The deep learning architecture used for this segmentation. Different models have different strengths.
—
Model
—
Time taken for the model to process the image and produce the segmentation. Lower is faster.
—
Inference Time
milliseconds
Average probability the model assigns to pixels it identifies as lung. Higher confidence means the model is more certain about its segmentation boundaries.
—
Confidence
mean probability
Percentage of the total image area identified as lung tissue. For a well-positioned PA chest X-ray, expect 20-40%.
—
Lung Coverage
% of image
Ratio of smaller to larger lung field area (0-1). Values > 0.85 indicate symmetric lungs. Low symmetry may suggest rotation, effusion, or volume loss.
—
Symmetry
Dimensions of the original uploaded image in pixels.
—
Image Size
pixels