AUTHOR=Zhou Xi Yang , Tang Chun Xiang , Guo Ying Kun , Tao Xin Wei , Chen Wen Cui , Guo Jin Zhou , Ren Gui Sheng , Li Xiao , Luo Song , Li Jun Hao , Huang Wei Wei , Lu Guang Ming , Zhang Long Jiang , Huang Xiang Hua , Wang Yi Ning , Yang Gui Fen 

TITLE=Diagnosis of Cardiac Amyloidosis Using a Radiomics Approach Applied to Late Gadolinium-Enhanced Cardiac Magnetic Resonance Images: A Retrospective, Multicohort, Diagnostic Study

JOURNAL=Frontiers in Cardiovascular Medicine

VOLUME=Volume 9 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/cardiovascular-medicine/articles/10.3389/fcvm.2022.818957

DOI=10.3389/fcvm.2022.818957

ISSN=2297-055X

ABSTRACT=Objectives: To assess the potential of a radiomics approach of late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR) in the diagnosis of cardiac amyloidosis (CA). 
Materials and Methods: This retrospective study included 200 patients with biopsy-proven light-chain amyloidosis. CA was diagnosed based on confirmed systemic amyloidosis with evidence of cardiac involvement by imaging and clinical biomarkers. One hundred thirty-nine patients (54±8 years, 75 [54%] men) in our institution were divided into training cohort (n=97, mean age of 53±8 years,54 [56%] men) and internal validation cohort (n=42, mean age of 56±8 years, 21 [50%] men) with ratio of 7:3, while 61 patients (mean age of 60±9 years,42 [69%]men) from other two institutions were enrolled for external validation. Radiomics features were extracted from global (from base to apex) left ventricular (LV) myocardium and three different segments (basal, mid-ventricular and apex) on short-axis LGE images.Boruta algorithm was used to select the radiomic features. The model was built using XGBoost algorithm. Qualitative and semi-quantitative assessment of LGE images were conducted by two readers, while quantitative assessment was measured using a semi-automatical CMR software. The diagnostic performance of the radiomics and other parameters were compared by receiver operating characteristic (ROC) curve analysis. Correlation between radiomics and the degree of myocardial involvement by amyloidosis was investigated
Results: A total of 1906 radiomics features were extracted for each LV section, respectively. No statistical significance was indicated between any two slices to diagnose CA and the highest area under the curve (AUC) was found in basal section (0.92 [95% CI, 0.86–0.97]  in the training set, 0.89 [95% CI, 0.79–1.00] in the internal validation set, 0.92 [95% CI, 0.85–0.99] in the external validation set), which was superior to qualitative and quantitative LGE assessment. Moderate correlations were reported between global or basal radiomics scores and Mayo stage in all patients (Spearman’s rho=0.61, 0.62; all p<0.01).
Conclusion: Radiomics analysis of LGE images provides incremental information compared with visual assessment and quantitative parameters on CMR to diagnose CA. Radiomics was moderately correlated with the severity of CA. Further studies are needed to assess the prognostic significance of radiomics in patients with CA.