Detection of Moderate Hepatic Steatosis on Portal Venous Phase Contrast-Enhanced CT: Evaluation Using an Automated Artificial Intelligence Tool.

Researchers

Journal

Modalities

Models

Abstract

Background: Pre-contrast CT is established to evaluate for hepatic steatosis; post-contrast CT has historically been limited for this purpose. Objective: To evaluate the diagnostic performance of portal venous phase post-contrast CT in detecting at least moderate hepatic steatosis using liver and spleen attenuation measurements determined by an automated artificial intelligence (AI) tool. Methods: This retrospective study included 2917 patients (mean age, 56.8 years; 1381 men, 1536 women) who underwent a CT examination including at least two series through the liver. Examinations were obtained from an AI vendor’s data lake from 24 centers from one U.S. healthcare network and 29 centers from one Israeli healthcare network. An automated deep learning tool extracted liver and spleen attenuation measurements. Reference for at least moderate steatosis was pre-contrast liver attenuation <40.0 HU (i.e., estimated liver fat >15%). A radiologist manually reviewed examinations with outlier AI results to confirm portal venous timing and identify issues impacting attenuation measurements. Results: After outlier review, analysis included 2777 patients with portal venous phase images. Prevalence of at least moderate steatosis was 13.9% (387/2777). Patients without and with at least moderate steatosis had mean post-contrast liver attenuation of 104.5±18.1 HU and 67.1±18.6 HU (p<.001), mean post-contrast liver-spleen attenuation difference of -7.6±16.4 HU and -31.8±20.3 HU (p<.001), and mean liver enhancement of 49.3±15.9 HU vs 38.6±13.6 HU (p<.001). Diagnostic performance for at least moderate steatosis was higher for post-contrast liver attenuation (AUC=0.938) than post-contrast liver-spleen attenuation difference (AUC=0.832) (p<.001). Post-contrast liver attenuation had sensitivity and specificity for at least moderate steatosis of 77.8% and 93.2% at <80 HU and 90.5% and 78.4% at <90 HU; post-contrast liver-spleen attenuation difference had sensitivity and specificity of 71.4% and 79.3% at <-20 HU, and 87.0% and 62.1% at <-10 HU. Conclusion: Post-contrast liver attenuation outperformed post-contrast liver-spleen attenuation difference for detecting at least moderate steatosis in a heterogeneous patient sample, evaluated using an automated AI tool. Splenic attenuation is likely not needed to assess for at least moderate steatosis on post-contrast images, Clinical Impact: The technique could promote early detection of clinically significant non-alcoholic fatty liver disease through individualized or large-scale opportunistic evaluation.