A 'habitat' AI model outperforms standard 2D approaches in stratifying lung adenocarcinoma risk in subsolid nodules on low-dose CT scans.
Key Details
- 1The study evaluated a 'habitat' AI model, 2D model, radiomic model, and a combined model for classifying invasiveness and grade of lung adenocarcinoma presenting as subsolid nodules on LDCT.
- 2747 patients with 834 resected lung adenocarcinomas were included, split into training, internal, and external test sets.
- 3On the external test set, the macro-average AUCs were: 2D model 0.87, habitat model 0.92, radiomic model 0.92, and combined model 0.93.
- 4Habitat imaging quantifies spatial heterogeneity by segmenting nodules into subregions based on characteristics like signal intensity.
- 5Habitat and radiomic models both significantly outperformed the traditional 2D approach.
Why It Matters
Habitat AI models offer a novel, more accurate, and interpretable tool for noninvasive risk stratification of subsolid lung nodules, which could enhance early lung cancer screening workflows and reduce interobserver variability among radiologists.
Source
AuntMinnie
Related News
•AuntMinnie
Study: Computer Vision Models Best LLMs in Chest CT Breast Abnormality Detection
Computer vision models (CVMs) surpass large language models (LLMs) in accurately labeling incidental breast abnormalities on chest CT scans.
•Cardiovascular Business
Radiology Maintains Lead in FDA-Cleared AI Algorithms, Cardiology Follows
Radiology remains the top specialty for FDA-cleared AI, with cardiology as a strong second, particularly in cardiovascular imaging.
•AuntMinnie
Deep Learning Models Rival Radiologists for Pancreatic Cancer Detection on CT
Deep-learning models achieved comparable or superior accuracy to experienced radiologists in detecting pancreatic cancer on CT scans, especially for small tumors.