AI-powered spectral CT analysis for clinical decision support in carotid vulnerable plaque detection: a deep learning approach.
6 affiliations
- 1Department of Radiology, Xiong'an Xuanwu Hospital, Rongcheng, Hebei, 071702, China.
- 2Imaging Department, The Second Affiliated Hospital of Qiqihar Medical College, Qiqihar, 161000, China.
- 3Neuroelectrophysiology Department, The Second Affiliated Hospital of Qiqihar Medical College, 37 Zhonghua West Road, Jianhua District, Qiqihar, Heilongjiang Province, 161000, China.
- 4Department of Neurology, The Second Affiliated Hospital of Qiqihar Medical College, Qiqihar, 161000, China.
- 5Department of General Medicine, The Second Affiliated Hospital of Qiqihar Medical College, Qiqihar, 161000, China.
- 6Neuroelectrophysiology Department, The Second Affiliated Hospital of Qiqihar Medical College, 37 Zhonghua West Road, Jianhua District, Qiqihar, Heilongjiang Province, 161000, China. [email protected].
Abstract
Background
Carotid vulnerable plaques (CVPs) represent a major cause of ischemic stroke, yet current diagnostic methods lack sufficient precision for early detection. Spectral computed tomography (CT) enables detailed plaque characterization, but its clinical utility depends on advanced analytical techniques.Methods
A Fourier Transform-enhanced U-Net model was developed using spectral CT data from 214 patients at Qiqihar Medical College. Transfer learning was applied for carotid segmentation, supported by extensive data augmentation and ten-fold cross-validation. Model performance was assessed using sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC). A patient-level independent test set was reserved to evaluate model performance.Results
The FTC-enhanced U-Net achieved a Dice similarity coefficient (DSC) of 0.89 and an intersection over union (IoU) of 0.80, accurately delineating arterial and plaque boundaries. Stable classification performance was maintained across plaque risk levels, with accuracies of 86% for low-risk, 83% for medium-risk, and 87% for high-risk plaques. Ten-fold cross-validation yielded an AUC of 0.88 ± 0.03, an accuracy of 85.3% ± 2.1%, and an F1-score of 0.84 ± 0.02, confirming strong robustness and generalizability.Conclusions
Artificial intelligence (AI)-based predictive models derived from spectral CT imaging show substantial potential for early diagnosis and prognostic assessment of CVPs. The FTC-enhanced U-Net provides an efficient and reliable tool for clinical decision support, facilitating broader adoption of AI in vascular imaging and improving diagnostic precision and patient management.
Related
- Reliable Detection, Unreliable Clinical Decision-making: Artificial Intelligence Models in Burn Assessment
- Development and usability testing of a machine learning clinical decision support tool for suggesting etiologies and treatment of clinical deterioration
- Dual-domain self-supervised feature alignment via spectral–spatial representation learning for deepfake anomaly detection
- Machine Learning-Augmented Precision Dosing: Integrating Pharmacokinetic Modeling With Clinical Decision Support.
- Machine Learning and Deep Learning for Early Disease Detection Across Modalities: A Comprehensive Review of Methods, Applications, and Clinical Translation
BibTeX
@article{ni2026aipowered,
title = {{AI-powered spectral CT analysis for clinical decision support in carotid vulnerable plaque detection: a deep learning approach.}},
author = {Yunzhe Ni and Tianyu Zhang and Zonghui Huang and Yuguang Wang and Guochao Han and Yuan Lin and Bo Liu and Wennan Lin and Hui Li},
journal = {BMC medical informatics and decision making},
year = {2026},
month = jul,
doi = {10.1186/s12911-026-03611-0},
url = {https://europepmc.org/article/MED/42897593},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - AI-powered spectral CT analysis for clinical decision support in carotid vulnerable plaque detection: a deep learning approach.
AU - Yunzhe Ni
AU - Tianyu Zhang
AU - Zonghui Huang
AU - Yuguang Wang
AU - Guochao Han
AU - Yuan Lin
AU - Bo Liu
AU - Wennan Lin
AU - Hui Li
PY - 2026
DA - 2026/07/01
JO - BMC medical informatics and decision making
DO - 10.1186/s12911-026-03611-0
UR - https://europepmc.org/article/MED/42897593
AB - Background Carotid vulnerable plaques (CVPs) represent a major cause of ischemic stroke, yet current diagnostic methods lack sufficient precision for early detection. Spectral computed tomography (CT) enables detailed plaque characterization, but its clinical utility depends on advanced analytical techniques.Methods A Fourier Transform-enhanced U-Net model was developed using spectral CT data from 214 patients at Qiqihar Medical College. Transfer learning was applied for carotid segmentation, supported by extensive data augmentation and ten-fold cross-validation. Model performance was assessed using sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC). A patient-level independent test set was reserved to evaluate model performance.Results The FTC-enhanced U-Net achieved a Dice similarity coefficient (DSC) of 0.89 and an intersection over union (IoU) of 0.80, accurately delineating arterial and plaque boundaries. Stable classification performance was maintained across plaque risk levels, with accuracies of 86% for low-risk, 83% for medium-risk, and 87% for high-risk plaques. Ten-fold cross-validation yielded an AUC of 0.88 ± 0.03, an accuracy of 85.3% ± 2.1%, and an F1-score of 0.84 ± 0.02, confirming strong robustness and generalizability.Conclusions Artificial intelligence (AI)-based predictive models derived from spectral CT imaging show substantial potential for early diagnosis and prognostic assessment of CVPs. The FTC-enhanced U-Net provides an efficient and reliable tool for clinical decision support, facilitating broader adoption of AI in vascular imaging and improving diagnostic precision and patient management.
SN - 1472-6947
ER -