Artificial Intelligence Aids Cardiac Image Quality Assessment for Improving Precision in Strain Measurements.

OBJECTIVES: The aim of this study was to develop an artificial intelligence tool to assess echocardiographic image quality objectively. BACKGROUND: Left ventricular global longitudinal strain (LVGLS) has recently been used to monitor cancer therapeutics-related cardiac dysfunction (CTRCD) but image quality limits its reliability. METHODS: A DenseNet-121 convolutional neural network was developed for view identification from an athlete's echocardiographic dataset. To prove the concept that classification confidence (CC) can serve as a quality marker, values of longitudinal strain derived from feature tracking of cardiac magnetic resonance (CMR) imaging and strain analysis of echocardiography were compared. The CC was then applied to patients with breast cancer free from CTRCD to investigate the effects of image quality on the reliability of strain analysis. RESULTS: CC of the apical 4-chamber view (A4C) was significantly correlated with the endocardial border delineation index. CC of A4C >900 significantly predicted a <15% relative difference in longitudinal strain between CMR feature tracking and automated echocardiographic analysis. Echocardiographic studies (n =752) of 102 patients with breast cancer without CTRCD were investigated. The strain analysis showed higher parallel forms, inter-rater, and test-retest reliabilities in patients with CC of A4C >900. During sequential comparisons of automated LVGLS in individual patients, those with CC of A4C >900 had a lower false positive detection rate of CTRCD. CONCLUSIONS: CC of A4C was associated with the reliability of automated LVGLS and could also potentially be used as a filter to select comparable images from sequential echocardiographic studies in individual patients and reduce the false positive detection rate of CTRCD.

Is a timely assessment of the hematocrit necessary for cardiovascular magnetic resonance-derived extracellular volume measurements?

BACKGROUND: Cardiovascular magnetic resonance (CMR)-derived extracellular volume (ECV) requires a hematocrit (Hct) to correct contrast volume distributions in blood. However, the timely assessment of Hct can be challenging and has limited the routine clinical application of ECV. The goal of the present study was to evaluate whether ECV measurements lead to significant error if a venous Hct was unavailable on the day of CMR. METHODS: 109 patients with CMR T1 mapping and two venous Hcts (Hct0: a Hct from the day of CMR, and Hct1: a Hct from a different day) were retrospectively identified. A synthetic Hct (Hctsyn) derived from native blood T1 was also assessed. The study used two different ECV methods, (1) a conventional method in which ECV was estimated from native and postcontrast T1 maps using a region-based method, and (2) an inline method in which ECV was directly measured from inline ECV mapping. ECVs measured with Hct0, Hct1, and Hctsyn were compared for each method, and the reference ECV (ECV0) was defined using the Hct0. The error between synthetic (ECVsyn) and ECV0was analyzed for the two ECV methods. RESULTS: ECV measured using Hct1 and Hctsyn were significantly correlated with ECV0 for each method. No significant differences were observed between ECV0 and ECV measured with Hct1 (ECV1; 28.4 ± 6.6% vs. 28.3 ± 6.1%, p = 0.789) and between ECV0 and ECV calculated with Hctsyn (ECVsyn; 28.4 ± 6.6% vs. 28.2 ± 6.2%, p = 0.45) using the conventional method. Similarly, ECV0 was not significantly different from ECV1 (28.5 ± 6.7% vs. 28.5 ± 6.2, p = 0.801) and ECVsyn (28.5 ± 6.7% vs. 28.4 ± 6.0, p = 0.974) using inline method. ECVsyn values revealed relatively large discrepancies in patients with lower Hcts compared with those with higher Hcts. CONCLUSIONS: Venous Hcts measured on a different day from that of the CMR examination can still be used to measure ECV. ECVsyn can provide an alternative method to quantify ECV without needing a blood sample, but significant ECV errors occur in patients with severe anemia.