REVIEW 4 major objections 5 minor 3 references
Generalised Label-free Artefact Cleaning for Real-time Medical Pulsatile Time Series
T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash
Pith's one-line read A label-free variational autoencoder trained on 180,000 ten-second blood-pressure samples cleans artefacts across patients, cohorts, and signal types, and runs in real time.
desk verdict A useful, well-scoped engineering extension of DeepClean with a real-time deployment story, but the evaluation is under-specified (labels, cohort size, error bars) and needs major revision before the generalization claims carry their weight. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is a one-dimensional convolutional variational autoencoder trained with the standard variational-autoencoder objective (the evidence lower bound), with the artefact decision made from mean-squared reconstruction error. Two modules wrap the autoencoder: a frequency adapter that resamples any input rate to the training rate and back, and a reversible instance-normalisation layer that standardises each segment's mean and variance before encoding and restores them after decoding. The reconstruction error is thresholded at the 90th percentile of the validation set, and segments above the threshold are flagged as artefacts. A latent dimension of roughly 20 is reported as the best balance between reconstruction fidelity and class separation. This machinery learns a probabilistic prior over clean pulsatile waveforms and classifies by deviation from that prior, requiring no labels.
What would settle it
Collect a sample of the held-out arterial-blood-pressure segments and have two independent clinicians annotate each as artefactual or clean; if inter-rater agreement is low or the model's errors concentrate on segments clinicians disagree about, then the reported 90% accuracy and 0.89 F1-score are not anchored to a reliable ground truth. Alternatively, feed the model a set of clean but atypical rhythm segments (e.g., from patients with atrial fibrillation): if it flags most of them as artefacts, the claimed generalisation to unseen physiology fails.
Extended reading notes
Core claim
GenClean's central discovery is that a variational autoencoder trained only to reconstruct clean pulsatile waveforms doubles as a generalised artefact detector: a segment is labelled artefactual when its reconstruction deviates from the input beyond a threshold, and this rule transfers to new patients, new recording sites, different sampling rates, and even a different sensor modality. The paper demonstrates this by training on one hospital's trauma cohort and testing on held-out patients and on an external intensive-care arterial-blood-pressure cohort, where it reports 95.6% accuracy, and by adapting the same training to a wrist-worn photoplethysmography dataset, where it reports 85.8% accuracy. The authors argue that the generalisation comes from three design choices working together: resampling inputs to a common frequency, reversible instance normalisation that strips patient-specific mean and variance before reconstruction, and a loss computed before rescaling so one global threshold works across patients.
Load-bearing premise
The reported accuracies assume the expert labels for the held-out and external test sets are correct and consistently defined, yet the paper never describes how those labels were generated or validated, so the numbers could be measuring agreement with unverified annotations.
Editorial extensions
If this is right
- A model trained on one centre's trauma cohort can be applied directly to arterial blood pressure from a different intensive-care population recorded at a different sampling rate, with reported accuracy above 95% and no retraining.
- Removing artefact-contaminated segments cuts the number of counted hypertensive events by about 22% on the external cohort, suggesting that alarms derived from raw waveforms could be reduced substantially.
- The same variational-autoencoder recipe transfers to photoplethysmography with accuracy above 85%, so the method is not tied to invasive blood-pressure morphology.
- With roughly 19–39 ms of processing per ten-second segment and only a few mFLOPS, the method runs on bedside CPU hardware rather than requiring a GPU server.
- The latent space becomes visibly more separated between artefactual and clean segments, so the representation itself could support downstream monitoring tasks beyond cleaning.
Reading between the lines
- Inference: the 90th-percentile validation threshold ties the operating point to the artefact prevalence of the validation set; deploying in a unit with a different artefact rate would require recalibrating the threshold, which the paper does not address.
- Inference: because the variational autoencoder reconstructs the most probable clean waveform, rare but physiologically normal patterns such as arrhythmias may be flagged as artefacts; a targeted test on arrhythmia segments would reveal whether this failure mode is real.
- Inference: the reported 22% reduction in hypertensive events should not be read as pure error correction; without an independent human-annotated reference, some removed 'events' could be genuine but atypical physiology, so a gold-standard clinical validation is a natural next step.
- Inference: the same label-free recipe should transfer to other cardiac-driven pulsatile signals such as intracranial pressure or fetal heart rate, since the only requirement is a learnable waveform prior for clean segments; the paper suggests but does not test this.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes GenClean, a label-free variational-autoencoder-based artefact cleaning framework for pulsatile medical time series. The method is trained on 180,000 ten-second arterial blood pressure (ABP) segments from 150 traumatic brain injury patients and is claimed to generalize across patient-level distribution shifts, across disease cohorts (MIMIC-III), and to other signal modalities (PPG). The authors report at least 90% accuracy and 0.89 F1-score on an internal held-out set, 95.6% accuracy on MIMIC-III ABP, 85.8% accuracy on WESAD PPG, and a 22% reduction in detected hypertension events after cleaning. They also demonstrate integration into the ICM+ monitoring software with modest CPU and memory overhead and provide code.
Significance. If the reported results hold, GenClean would be a valuable contribution: it is among the first label-free artefact-cleaning methods to explicitly target patient- and site-level generalisation, it uses a large in-house training corpus, the code is made available, and the real-time ICM+ integration addresses a practically important deployment constraint. The honest limitations section and the explicit investigation of distribution shifts are also strengths. However, the central evaluation numbers depend on artefact labels whose generation is not described, so the current evidence is not sufficient to certify the generalisation claims.
major comments (4)
- [Methods, Artefact Cleaning Framework; Results, Clinical Datasets] The evaluation labels that support the central claims are never described. The internal held-out set is only said to be 'balance labelled (non-artefactual: artefactual = 1:1)', and the evaluation is said to use 'our experts-labelled data', but the paper gives no annotation protocol, no definition of what constitutes an artefact at the segment or waveform level, no inter-rater agreement, and no information about annotator blinding. For the MIMIC-III cross-disease cohort, the label generation is not described at all. Since accuracy, F1, sensitivity, and specificity are all computed against these labels, the headline numbers in Fig. 2d and Table 1 cannot be interpreted without knowing whether the labels are reliable and independent of the heuristic rules used in training.
- [Results, Generalisation on Cross-cohort; Table 1] The cross-disease generalisation claim rests on only five MIMIC-III patients and a single pooled accuracy of 95.6%, reported without the number of test segments, a per-patient breakdown, confidence intervals, or significance tests. The hypertension-event reduction is a raw change from 2134 to 1661 events over the same five patients. This sample size is too small to support the broad cross-disease and cross-site generalisation statement made in the abstract and discussion; the authors should report per-patient metrics, confidence intervals, and the total number of annotated segments.
- [Methods, Heuristic rules; Results, WESAD (PPG)] For PPG, the Methods define artefact classification by a band-power heuristic (segments with more than 30% of total signal power outside 0.5–3 Hz are classified as artefacts). If the WESAD test labels are generated with this same criterion, the reported 85.8% accuracy measures agreement with that heuristic rather than generalisation to independently established PPG artefacts. The paper should state explicitly how the WESAD test labels were obtained, whether they are independent of the training heuristics, and whether the same heuristic was used for both DaLiA training filtering and WESAD evaluation.
- [Methods, Artefact Cleaning Framework; Methods, Training methods] The artefact threshold is set as the 90th percentile of reconstruction error on the validation set, and the loss is computed before the rescaling step of the adaptive normalisation layer. These choices are reasonable in principle, but the paper reports no sensitivity analysis with respect to the threshold percentile or the validation-set composition. Without such an analysis, it is unclear whether the claimed accuracy and F1 are robust to this hand-set threshold or whether they would change materially under plausible variations in validation-set artefact prevalence.
minor comments (5)
- [Results, Distribution Shift Exploration] The sentence 'One of the patients (Test 5) consistently exhibited lower performance in both metrics, with no value above 0.6' is ambiguous: it is not clear whether this refers to the previous method or to GenClean. If it refers to GenClean, it appears to contradict the earlier claim of at least 90% accuracy for the generalised model; please clarify.
- [Real-time results and Table 2] The processing time is reported as '~19 ms for 120 Hz' and 'longest processing time reaching ~26 ms (150 Hz)' in the results, while Table 2 lists 'Processing Time 39 ms' for a ten-second segment. Please reconcile these numbers and specify whether they measure the same workload.
- [Methods, Artefact Cleaning Framework] In the ELBO equation, the text says 'p(x|z) is the encoder, q(z|x) is the decoder', but in the standard VAE notation used in the equation q(z|x) is the encoder and p(x|z) is the decoder. Please correct this description.
- [General editorial] There are several typographical and formatting issues: 'trauma brain injury' should be 'traumatic brain injury', 'V AE' has inconsistent spacing, the WESAD description contains 'neural' instead of 'neutral', and some references are incomplete (e.g., Ref. 28 reports 'Sensors 23, nan'). Please proofread the manuscript.
- [Figure 3c] The caption and text state that both MSE and MAE were compared as reconstruction error metrics, but only the MSE result is described in detail. Please report the MAE comparison or state explicitly that it was used only as a check.
Circularity Check
No circular derivation found; GenClean's reported accuracy/F1 numbers are out-of-sample, though test-label provenance is under-described.
full rationale
The claimed derivation chain is not circular. GenClean's artefact threshold is set as the 90th percentile of the validation reconstruction error (Methods, Artefact Cleaning Framework) and is then applied to held-out patients and external cohorts, so the internal ≥90% accuracy/0.89 F1 and MIMIC-III 95.6% accuracy are not fitted to the evaluation labels. The frequency adapter and reversible instance normalisation come from standard cited prior work (ref. 43) and are tested across sampling rates, not smuggled in as an unverified self-citation. The paper's self-citations, chiefly DeepClean (ref. 8) and ICM+ (ref. 17), motivate the VAE backbone and the real-time platform, but the central generalisation claim is independently supported by the Fig. 3d ablations and by external MIMIC-III/WESAD evaluations, so those self-citations are not load-bearing circularity. The main weakness is not circularity but missing evidence: the paper reports 'experts-labelled data' and MIMIC-III sensitivity/specificity without giving an annotation protocol, so the ground truth for the accuracy numbers is not verifiable. That is a validity/correctness risk, not a definitional reduction. The Discussion also acknowledges generalisation limits (paediatrics, arrhythmias, bedside deployment), which is consistent with an honest, non-circular presentation.
Assumptions & free parameters
free parameters (5)
- Latent dimension =
20
- Artefact threshold =
90th percentile of validation reconstruction MSE
- ABP amplitude range =
0-300 mmHg
- ABP peak-to-peak minimum =
15 mmHg
- PPG band-power criterion =
0.5-3 Hz; >30% power outside band -> artefact
assumptions (4)
- domain assumption Variational autoencoder trained on label-free heuristic-filtered data learns a distribution representative of clean pulsatile signals
- domain assumption Expert labels on test sets (internal held-out, MIMIC-III, WESAD) are accurate ground truth
- standard math The ELBO loss for VAEs (Kingma and Welling) is a valid objective for learning clean-signal distributions
- standard math Reversible instance normalisation preserves the information needed for reconstruction
Cite this review
Pith. "Pith review of Generalised Label-free Artefact Cleaning for Real-time Medical Pulsatile Time Series." pith.science (2026). https://pith.science/paper/WSI3XELT
@misc{pith2026250421209,
author = {Pith},
title = {Pith review of: Generalised Label-free Artefact Cleaning for Real-time Medical Pulsatile Time Series},
year = {2026},
howpublished = {\url{https://pith.science/paper/WSI3XELT}},
note = {Machine review of arXiv:2504.21209}
}
read the original abstract
Artefacts compromise clinical decision-making in the use of medical time series. Pulsatile waveforms offer probabilities for accurate artefact detection, yet most approaches rely on supervised manners and overlook patient-level distribution shifts. To address these issues, we introduce a generalised label-free framework, GenClean, for real-time artefact cleaning and leverage an in-house dataset of 180,000 ten-second arterial blood pressure (ABP) samples for training. We first investigate patient-level generalisation, demonstrating robust performances under both intra- and inter-patient distribution shifts. We further validate its effectiveness through challenging cross-disease cohort experiments on the MIMIC-III database. Additionally, we extend our method to photoplethysmography (PPG), highlighting its applicability to diverse medical pulsatile signals. Finally, its integration into ICM+, a clinical research monitoring software, confirms the real-time feasibility of our framework, emphasising its practical utility in continuous physiological monitoring. This work provides a foundational step toward precision medicine in improving the reliability of high-resolution medical time series analysis
Figures
Reference graph
Works this paper leans on
-
[18]
Czosnyka, M. et al. Continuous assessment of the cerebral vasomotor reactivity in head injury. Neurosurgery 41, 11–19 (1997). 19. Johnson, A. E. W. et al. MIMIC-III, a freely accessible critical care database. Sci. Data 3, 160035 (2016). 20. Schmidt, P ., Reiss, A., Duerichen, R., Marberger, C. & V an Laerhoven, K. Introducing WESAD, a Multimodal Dataset ...
arXiv 1997
-
[26]
Haule, H. et al. VA E-IF: Deep feature extraction with averaging for unsupervised artifact detection in routine acquired ICU time-series. Preprint at http://arxiv.org/abs/2312.05959 (2023). 27. Nowroozilarki, Z., Mortazavi, B. J. & Jafari, R. V ariational Autoencoders for Biomedical Signal Morphology Clustering and Noise Detection. IEEE J. Biomed. Health ...
work page Pith review arXiv 2023
-
[35]
I, M.-T., JE, W., M, J. & M, A. Empirical Mode Decomposition-Based Method for Artefact Removal in Raw Intracranial Pressure Signals. Acta Neurochir . Suppl. 131, 201–205 (2021). 36. M, C., M, F. & P , S. Automatic Pulse Classification for Artefact Removal Using SAX Strings, a CENTER-TBI Study. Acta Neurochir . Suppl. 131, 231–234 (2021). 37. Deng, A. & Ho...
arXiv 2021
Reviewed August 16, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.