Enterprise AI Analysis
CRESTOMICS: ANALYZING CAROTID PLAQUES IN THE CREST-2 TRIAL WITH A NEW ADDITIVE CLASSIFICATION MODEL
Authors: Pranav Kulkarni¹,², Brajesh K. Lal¹³, Georges Jreij³, Sai Vallamchetla⁴, Langford Green³, Jenifer Voeks, John Huston⁴, Lloyd Edwards⁶, George Howard⁷, Bradley A. Maron¹³, Thomas G. Brott⁴, James F. Meschia⁵, Florence X. Doo¹,³*, Heng Huang¹²*
Abstract: Accurate characterization of carotid plaques is critical for stroke prevention in patients with carotid stenosis. We analyze 500 plaques from CREST-2, a multi-center clinical trial, to identify radiomics-based markers from B-mode ultrasound images linked with high-risk. We propose a new kernel-based additive model, combining coherence loss with group-sparse regularization for nonlinear classification. Group-wise additive effects of each feature group are visualized using partial dependence plots. Results indicate our method accurately and interpretably assesses plaques, revealing a strong association between plaque texture and clinical risk.
Executive Impact: Pioneering Predictive Analytics in Healthcare
This research introduces a novel AI model that significantly enhances the ability to predict high-risk carotid plaques, a critical factor in stroke prevention. By leveraging advanced radiomics, enterprises can implement more precise diagnostic tools, leading to improved patient outcomes and substantial operational efficiencies.
0.95
Model AUROC Performance
97.20%
Predictive Accuracy Achieved
88.11%
F1 Score for Robustness
1
Improved Stroke Risk Prediction
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
AI for Enhanced Medical Diagnosis
This category highlights the application of advanced machine learning techniques, specifically radiomics, to extract nuanced information from medical images for improved diagnostic accuracy. The study showcases how AI can provide highly interpretable insights into complex biological processes, such as plaque vulnerability in carotid arteries.
0.95
Model AUROC Performance
Our new kernel-based additive model achieved a superior AUROC of 0.95, indicating excellent discriminatory power for identifying high-risk carotid plaques, a critical factor in stroke prevention.
Enterprise Process Flow
Data Extraction (500 plaques from CREST-2)
→
Radiomics Feature Extraction (102 features)
→
Feature Selection (Top-10 via ElasticNet)
→
Model Training (Kernel-based additive model)
→
Clinical Risk Assessment (Predicting Gray-Weale types I & II)
→
Interpretability Analysis (Partial dependence plots & group contributions)
| Model | AUROC | Accuracy | F1 Score |
|---|---|---|---|
| Proposed Model | 0.95 | 97.20% | 88.11% |
| GaussianSVM (Best Baseline) | 0.94 | 97.20% | 86.55% |
| XGBoost | 0.93 | 96.40% | 82.18% |
| Logistic Regression | 0.93 | 96.00% | 79.52% |
| Linear SVM (L1 Penalty) | 0.90 | 96.00% | 79.52% |
Case Study: High-risk Plaque Identification in CREST-2
Challenge: Accurately identifying high-risk carotid plaques from ultrasound images is crucial for stroke prevention but complex due to subtle variations.
Solution: Researchers applied a novel kernel-based additive model to 500 carotid plaques from the CREST-2 trial, extracting 102 radiomic features to analyze plaque texture, shape, and composition.
Impact: The model achieved an AUROC of 0.95, significantly outperforming traditional methods. It revealed strong associations between GLCM texture features and clinical risk, providing a highly interpretable assessment of plaque vulnerability. This can lead to earlier, more precise interventions and improved patient outcomes.
Calculate Your Potential ROI with AI
Discover the tangible benefits of integrating advanced AI solutions like the one in this study into your enterprise. Use our interactive calculator to estimate your potential annual savings and reclaimed operational hours.
Potential Annual Savings
$0
Annual Hours Reclaimed
0
Your AI Implementation Roadmap
Embark on a structured journey to integrate cutting-edge AI into your operations. Our phased approach ensures seamless adoption and measurable success.
Discovery & Strategy
We begin with a deep dive into your current workflows, identifying key pain points and opportunities for AI leverage. This phase defines project scope, objectives, and success metrics.
Data Preparation & Model Development
Our experts collect and prepare your data, then design and train custom AI models based on the validated techniques, like the advanced additive model from this research.
Integration & Deployment
The developed AI solutions are seamlessly integrated into your existing systems, ensuring minimal disruption and maximum compatibility. Rigorous testing is performed to guarantee performance.
Monitoring & Optimization
Post-deployment, we continuously monitor model performance, gathering feedback and implementing iterative improvements to ensure sustained value and adapt to evolving needs.
Ready to Transform Your Enterprise with AI?
The insights from this research demonstrate the profound impact of specialized AI models in complex fields. Connect with our AI strategists to explore how similar innovations can drive efficiency, accuracy, and competitive advantage in your organization.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat