Key result
PanEcho on single-view echo detects cancer therapy-related cardiac dysfunction with ~87% accuracy.
Why the study?
Surveillance for cancer therapy-related cardiac dysfunction (CTRCD) relies on LVEF and GLS measurements that require dedicated software and trained personnel. Whether PanEcho pretrained deep learning weights fine-tuned on apical four-chamber videos can estimate LVEF and GLS to support CTRCD assessment needed evaluation.
Does a fine-tuned PanEcho deep learning model applied to apical four-chamber echocardiographic videos accurately estimate LVEF and GLS and support exam-level cancer therapy-related cardiac dysfunction assessment in a longitudinal cardio-oncology cohort?
Cohort (n=185)
Does a fine-tuned PanEcho deep learning model applied to apical four-chamber echocardiographic videos accurately estimate LVEF and GLS and support exam-level cancer therapy-related cardiac dysfunction assessment in a longitudinal cardio-oncology cohort?
Answers are grounded in this paper and cite their sources.
A fine-tuned PanEcho deep learning model applied to routine apical four-chamber echocardiographic videos provides clinically interpretable estimates of LVEF and GLS and demonstrates high specificity (91.4%) and moderate discrimination (AUC 0.714) for detecting cancer therapy-related cardiac dysfunction.
Background Cancer therapy-related cardiac dysfunction (CTRCD) surveillance relies on left ventricular ejection fraction (LVEF) and global longitudinal strain (GLS), measures that often require dedicated software and trained personnel. Video-based deep learning may reduce these barriers by estimating LVEF/GLS directly from routine echocardiographic cine loops; PanEcho is a pretrained model designed for this task. Purpose To evaluate whether PanEcho pretrained weights, fine-tuned on apical four-chamber (A4C) videos, can estimate LVEF and GLS and support exam-level CTRCD assessment in a longitudinal cardio-oncology cohort. Methods We analysed 185 patients contributing 665 longitudinal echocardiographic A4C examinations with reference LVEF and GLS. Videos were standardised using an EchoNet-Dynamic-style preprocessing pipeline to match PanEcho's input domain. Patient-level splits comprised 130/28/27 patients (469/96/100 examinations) for training/validation/test. Starting from pretrained PanEcho weights, we fine-tuned models to predict LVEF and GLS; results are reported for the default configuration selected after exploring multiple parameter settings. Test-set performance (n=100 for LVEF and GLS) was summarised with MAE, RMSE, and Pearson r. Reference CTRCD labels were derived from measured baseline-to-follow-up LVEF/GLS changes per the 2021 ICOS consensus (aligned with 2022 ESC cardio-oncology guidance). CTRCD was inferred using measured baseline values with predicted follow-up values and evaluated in follow-up examinations with complete baseline measurements, reference labels, and corresponding predictions. Results On the test set, LVEF estimation achieved MAE 3.85%, RMSE 4.78%, and Pearson r=0.503 (Fig. 1); GLS estimation achieved MAE 1.89%, RMSE 2.43%, and Pearson r=0.417 (Fig. 2). For CTRCD, the inference was evaluable in 480 follow-up examinations (50 reference-positive) and achieved sensitivity 50.0%, specificity 91.4%, accuracy 87.1%, and ROC AUC 0.714. Conclusions Fine-tuned PanEcho applied to EchoNet-Dynamic-style-preprocessed A4C videos provided clinically interpretable estimates of LVEF and GLS and supported feasible exam-level CTRCD assessment when measured baseline values were retained. To strengthen generalisability and longitudinal fidelity, future analyses should incorporate multi-view inputs, mitigate class imbalance, validate performance in independent external cohorts, and align follow-up using recorded study dates rather than exam-order suffixes.Fig 1.Test-set LVEF: reference vs predictedFig 2.
No takes yet. Share an insight, caveat, or question.
Guerra et al. (2026) conducted a cohort in Cancer therapy-related cardiac dysfunction (CTRCD) (n=185). Fine-tuned PanEcho deep learning model on apical four-chamber echocardiographic videos vs. Reference measured LVEF, GLS, and 2021 ICOS consensus CTRCD labels was evaluated on Estimation of LVEF and GLS (MAE, RMSE, Pearson r) and exam-level CTRCD assessment (ROC AUC, accuracy, sensitivity, specificity) (ROC AUC 0.714 (LVEF MAE 3.85%, r=0.503; GLS MAE 1.89%, r=0.417)). Fine-tuned PanEcho on A4C videos estimated LVEF (MAE 3.85%, r=0.503) and GLS (MAE 1.89%, r=0.417), detecting CTRCD with 87.1% accuracy and AUC 0.714.
Synapse has enriched 10 related papers. Consider them for comparative context: