arXiv subjects

Kévin Albert

Publications and source records attributed to Kévin Albert.

3 recordsLinked to original sources

arXiv2026

GRN-Transformer: Enhancing Motion Artifact Detection in PICU Photoplethysmogram Signals

Photoplethysmogram (PPG) signals, optical measurements of pulsatile blood flow used continuously in intensive care monitoring, are frequently contaminated by motion, low perfusion, or sensor displacement, producing waveform artifacts that propagate into downstream estimates such as SpO\textsubscript{2} and trigger spurious clinical alarms. Automated artifact detection is therefore a prerequisite for reliable bedside decision support. Transformer classifiers, which use self-attention to weight contributions from every part of the input pulse, are well suited to learning artifact morphology, but their performance is known to degrade on the small, imbalanced datasets typical of single-center clinical studies. We propose the \emph{GRN-Transformer}, which integrates a single Gated Residual Network (GRN) block atop a standard Transformer encoder stack, serving as a small-data regularizer. On a labeled Pediatric Intensive Care Unit (PICU) PPG dataset from CHU Sainte-Justine Hospital (CHUSJ), the GRN-Transformer reaches $98\%$ accuracy, $90\%$ precision, $97\%$ recall, and $93\%$ F1-score using only $5\%$ of the annotated pulses, substantially improving recall over the baseline Transformer ($+11$ points) without sacrificing precision. Cross-validated evaluation indicates that clean-data accuracy gains are modest once fold-to-fold variability is accounted for, but the GRN-Transformer is markedly more robust to realistic signal degradations (noise, baseline wander, sensor dropout), trains approximately $2.7\times$ faster than the baseline, and runs at $6.33$~ms p99 latency on a CPU-only consumer laptop. A retrospective simulation suggests the model could meaningfully reduce clinician review burden when used as a pre-filter for PPG-driven alarms. These results support the GRN-Transformer as a deployable artifact-detection component for resource-constrained pediatric clinical settings.

eess.SP↗

arXiv2025

Transformer Meets Gated Residual Networks To Enhance Photoplethysmogram Artifact Detection Informed by Mutual Information Neural Estimation

This study delves into the effectiveness of various learning methods in improving Transformer models, focusing particularly on the Gated Residual Network Transformer (GRN-Transformer) in the context of pediatric intensive care units (PICU) with limited data availability. Our findings indicate that Transformers trained via supervised learning are less effective compared to MLP, CNN, and LSTM networks in such environments. Yet, leveraging unsupervised and self-supervised learning on unannotated data, with subsequent fine-tuning on annotated data, notably enhances Transformer performance, although not to the level of the GRN-Transformer. Central to our research is the analysis of different activation functions for the Gated Linear Unit (GLU), a crucial element of the GRN structure. We also employ Mutual Information Neural Estimation (MINE) to evaluate the GRN's contribution. Additionally, the study examines the effects of integrating GRN within the Transformer's Attention mechanism versus using it as a separate intermediary layer. Our results highlight that GLU with sigmoid activation stands out, achieving 0.98 accuracy, 0.91 precision, 0.96 recall, and 0.94 F1 score. The MINE analysis supports the hypothesis that GRN enhances the mutual information between the hidden representations and the output. Moreover, the use of GRN as an intermediate filter layer proves more beneficial than incorporating it within the Attention mechanism. In summary, this research clarifies how GRN bolsters GRN-Transformer's performance, surpassing other learning techniques. These findings offer a promising avenue for adopting sophisticated models like Transformers in data-constrained environments, such as PPG artifact detection in PICU settings.

eess.SP↗

arXiv2025

A Novel Transformer-Based Self-Supervised Learning Method to Enhance Photoplethysmogram Signal Artifact Detection

Recent research at CHU Sainte Justine's Pediatric Critical Care Unit (PICU) has revealed that traditional machine learning methods, such as semi-supervised label propagation and K-nearest neighbors, outperform Transformer-based models in artifact detection from PPG signals, mainly when data is limited. This study addresses the underutilization of abundant unlabeled data by employing self-supervised learning (SSL) to extract latent features from these data, followed by fine-tuning on labeled data. Our experiments demonstrate that SSL significantly enhances the Transformer model's ability to learn representations, improving its robustness in artifact classification tasks. Among various SSL techniques, including masking, contrastive learning, and DINO (self-distillation with no labels)-contrastive learning exhibited the most stable and superior performance in small PPG datasets. Further, we delve into optimizing contrastive loss functions, which are crucial for contrastive SSL. Inspired by InfoNCE, we introduce a novel contrastive loss function that facilitates smoother training and better convergence, thereby enhancing performance in artifact classification. In summary, this study establishes the efficacy of SSL in leveraging unlabeled data, particularly in enhancing the capabilities of the Transformer model. This approach holds promise for broader applications in PICU environments, where annotated data is often limited.

cs.LG↗