Abstract Body (Do not enter title and authors here): In patients with tetralogy of Fallot (TOF) repaired with a transannular patch (TAP), right ventricular outflow tracts (RVOT) are classified by shape using complex 3D measurements. RVOT shape is crucial for evaluating transcatheter pulmonary valve replacement (TPVR).

Aim:
To evaluate if a simple graphical representation of the RVOT in TOF patients repaired with TAP can classify RVOTs than Level III imaging experts in a Latin American tertiary care center.

Methods:
This retrospective single-center study included TOF post-TAP patients who underwent cardiac magnetic resonance (CMR) between 2014 and 2023. Dynamic CMR angiography images were used to segment and create 3D reconstructions of the RVOT and pulmonary artery. Centerline-based measurements derived from perimeter were taken at six RVOT points (bifurcation, mid trunk, supravalvular, pulmonary valve annulus, subvalvular, and RVOT) following the TPVR planning protocol by Canan et al. to measure the area and create a graphical representation (Figure 1). The graphs and 3D reconstructions were categorized into five RVOT types based on Schievano et al.'s classification. Graphs were compared to the original classification by overlaying graph forms, and 3D files were classified by three Level III cardiac imagers. The correlation between evaluators and graphs was measured using the Kappa coefficient.

Results:
The study population included 60 patients (55% male, mean age 17 ± 12 years at CMR). The RVOT type distribution, according to the graph representation, was 53.3% type 1, 13.3% type 2, 5% type 3, 10% type 4, and 15% type 5. Only 3% of the graphs did not fit the established classification. The initial agreement among the three evaluators regarding 3D classification was 28%. Compared with the graphs, evaluator 1 agreed fairly (Kappa=0.24, p=0.0006), evaluator 2 slightly (Kappa=0.197, p=0.0008), and evaluator 3 poorly (Kappa=0.07, p=0.15).

Conclusions:
Our approach reclassified RVOT shapes more effectively than imaging experts, showing a high fit in classification and low agreement among cardiac imagers. This method could be an additional tool to identify TOF patients suitable for TPVR post-TAP.