We provide new insights into the performance of camera-based heart and respiration rate extraction and evaluate its usability for replacing spot checks conducted in the general ward. A study was performed comprising of 36 ICU patients recorded for a total time of 699 h. The 5 beats/minute agreement between camera and ECG-based heart rate measurements was 81.5%, with a coverage of 81.9%, where the largest gap between measurements was 239 min. The challenges encountered in heart rate monitoring were limited visibility of the patient’s face and irregular heart rates, which led to poor agreement between camera- and ECG-based heart rate measurements. To prevent non-breathing motion from causing error in respiration rate extraction, we developed a metric which was used to detect non-breathing motion. The 3 breaths/minute agreement between the camera- and contact-based respiration rate measurements was 91.1%, with a coverage of 59.1%, where the largest gap between measurements was 114 min. Encountered challenges were the morphology of the respiration signal and irregular breathing. While a few challenges need to be overcome, the results show promise for the usability of camera-based heart and respiration rate monitoring as a replacement for spot checks of these vital parameters conducted in the general ward.

Unobtrusive pulse rate monitoring by continuous video recording, based on remote photoplethysmography (rPPG), might enable early detection of perioperative arrhythmias in general ward patients. However, the accuracy of an rPPG-based machine learning model to monitor the pulse rate during sinus rhythm and arrhythmias is unknown. We conducted a prospective, observational diagnostic study in a cohort with a high prevalence of arrhythmias (patients undergoing elective electrical cardioversion). Pulse rate was assessed with rPPG via a visible light camera and ECG as reference, before and after cardioversion. A cardiologist categorized ECGs into normal sinus rhythm or arrhythmias requiring further investigation. A supervised machine learning model (support vector machine with Gaussian kernel) was trained using rPPG signal features from 60-s intervals and validated via leave-one-subject-out. Pulse rate measurement performance was evaluated with Bland–Altman analysis. Of 72 patients screened, 51 patients were included in the analyses, including 444 60-s intervals with normal sinus rhythm and 1130 60-s intervals of clinically relevant arrhythmias. The model showed robust discrimination (AUC 0.95 [0.93–0.96]) and good calibration. For pulse rate measurement, the bias and limits of agreement for sinus rhythm were 1.21 [− 8.60 to 11.02], while for arrhythmia, they were − 7.45 [− 35.75 to 20.86]. The machine learning model accurately identified sinus rhythm and arrhythmias using rPPG in real-world conditions. Heart rate underestimation during arrhythmias highlights the need for optimization.

Objective: Contactless monitoring of instantaneous heart rate and respiration rate has a significant clinical relevance. This work aims to use Speckle Vibrometry (i.e., based on the secondary laser speckle effect) to contactlessly measure these two vital signs in an intensive care unit. Methods: In this work, we propose an algorithm for the estimation of instantaneous heart rate and respiration rate from mechanically ventilated patients. The algorithm uses multiple regions, principal component analysis, and dominant angle analysis. A semi-automated peak detection method is implemented to precisely label the aortic valve opening peak within the cardiac waveform. Results: Compared with electrocardiography, the present work achieves limits of agreement of [−2.19, 1.73] beats per minute of instantaneous heart rate. The measurement spot is on the chest covered with two to three layers of duvet blankets. Compared with the airway flow signal measured by the mechanical ventilator, the present work achieves limits of agreement of [−0.68, 0.46] respirations per minute of instantaneous respiration rate. Conclusions: These results showcased Speckle Vibrometry’s potential in vital sign monitoring in a clinical setting. Significance: This is the first human clinical study for Speckle Vibrometry.

Artificial intelligence (AI) is playing an increasingly prominent role in healthcare technology, particularly in patient monitoring and diagnosis. While AI offers significant benefits, it raises concerns about the patient-provider relationship and key care values. To mitigate these risks and align technology with societal values, experts stress the importance of user involvement in technology development. However, input from patients or nurses during AI development remains uncommon. The rapidly digitizing healthcare thus demonstrates a context where technological citizenship is not yet thriving, underscoring the need for improvement in this area. This article examines a case study where nurses contributed to the development of an AI-driven monitoring camera for the cardiothoracic surgery department. Based on interview data, our study shows that involving users, particularly nurses, can lead to the improved innovation of a socio-technical care practice, greater recognition of the importance of user involvement among technology developers, and increased empowerment of nurses with regards to the technology that impacts their work practice. To promote technological citizenship in the evolving healthcare landscape, this article offers recommendations for fostering greater user involvement during the development and implementation of new technologies.

Purpose: Video-based monitoring technologies enable continuous and contactless monitoring of vital signs. This study evaluates the clinical concordance and determinants of performance of contactless video-based heart and respiratory rate monitoring compared with reference standards in a heterogeneous population of critically ill patients. Methods: In this prospective observational study, 35 intensive care unit patients were continuously monitored for 24 h. Video-based heart rate and respiratory rate were compared with the clinical reference standard. Agreement was assessed using Bland–Altman plots, intraclass correlation coefficients (ICC), and error grid analyses. Generalized estimating equations (GEE) identified factors affecting agreement. Results: For heart rate, bias was 2.1 bpm (limits − 33.6 to 37.7), with 81.9% within ± 5 bpm and 99.3% in error grid zones A/B. ICC was 0.43. For respiratory rate, bias was − 2.4 breaths/min (limits − 14.3 to 9.5), with 63.5% within ± 3 breaths and 87.5% in zones A/B. ICC was 0.41. High heart rate, atrial fibrillation, norepinephrine administration, and movement reduced agreement for heart rate; movement reduced agreement for respiratory rate. Conclusion: Video-based monitoring shows promise for detecting abnormal vital signs in critically ill patients, but improved robustness to motion is needed for reliable clinical implementation.