prediction. The results show lack multi-institutional collaboration stresses need for bigger datasets order to be clinically relevant manner.

Language: Английский

Citations

6
Radiomics and deep learning models for CT pre-operative lymph node staging in pancreatic ductal adenocarcinoma: A systematic review and meta-analysis DOI
Roberto Castellana, Salvatore Claudio Fanni, Claudia Roncella

et al.

European Journal of Radiology, Journal Year: 2024, Volume and Issue: 176, P. 111510 - 111510

Published: May 18, 2024

Language: Английский

Citations

5
Contrast-enhanced MRI-based intratumoral heterogeneity assessment for predicting lymph node metastasis in resectable pancreatic ductal adenocarcinoma DOI Creative Commons

Junjian Shen,

Qin Li, Lei Li

et al.

Insights into Imaging, Journal Year: 2025, Volume and Issue: 16(1)

Published: March 30, 2025

Abstract Objectives To develop and validate a contrast-enhanced MRI-based intratumoral heterogeneity (ITH) model for predicting lymph node (LN) metastasis in resectable pancreatic ductal adenocarcinoma (PDAC). Methods Lesions were encoded into different habitats based on enhancement ratios at arterial, venous, delayed phases of MRI. Habitat models enhanced ratio mapping single sequences, radiomic models, clinical developed evaluating LN metastasis. The performance the was evaluated via metrics. Additionally, patients stratified high-risk low-risk groups an ensembled to assess prognosis after adjuvant therapy. Results We radiomics–habitat–clinical (RHC) that integrates radiomics, habitat, data precise prediction PDAC. RHC showed strong predictive performance, with area under curve (AUC) values 0.805, 0.779, 0.615 derivation, internal validation, external validation cohorts, respectively. Using optimal threshold 0.46, effectively patients, revealing significant differences recurrence-free survival overall (OS) ( p = 0.004 < 0.001). Adjuvant therapy improved OS group 0.004), but no benefit observed 0.069). Conclusion ITH provides reliable estimates PDAC may offer additional value guiding decision-making. Critical relevance statement This ensemble facilitates preoperative using offers foundation prognostic assessment supports management personalized treatment strategies. Key Points habitat can predict Both radiomics characteristics useful have potential enhance accuracy inform therapeutic decisions. Graphical

Language: Английский

Citations

0
Radiomics diagnostic performance for predicting lymph node metastasis in esophageal cancer: a systematic review and meta-analysis DOI Creative Commons

Dong Ma,

Teli Zhou,

Jing Chen

et al.

BMC Medical Imaging, Journal Year: 2024, Volume and Issue: 24(1)

Published: June 12, 2024

Esophageal cancer, a global health concern, impacts predominantly men, particularly in Eastern Asia. Lymph node metastasis (LNM) significantly influences prognosis, and current imaging methods exhibit limitations accurate detection. The integration of radiomics, an artificial intelligence (AI) driven approach medical imaging, offers transformative potential. This meta-analysis evaluates existing evidence on the accuracy radiomics models for predicting LNM esophageal cancer.

Language: Английский

Citations

3
Pancreatic ductal adenocarcinoma staging: A narrative review of radiologic techniques and advances DOI Creative Commons
Linda C. Chu, Elliot K. Fishman

International Journal of Surgery, Journal Year: 2023, Volume and Issue: unknown

Published: Dec. 4, 2023

Radiology plays an important role in the initial diagnosis and staging of patients with pancreatic ductal adenocarcinoma (PDAC). CT is preferred modality over MRI, due to wider availability, greater consistency image quality, lower cost. MRI PET/CT are usually reserved as problem-solving tools select patients. The National Comprehensive Cancer Network (NCCN) guidelines define resectability criteria based on tumor involvement arteries veins, triage into resectable, borderline locally advanced, metastatic categories. Patients resectable disease eligible for upfront surgical resection, while high-stage treated neoadjuvant chemotherapy and/or radiation therapy hopes downstaging disease. accuracy critically depends imaging technique experience radiologists. Several challenges accurate preoperative include prediction lymph node metastases, detection subtle liver peritoneal restaging following therapy. Artificial intelligence (AI) has potential function "second readers" improve upon radiologists' small early-stage tumors, which can shift more toward resection potentially curable cancer. AI may also provide biomarkers that predict recurrence patient survival after assist selection most likely benefit from surgery thus improving outcomes.

Language: Английский

Citations

8
Radiogenomic analysis for predicting lymph node metastasis and molecular annotation of radiomic features in pancreatic cancer DOI Creative Commons
Yi Tang,

Yi-xi Su,

Jin-mei Zheng

et al.

Journal of Translational Medicine, Journal Year: 2024, Volume and Issue: 22(1)

Published: July 29, 2024

Abstract Background To provide a preoperative prediction model for lymph node metastasis in pancreatic cancer patients and molecular information of key radiomic features. Methods Two cohorts comprising 151 54 were included the analysis. Radiomic features from tumor region interests extracted by using PyRadiomics software. We used framework that incorporated 10 machine learning algorithms generated 77 combinations to construct radiomics-based models prediction. Weighted gene coexpression network analysis (WGCNA) was subsequently performed determine relationships between expression levels Molecular pathways enrichment uncover underlying Results Patients in-house cohort (mean age, 61.3 years ± 9.6 [SD]; 91 men [60%]) separated into training ( n = 105, 70%) validation 46, 30%) cohorts. A total 1,239 subjected algorithms. The showed moderate performance predicting metastasis, combination StepGBM Enet had best (AUC 0.84, 95% CI 0.77–0.91) 0.85, 0.73–0.98) determined 15 core variables metastasis. Proliferation-related processes may respond main alterations these Conclusions Machine learning-based radiomics could predict status cancer, which is associated with proliferation-related alterations.

Language: Английский

Citations

1