Clinical validation of an AI-based automatic quantification tool for lung lobes in SPECT/CT

Abstract

Background: Lung lobar ventilation and perfusion (V/Q) quantification is generally obtained by generating planar scintigraphy images and then imposing three equally sized regions of interest on the data of each lung. This method is fast but not as accurate as SPECT/CT imaging, which provides three-dimensional data and therefore allows more precise lobar quantification. However, the manual delineation of each lobe is time-consuming, which makes SPECT/CT incompatible with the clinical workflow for V/Q estimation. An alternative may be to use artificial intelligence-based auto-segmentation tools such as AutoLung3D (Siemens Healthineers, Knoxville, USA), which automatically delineate the lung lobes on the CT data acquired with the SPECT data. The present study assessed the clinical validity of this approach relative to planar scintigraphy and manual quantification in SPECT/CT.

Methods: The Autolung3D software was tested on the retrospective SPECT/CT data of 43 patients who underwent V/Q scintigraphy with ^99mTc-macroaggregated albumin and ^99mTc-labeled aerosol. It was compared to planar scintigraphy and SPECT/CT using the manual quantification method in terms of relative lobar V/Q quantification values and interobserver variability.

Results: The three methods provided similar V/Q estimates for the left lung lobes and total lungs. However, compared to the manual SPECT/CT method, planar scintigraphy yielded significantly higher estimates for the middle right lobe and significantly lower estimates for the superior and inferior right lobes. The estimates of the manual and automated SPECT/CT methods were similar. However, the post-processing time in the automated method was approximately 5 min compared to 2 h for the manual method. Moreover, the automated method associated with a drastic reduction in interobserver variability: Its maximal relative standard deviation was only 5%, compared to 23% for planar scintigraphy and 19% for the manual SPECT/CT method.

Conclusions: This study validated the AutoLung3D software for general clinical use since it rapidly provides accurate lobar quantification in V/Q scans with markedly less interobserver variability than planar scintigraphy or the manual SPECT/CT method.

Keywords: AI-based segmentation; Lobar quantification; Perfusion SPECT/CT; Ventilation SPECT/CT.

Fig. 1

Lung segmentation on the planar scintigraphy data for perfusion and ventilation. Each lung was divided into three regions of interest by using a template that imposes three rectangular equally sized areas. Representative data of a single patient are shown

Fig. 2

Automated lobar segmentation on the perfusion SPECT/CT dataset using the AI-based algorithm AutoLung3D. Representative data of a single patient (the same patient in Fig. 1) are shown

Fig. 3

Manual lobar segmentation on the perfusion SPECT/CT dataset imported in Eclipse. The segmentation was conducted by an experienced nuclear physician. Representative data of a single patient (the same patient in Figs. 1 and 2) are shown

Fig. 4

Comparison of the three quantification methods in terms of relative ^99mTc-labeled aerosol distribution (%) during lung ventilation analysis. *p < 0.05 for the indicated lobe, as determined by Welch’s test-based ANOVA. Significant differences between the methods can be seen by viewing the 95% confidence intervals: If two methods do not show overlap of these bars, they differ significantly in relative distribution

Fig. 5

Comparison of the three quantification methods in terms of relative ^99mTc-macroaggregated albumin distribution (%) during lung perfusion analysis. *p < 0.05 for the indicated lobe, as determined by Welch’s test-based ANOVA. Significant differences between the methods can be seen by viewing the 95% confidence intervals: If two methods do not show overlap of these bars, they differ significantly in relative distribution

Fig. 6

Interobserver variability of the three methods. Ventilation quantifications were performed by three experienced nuclear physicians with planar, manual, and automated segmentation in ten patients. The data are expressed as average relative standard deviation (%) of the relative distribution in each lobe/lung. The gray error bars indicate the standard deviation of the ten patients

Fig. 7

Interobserver variability of the three methods. Perfusion quantifications were performed by three experienced nuclear physicians with planar, manual, and automated segmentation in ten patients. The data are expressed as average relative standard deviation (%) of the relative distribution in each lobe/lung. The gray error bars indicate the standard deviation of the ten patients