ウェアラブルセンサーとAIによる新しい血圧モニタリング手法（A Better Way to Monitor Blood Pressure?）

2026-08-17 ジョンズ・ホプキンス大学（JHU）

＜関連情報＞

• https://hub.jhu.edu/2026/08/17/wearable-sensors-ai-monitor-blood-pressure-icu/
• https://www.sciencedirect.com/science/article/abs/pii/S0010482526004257

重症患者における非侵襲的動脈血圧波形生成：センサーベースの深層学習アプローチ Non-invasive arterial blood pressure waveform generation in critically ill patients: A sensor-based deep learning approach

Carl Harris, Bright Nnadi, Sampath Rapuri, John Rattray, Francesco Tenore, Ralph Etienne-Cummings, Robert D. Stevens
Computers in Biology and Medicine  Available online: 15 July 2026
DOI:https://doi.org/10.1016/j.compbiomed.2026.111861

Highlights

• MOSAIC wearable sensors enable non-invasive continuous BP estimation in the ICU.
• A CNN/LSTM hybrid model achieved , on ABP waveform prediction.
• Split-conformal prediction gave calibrated, value-dependent uncertainty bands.
• MOSAIC may reduce reliance on invasive arterial catheterization in the ICU.

Abstract

Continuous monitoring of Arterial Blood Pressure (ABP) in critically ill patients requires invasive arterial catheterization, which carries risks of thrombosis, vascular injury and infection. Here, we train and validate a computational model for continuous non-invasive ABP estimation in Intensive Care Unit (ICU) patients using a novel wearable sensor array. The sensor acquires continuous high frequency photoplethysmography (PPG) and electrocardiography (ECG) signals which are used as inputs in a deep learning algorithm for beat-to-beat reconstruction of ABP waveforms. We include 28 patients enrolled in four ICU units at Johns Hopkins Hospital, comprising 15,489 5-s ECG and PPG segments. A CNN/LSTM hybrid architecture achieved an R² of 0.732 and a mean absolute error (MAE) of 6.42 ± 3.82 mmHg, with systolic and diastolic blood pressure MAEs of 6.21 ± 3.89 and 3.41 ± 3.02 mmHg, respectively. This performance closely approached an upper-bound model trained on contemporaneously acquired ground truth ECG and PPG signals (R² =0.799, MAE=5.67mmHg), indicating that the sensors retain most hemodynamically relevant information. Split-conformal prediction provided calibrated uncertainty intervals with coverage meeting nominal targets, offering a principled framework for bedside confidence assessment. These findings demonstrate the feasibility of accurate, continuous, non-invasive ABP waveform estimation from wearable biosignals in critically ill patients, establishing a foundation for reducing dependence on invasive arterial monitoring while preserving the waveform-level information essential for hemodynamic management.