- Rating
- 2
- Confidence
- 3
- Informativeness
- 3
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper proposes BeatFlow-ECG, a conditional rectified-flow model for reconstructing single-channel ECG from synchronized PPG and inertial signals. The framework combines a convolutional encoder-decoder, Transformer bottleneck, flow-time conditioning, motion-aware training, and an STFT loss. Evaluation on PPG-DaLiA and WESAD under LOSO protocols shows improved waveform and beat-timing performance over the considered baselines.
Critical strengths
The paper addresses a relevant wearable cardiac-monitoring problem and presents a coherent conditional flow-based framework. The evaluation is reasonably comprehensive, including waveform metrics, R-peak/heart-rate metrics, regional P/QRS/T correlations, motion-stratified analysis, and ablations.
Critical weaknesses
- The scope of “ECG reconstruction” is unclear. Each 10-s ECG target is independently z-score normalized, which removes absolute baseline and amplitude information. Therefore, the model mainly reconstructs normalized morphology and timing rather than the physical ECG waveform. Pearson correlation and R-peak F1 also do not directly assess peak-amplitude fidelity. The manuscript should either restrict its claims accordingly or provide amplitude-sensitive evaluation.
- The contribution of motion-aware training is not fully isolated. The ablation removes motion weighting and STFT loss together, so the specific effect of motion-dependent loss weighting cannot be determined. In addition, LOSO results are reported without clear subject-level variability or confidence intervals, making it difficult to assess consistency across subjects.
Questions for authors
-
Is absolute ECG amplitude reconstruction outside the intended scope of the method, given the per-window z-score normalization? Are any target ECG statistics used during inference or inverse normalization?
-
Are the P/QRS/T regional correlations computed using ground-truth or predicted R peaks?
-
Can the authors provide an ablation removing only motion-dependent loss weighting while retaining the STFT loss?
-
Can subject-level variability or confidence intervals across LOSO folds be reported?
-
There are some abbreviations without any explanation
-
In the results section, the authors only present the comparison results between baseline models and the proposed model using PPG-DaLiA. What about the WESAD dataset?
-
There are some details in the network structure not clearly mentioned, like AdaRMSNorm, how is it implemented in the current model?