XGBoost-based lung cancer risk prediction model shows greater accuracy than logistic regression in a large screening cohort.
Key Details
- 1Study analyzed data from 11,708 participants in the Guangzhou Lung-Care Project.
- 2Data split into 70% training and 30% validation cohorts.
- 3XGBoost machine learning algorithm achieved an AUC of 0.658, compared to 0.647 for logistic regression.
- 4Childhood exposure to cooking fuels was found to significantly affect lung cancer risk.
- 5Study addresses a gap in machine learning-based lung cancer risk prediction research in China.
Why It Matters
Machine learning models like XGBoost can provide more accurate and robust risk predictions for lung cancer screening, potentially improving early identification and targeting of high-risk individuals. This enhances the effectiveness of radiology-led screening programs and supports broader adoption of AI-driven risk assessments.
Source
AuntMinnie
Related News
•AuntMinnie
Study Reveals Women's Willingness to Pay for AI in Mammography
Awareness of AI accuracy, error rates, and advertising affect women's out-of-pocket willingness to pay for AI-supported mammography.
•Radiology Business
GE HealthCare’s True Definition DL Receives FDA Clearance for CT Imaging
GE HealthCare's True Definition DL, a deep learning-based CT image reconstruction tool, has received FDA 510(k) clearance.
•AuntMinnie
Deep Learning Chest X-ray Aging Estimates Predict Mortality Risk
Chest x-ray-based biologic aging and aging velocity estimated by deep learning are linked to all-cause and disease-specific mortality.