Computed tomography enterography-based radiomics for assessing mucosal healing in patients with small bowel Crohn's disease DOI
Hao Ding, Yuanyuan Fang, Wenjie Fan

et al.

World Journal of Gastroenterology, Journal Year: 2024, Volume and Issue: 31(3)

Published: Dec. 17, 2024

Mucosal healing (MH) is the major therapeutic target for Crohn's disease (CD). As most commonly involved intestinal segment, small bowel (SB) assessment crucial CD patients. Yet, it poses a significant challenge due to its limited accessibility through conventional endoscopic methods. To establish noninvasive radiomic model based on computed tomography enterography (CTE) MH in SBCD Seventy-three patients diagnosed with were included and divided into training cohort (n = 55) test 18). Radiomic features obtained from CTE images model. Patient demographics analysed clinical A radiomic-clinical nomogram was constructed by combining features. The diagnostic efficacy benefit evaluated via receiver operating characteristic (ROC) curve analysis decision (DCA), respectively. Of 73 enrolled, 25 achieved MH. had an area under ROC of 0.961 (95% confidence interval: 0.886-1.000) 0.958 (0.877-1.000) provided superior either or models alone, as demonstrated DCA. These results indicate that CTE-based promising imaging biomarker serves potential alternative enteroscopy

Language: Английский

A Nomogram Based on Laboratory Data, Inflammatory Bowel Disease Questionnaire and CT Enterography for Activity Evaluation in Crohn’s Disease DOI Creative Commons
Han Zhang, Yijun Shen, Bo Cao

et al.

Journal of Inflammation Research, Journal Year: 2025, Volume and Issue: Volume 18, P. 183 - 194

Published: Jan. 1, 2025

Accurately assessing the activity of Crohn's disease (CD) is crucial for determining prognosis and guiding treatment strategies CD patients. This study aimed to develop validate a nomogram activity. The semi-automatic segmentation method PyRadiomics software were employed segment extract radiomics features from spectral CT enterography images lesions in 107 radiomic score (rad-score) was calculated using signature formula. Multivariate logistic regression analysis identified independent risk factors erythrocyte sedimentation rate, fecal calprotectin, Inflammatory Bowel Disease Questionnaire (IBDQ), constructed combination with rad-score. underwent evaluation testing training set (n = 84) validation 23), respectively. discrimination performance combined (AUC 0.877) marginally superior that IBDQ + clinical 0.854). However, there no significant difference AUC between two models (P 0.206). outperformed 0.808), 0.746), 0.688). Significant differences observed (radiomic vs clinical, P 0.026; 0.011; combined, 0.008; set). nomogram, laboratory data, rad-score, presents an accurate reliable enhances potential personalized plans better management, making it valuable tool practice.

Language: Английский

Citations

1
AI-luminating Artificial Intelligence in Inflammatory Bowel Diseases: A Narrative Review on the Role of AI in Endoscopy, Histology, and Imaging for IBD DOI
Phillip Gu,

Oreen Mendonca,

Dan Carter

et al.

Inflammatory Bowel Diseases, Journal Year: 2024, Volume and Issue: 30(12), P. 2467 - 2485

Published: March 7, 2024

Endoscopy, histology, and cross-sectional imaging serve as fundamental pillars in the detection, monitoring, prognostication of inflammatory bowel disease (IBD). However, interpretation these studies often relies on subjective human judgment, which can lead to delays, intra- interobserver variability, potential diagnostic discrepancies. With rising incidence IBD globally coupled with exponential digitization data, there is a growing demand for innovative approaches streamline diagnosis elevate clinical decision-making. In this context, artificial intelligence (AI) technologies emerge timely solution address evolving challenges IBD. Early using deep learning radiomics endoscopy, have demonstrated promising results AI detect, diagnose, characterize, phenotype, prognosticate Nonetheless, available literature has inherent limitations knowledge gaps that need be addressed before transition into mainstream tool To better understand value integrating IBD, we review summarize our current understanding identify inform future investigations.

Language: Английский

Citations

8
Artificial Intelligence for Quantifying Cumulative Small Bowel Disease Severity on CT-Enterography in Crohn's Disease DOI
Ryan W. Stidham, Binu Enchakalody, Stewart C. Wang

et al.

The American Journal of Gastroenterology, Journal Year: 2024, Volume and Issue: 119(9), P. 1885 - 1893

Published: April 25, 2024

INTRODUCTION: Assessing the cumulative degree of bowel injury in ileal Crohn's disease (CD) is difficult. We aimed to develop machine learning (ML) methodologies for automated estimation on computed tomography-enterography (CTE) help predict future surgery. METHODS: Adults with CD using biologic therapy at a tertiary care center underwent ML analysis CTE scans. Two fellowship-trained radiologists graded severity granular spatial increments along ileum (1 cm), called mini-segments. segmentation methods were trained radiologist grading predicted and then spatially mapped ileum. Cumulative was calculated as sum (S-CIDSS) mean grades Multivariate models small resection compared metrics traditional measures, adjusting laboratory values, medications, prior surgery time CTE. RESULTS: In 229 scans, 8,424 mini-segments analysis. Agreement between strong (κ = 0.80, 95% confidence interval 0.79–0.81) similar inter-radiologist agreement 0.87, 0.85–0.88). S-CIDSS (46.6 vs 30.4, P 0.0007) grade scores (1.80 1.42, < 0.0001) greater users that went Models (area under curve 0.76) outperformed conventional medical history 0.62) predicting users. DISCUSSION: Automated show promise improving prediction outcomes CD. Beyond replicating expert judgment, enterography can augment personalization assessment

Language: Английский

Citations

6
A radiomics nomogram based on MSCT and clinical factors can stratify fibrosis in inflammatory bowel disease DOI Creative Commons
Xu Zeng, Huijie Jiang,

Yanmei Dai

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Jan. 12, 2024

Abstract Intestinal fibrosis is one of the major complications inflammatory bowel disease (IBD) and a pathological process that significantly impacts patient prognosis treatment selection. Although current imaging assessment clinical markers are widely used for diagnosis stratification fibrosis, these methods suffer from subjectivity limitations. In this study, we aim to develop radiomics diagnostic model based on multi-slice computed tomography (MSCT) factors. MSCT images relevant data were collected 218 IBD patients, large number quantitative image features extracted. Using features, constructed transformed it into user-friendly nomogram. A nomogram was developed predict in by integrating multiple The exhibited favorable discriminative ability, with an AUC 0.865 validation sets, surpassing both logistic regression (LR) (AUC = 0.821) 0.602) test set. train set, LR achieved 0.975, while had 0.735. demonstrated superior performance 0.971, suggesting its potential as valuable tool predicting improving decision-making. nomogram, incorporating factors, demonstrates promise stratifying IBD. outperforms traditional models offers personalized risk assessment. However, further addressing identified limitations necessary enhance applicability.

Language: Английский

Citations

4
Artificial Intelligence and IBD: Where are We Now and Where Will We Be in the Future? DOI
Mehwish Ahmed, Molly L. Stone, Ryan W. Stidham

et al.

Current Gastroenterology Reports, Journal Year: 2024, Volume and Issue: 26(5), P. 137 - 144

Published: Feb. 27, 2024

Language: Английский

Citations

4
Automatic Segmentation and Radiomics for Identification and Activity Assessment of CTE Lesions in Crohn’s Disease DOI
Yankun Gao, Bo Zhang,

Dehan Zhao

et al.

Inflammatory Bowel Diseases, Journal Year: 2023, Volume and Issue: unknown

Published: Nov. 27, 2023

The purpose of this article is to develop a deep learning automatic segmentation model for the Crohn's disease (CD) lesions in computed tomography enterography (CTE) images. Additionally, radiomics features extracted from segmented CD will be analyzed and multiple machine classifiers built distinguish activity.This was retrospective study with 2 sets CTE image data. Segmentation datasets were used establish nnU-Net neural network's model. classification dataset processed using obtain results extract features. most optimal then selected build 5 activity. performance evaluated Dice similarity coefficient, while classifier area under curve, sensitivity, specificity, accuracy.The had 84 examinations patients (mean age 31 ± 13 years , 60 males), 193 12 136 males). achieved coefficient 0.824 on testing set. logistic regression showed best among set, an accuracy 0.862, 0.697, 0.840, 0.759, respectively.The automated accurately segments lesions, distinguishes activity well. This method can assist radiologists promptly precisely evaluating activity.The images quickly identifying Crohn’s

Language: Английский

Citations

10
Computed tomography enterography-based radiomics nomograms to predict inflammatory activity for ileocolonic Crohn’s disease: a preliminary single-center retrospective study DOI Creative Commons

Yuping Ma,

Luanxin Zhu,

Bota Cui

et al.

BMC Medical Imaging, Journal Year: 2025, Volume and Issue: 25(1)

Published: Jan. 27, 2025

This study aims to develop and validate nomograms that utilize morphological radiomics features derived from computed tomography enterography (CTE) evaluate inflammatory activity in patients with ileocolonic Crohn's disease (CD). A total of 54 CD (237 bowel segments) clinically confirmed were retrospectively analyzed. The Simple Endoscopic Score for Disease (SES-CD) was used as a reference standard quantify the degree mucosal inflammation assess severity. We extracted training cohort create model (M-score) (Rad-score). combined nomogram generated by integrating M-score Rad-score. predictive performance each evaluated using receiver operating characteristic (ROC) curve analysis. Additionally, calibration decision analysis (DCA) employed accuracy clinical applicability testing cohort. area under ROC (AUC) nomogram, which included stenosis, comb sign, Rad-score, 0.834 [95% confidence interval (CI): 0.728–0.940] distinguishing between active remissive disease. Furthermore, created sign Rad-score achieved satisfactory AUC 0.781 (95% CI: 0.611–0.951) differentiating mild moderate-to-severe activity. DCA both nomograms' utility. Nomograms CTE-based could serve valuable tools assessing activity, thereby supporting decision-making managing CD. Keypoints. 1. Radiomics CTE predict 2. most effective predicting 3. enhanced radiologists' ability

Language: Английский

Citations

0
Development and validation of a CT-based radiomic nomogram for predicting surgical resection risk in patients with adhesive small bowel obstruction DOI Creative Commons
Zhibo Wang, Ling Zhu, Shunli Liu

et al.

BMC Medical Imaging, Journal Year: 2025, Volume and Issue: 25(1)

Published: Feb. 11, 2025

Language: Английский

Citations

0
The global research of artificial intelligence on inflammatory bowel disease: A bibliometric analysis DOI Creative Commons
Suqi Zeng,

Chenyu Dong,

Chuan Liu

et al.

Digital Health, Journal Year: 2025, Volume and Issue: 11

Published: Jan. 1, 2025

Aims This study aimed to evaluate the related research on artificial intelligence (AI) in inflammatory bowel disease (IBD) through bibliometrics analysis and identified basis, current hotspots, future development. Methods The literature was acquired from Web of Science Core Collection (WoSCC) 31 December 2024. Co-occurrence cooperation relationship (cited) authors, institutions, countries, cited journals, references, keywords were carried out CiteSpace 6.1.R6 software Online Analysis platform Literature Metrology. Meanwhile, relevant knowledge maps drawn, clustering performed. Results According WoSCC, 1919 790 184 49 countries/regions published 176 AI-related papers IBD during 1999–2024. number has increased significantly since 2019, reaching a maximum by 2023. United States had highest publications closest collaboration with other countries. showed that earliest studies focused “psychometric value” then moved “deep learning model,” “intestinal ultrasound,” “new diagnostic strategies.” Conclusion is first bibliometric summarize status visually reveal development trends hotspots application AI IBD. still its infancy, focus this field will shift improving efficiency diagnosis treatment deep techniques, big data-based treatment, prognosis prediction.

Language: Английский

Citations

0
Machine learning-based prediction of postoperative mortality risk after abdominal surgery DOI
Jihong Yuan, Yongmei Jin,

Jing-Ye Xiang

et al.

World Journal of Gastrointestinal Surgery, Journal Year: 2025, Volume and Issue: 17(4)

Published: March 29, 2025

BACKGROUND Preoperative risk assessments are vital for identifying patients at high of postoperative mortality. However, traditional scoring systems can be time consuming. We hypothesized that the use machine learning models would enable rapid and accurate to performed. AIM To assess potential algorithms develop predictive mortality after abdominal surgery. METHODS This retrospective study included 230 individuals who underwent surgery Seventh People’s Hospital Shanghai University Traditional Chinese Medicine between January 2023 December 2023. Demographic surgery-related data were collected used nomogram, decision-tree, random-forest, gradient-boosting, support vector machine, naïve Bayesian predict 30-day Models assessed using receiver operating characteristic curves compared DeLong test. RESULTS Of patients, 52 died 178 survived. developed training cohort (n = 161) validation 68). The areas under gradient-boosting tree, 0.908 [95% confidence interval (CI): 0.824-0.992], 0.874 (95%CI: 0.785-0.963), 0.928 0.869-0.987), 0.907 0.837-0.976), 0.983 0.959-1.000), 0.807 0.702-0.911), respectively. CONCLUSION Nomogram, all demonstrate strong performances prediction selected based on clinical circumstances.

Language: Английский

Citations

0