What does this research mean for the field?

Temporal deep learning models significantly improve respiratory rate estimation from PPG-derived numerical features, achieving mean absolute error values as low as 0.521 breaths/min. Novelty: ClaimNovelty.CONFIRMATORY. Consensus alignment: ConsensusAlignment.SUPPORTS_CONSENSUS.

What question did this study set out to answer?

The aim is to evaluate deep learning models for estimating respiratory rate from photoplethysmography data.

March 8, 2026Open Access

Comparative Evaluation of Deep Learning Models for Respiratory Rate Estimation Using PPG-Derived Numerical Features

Key Result

Bidirectional LSTM models achieved the lowest MAE of 0.521 breaths/min for respiratory rate estimation from PPG features, outperforming feedforward models.

Key Points

The aim is to evaluate deep learning models for estimating respiratory rate from photoplethysmography data.
Constructed fixed-length sliding windows from PPG data
Trained five neural network architectures: DFNN, RNN, Bi-RNN, LSTM, Bi-LSTM
Assessed model performance using MAE, RMSE, R2, and computational runtime
Models with temporal dependencies achieved MAE as low as 0.521 breaths/min
Temporal deep learning models outperform static feedforward baselines
Models are competitive with previously reported PPG-based methods

Structured PICO

Do deep learning models incorporating temporal dependencies improve the accuracy of respiratory rate estimation from PPG-derived features compared to static feedforward networks?

Population

Dataset of photoplethysmography (PPG)-derived numerical features for respiratory rate estimation (patient demographics and sample size not specified)

Intervention

Deep learning models incorporating temporal dependencies (unidirectional and bidirectional Recurrent Neural Networks [RNN, Bi-RNN], and unidirectional and bidirectional Long Short-Term Memory networks [LSTM, Bi-LSTM])

Comparator

Static feedforward baseline (Deep Feedforward Neural Network [DFNN])

Outcome

Model performance assessed using mean absolute error (MAE), root mean squared error (RMSE), coefficient of determination (R2), and computational runtimesurrogate

Deep learning models with temporal dependencies (such as RNNs and LSTMs) can accurately estimate respiratory rate from non-invasive PPG data, achieving competitive accuracy for real-time monitoring applications.

Main Result

Absolute Event Rate: 0% vs 0%

Abstract

Respiratory rate (RR) is a critical vital sign for the early detection of hypoxia and respiratory deterioration, yet its continuous monitoring remains challenging in clinical environments. Photoplethysmography (PPG) provides a non-invasive source of physiological information from which respiratory dynamics can be inferred. In this study, numerical physiological features derived from PPG data were used to comparatively evaluate multiple deep learning models for respiratory rate estimation. Fixed-length sliding windows were constructed from the dataset and used to train five neural network architectures: a Deep Feedforward Neural Network (DFNN), unidirectional and bidirectional Recurrent Neural Networks (RNN, Bi-RNN), and unidirectional and bidirectional Long Short-Term Memory networks (LSTM, Bi-LSTM). Model performance was assessed using mean absolute error (MAE), root mean squared error (RMSE), coefficient of determination (R2), and computational runtime. Results indicate that models incorporating temporal dependencies outperform the static feedforward baseline, achieving MAE values as low as 0.521 breaths/min, making them competitive with or lower than previously reported PPG-based approaches. These findings highlight the effectiveness of temporal deep learning models for respiratory rate estimation from PPG-derived numerical features and provide insight into accuracy–efficiency trade-offs relevant to real-time monitoring applications.