"Can't Take the Pressure?": Examining the Challenges of Blood Pressure Estimation via Pulse Wave Analysis

Suril Mehta,Nipun Kwatra,Mohit Jain,Daniel McDuff

2023-04-24

Abstract:The use of observed wearable sensor data (e.g., photoplethysmograms [PPG]) to infer health measures (e.g., glucose level or blood pressure) is a very active area of research. Such technology can have a significant impact on health screening, chronic disease management and remote monitoring. A common approach is to collect sensor data and corresponding labels from a clinical grade device (e.g., blood pressure cuff), and train deep learning models to map one to the other. Although well intentioned, this approach often ignores a principled analysis of whether the input sensor data has enough information to predict the desired metric. We analyze the task of predicting blood pressure from PPG pulse wave analysis. Our review of the prior work reveals that many papers fall prey data leakage, and unrealistic constraints on the task and the preprocessing steps. We propose a set of tools to help determine if the input signal in question (e.g., PPG) is indeed a good predictor of the desired label (e.g., blood pressure). Using our proposed tools, we have found that blood pressure prediction using PPG has a high multi-valued mapping factor of 33.2% and low mutual information of 9.8%. In comparison, heart rate prediction using PPG, a well-established task, has a very low multi-valued mapping factor of 0.75% and high mutual information of 87.7%. We argue that these results provide a more realistic representation of the current progress towards to goal of wearable blood pressure measurement via PPG pulse wave analysis.

Signal Processing,Artificial Intelligence,Human-Computer Interaction,Machine Learning

What problem does this paper attempt to address?

The paper primarily explores the issue of non-invasively predicting Blood Pressure (BP) using Photoplethysmograms (PPG) through Pulse Wave Analysis (PWA). Specifically, the study focuses on the following aspects:

1. **Analyzing common issues in existing research**:
- **Data leakage**: There may be an overlap between the training set and the test set in the study, leading to artificially inflated model performance on the test set.
- **Over-constraining the task**: Some studies simplify the task difficulty by limiting the range of blood pressure values to achieve better prediction results, which may lead to neglecting the prediction capability in extreme cases.
- **Unreasonable calibration methods**: Although calibration can improve prediction accuracy, some calibration strategies may overly rely on data from specific time points, resulting in poor long-term prediction performance.
- **Unrealistic data preprocessing**: Some studies improve prediction accuracy by filtering out noisy data, but this approach may lead to poor model performance in real-world scenarios.

2. **Proposing evaluation tools**:
- **Multi-value mapping check**: Used to detect data samples that are close in the input space but far apart in the output space, to assess whether the prediction task is well-conditioned.
- **Mutual information evaluation**: Used to estimate the strength of the correlation between the input signal and the target output, thereby determining whether the input signal can well predict the target output.

3. **Empirical analysis**:
- Using the above tools to conduct an in-depth analysis of the task of predicting blood pressure from PPG signals, the results show that the task has a high multi-value mapping factor (33.2%) and low mutual information (9.8%), indicating that predicting blood pressure from PPG signals is an ill-conditioned problem.
- In contrast, the task of predicting heart rate from PPG signals has a low multi-value mapping factor (0.75%) and high mutual information (87.7%), indicating that PPG signals are a good predictor for heart rate.

In summary, this paper aims to reveal the challenges in the task of predicting blood pressure from PPG signals and to promote the development of future research by proposing new evaluation tools to more accurately assess the feasibility of such health parameter prediction tasks.