Search

Background: The partial volume correction (PVC) of cardiac PET datasets using anatomical side information during reconstruction is appealing but not straightforward. Other techniques, which do not make use of additional anatomical information, could be equally effective in improving the reconstructed myocardial activity.Methods: Resolution modeling in combination with different noise suppressing priors was evaluated as a means to perform PVC. Anatomical priors based on a high-resolution CT are compared to non-anatomical, edge-preserving priors (relative difference and total variation prior). The study is conducted on ex vivo datasets from ovine hearts. A simulation study additionally clarifies the relationship between prior effectiveness and myocardial wall thickness.Results: Simple resolution modeling during data reconstruction resulted in over- and underestimation of activity, which hampers the absolute left ventricular quantification when compared to the ground truth. Both the edge-preserving and the anatomy-based PVC techniques improve the absolute quantification, with comparable results (Student t-test, P = .17). The relative tracer distribution was preserved with any reconstruction technique (repeated ANOVA, P = .98).Conclusions: The use of edge-preserving priors emerged as optimal choice for quantification of tracer uptake in the left ventricular wall of the available datasets. Anatomical priors visually outperformed edge-preserving priors when the thinnest structures were of interest. [ABSTRACT FROM AUTHOR]

Background: In a previous study on ex vivo, static cardiac datasets, we investigated the benefits of performing partial volume correction (PVC) in cardiac 18F-Fluorodeoxyglucose(FDG) PET datasets. In the present study, we extend the analysis to in vivo cardiac datasets, with the aim of defining which reconstruction technique maximizes quantitative accuracy and, ultimately, makes PET a better diagnostic tool for cardiac pathologies.Methods: In vivo sheep datasets were acquired and reconstructed with/without motion correction and using several reconstruction algorithms (with/without resolution modeling, with/without non-anatomical priors). Corresponding ex vivo scans of the excised sheep hearts were performed on a small-animal PET scanner (Siemens Focus 220, microPET) to provide high-resolution reference data unaffected by respiratory and cardiac motion. A comparison between the in vivo cardiac reconstructions and the corresponding ex vivo ground truth was performed.Results: The use of an edge-preserving prior (Total Variation (TV) prior in this work) in combination with motion correction reduces the bias in absolute quantification when compared to the standard clinical reconstructions (- 0.83 vs - 3.74 SUV units), when the end-systolic gate is considered. At end-diastole, motion correction improves absolute quantification but the PVC with priors does not improve the similarity to the ground truth more than a regular iterative reconstruction with motion correction and without priors. Relative quantification was not influenced much by the chosen reconstruction algorithm.Conclusions: The relative ranking of the algorithms suggests superiority of the PVC reconstructions with dual gating in terms of overall absolute quantification and noise properties. A well-tuned edge-preserving prior, such as TV, enhances the noise properties of the resulting images of the heart. The end-systolic gate yields the most accurate quantification of cardiac datasets. [ABSTRACT FROM AUTHOR]

The relationship between diastolic dysfunction (DD) and outcome after CRT is debated. The purpose of this study was to evaluate the role of DD in predicting all-cause mortality in heart failure patients undergoing CRT. One-hundred ninety-three patients (age: 67 ± 11 years, QRS width: 167 ± 21 ms, LVEF 28 ± 8%) were included in this multicentre prospective study. Mitral filling pattern, mitral tissue Doppler velocity, tricuspid regurgitation velocity, and indexed left atrial volume were used to classify DD from grade I to III according to the 2016 recommendations of the American Society of Echocardiography/European Association of Cardiovascular Imaging. A reduction of LV end-systolic volume > 15% at 6-month follow-up (FU) identified CRT-responders and was observed in 132 (68%) patients. During a median 35 months FU, 29 (15%) patients died. Through multivariable analysis, coronary artery disease [HR 2.21 (1.00–4.89) P = 0.05], NYHA [HR 2.92 (1.45–5.87), P = 0.003] and grade I DD [HR 0.34 (0.13–0.89)] were shown to be independent predictors of prognosis (Table 1). Grade I DD was associated with a longer survival rate in both responders and non-responders (Fig. 1). Non-responders with grade II-III DD had the worse outcome [HR 12.5 (3.56–44.04), P < 0.0001] Better diastolic function at baseline is associated with improved survival after CRT implantation, independently of CRT-response. [ABSTRACT FROM AUTHOR]

Patients undergoing cardiac resynchronisation therapy are heterogeneous. Therefore, the prediction of CRT-response and outcome is challenging. The aim of this study was to evaluate the relative impact of clinical, electrocardiographic, and echocardiographic variables on the left ventricular (LV) remodelling and prognosis of CRT-candidates by the application of machine-learning (ML) approaches. In total, 193 patients undergoing CRT (mean age 67 ± 11 years, QRS width: 167 ± 21 msec) were prospectively included in this multicentre study. We used a combination of the Boruta algorithm and random forest methods (RF) to identify features predicting both CRT-volumetric response and prognosis. The model performance was tested by the area under the receiver-operating curve (AUC). We also applied the K-medoid method to identify clusters of phenotipically-similar patients. From 28 clinical, electrocardiographic, and echocardiographic-derived variables, 16 features were predictive of CRT-response, and 11 features were predictive of prognosis. Among the predictors of CRT-response, 6 variables (38%) pertained to right ventricular (RV) size or function. The tricuspid annular plane systolic excursion was the main feature associated with prognosis. The selected features were associated with a very good prediction of both CRT-response (AUC: 0.81, 95% CI: 0.74–0.87) and outcome (AUC 0.84, 95% CI: 0.75–0.93) (Fig. 1 A–B). An unsupervised ML approach allowed the identifications of two phenogroups of patients who differed significantly in clinical parameters, biventricular size and RV function and had significantly different prognosis (log-rank P < 0.0001; HR: 4.7, 95% CI: 2.1–10.0, P < 0.001) (Table 1). ML can reliably identify clinical and echocardiographic features associated with CRT-response and prognosis. Our results underscore the value of the assessment of RV-size and function parameters for the risk stratification of CRT-candidates. [ABSTRACT FROM AUTHOR]