Skip to main content

Table 3 A summary of the evaluations performed to assess the goodness of automatic HU conversion

From: A fully automated deep learning pipeline for micro-CT-imaging-based densitometry of lung fibrosis murine models

HU conversion evaluations

Raw Grey levels detected

Structure

Absolute percentage error

Airways

4.3% ± 1.9%

Heart

0.1% ± 0.1%

Transfer function

Parameter

Absolute percentage error

Slope

3.6% ± 0.2%

Intercept

3.4% ± 0.3%

Lung densitometry

Compartmental volume

Percent absolute error

Normo-aerated volume

0.5% ± 5.6%

Hypo-aerated volume

2.9% ± 4.8%

Non-aerated volume

0.4% ± 6.7%

Back to article page

Respiratory Research

ISSN: 1465-993X

Contact us