ORCID Profile
0000-0002-7675-7208
Current Organisations
Austin Health
,
Monash University School of Medicine
,
Peter MacCallum Cancer Centre
Does something not look right? The information on this page has been harvested from data sources that may not be up to date. We continue to work with information providers to improve coverage and quality. To report an issue, use the Feedback Form.
Publisher: AME Publishing Company
Date: 03-2023
DOI: 10.21037/TAU-22-690
Publisher: arXiv
Date: 2022
Publisher: Springer Science and Business Media LLC
Date: 17-06-2021
Publisher: Elsevier BV
Date: 09-2023
Publisher: Elsevier BV
Date: 2023
DOI: 10.1016/J.AUCC.2022.08.011
Abstract: Caregiver workload in the ICU setting is difficult to numerically quantify. Ambient Intelligence utilises computer vision-guided neural networks to continuously monitor multiple datapoints in video feeds, has become increasingly efficient at automatically tracking various aspects of human movement. To assess the feasibility of using Ambient Intelligence to track and quantify allpatient and caregiver activity within a bedspace over the course of an ICU admission and also to establish patient specific factors, and environmental factors such as time ofday, that might contribute to an increased workload in ICU workers. 5000 images were manually annotated and then used to train You Only LookOnce (YOLOv4), an open-source computer vision algorithm. Comparison of patientmotion and caregiver activity was then performed between these patients. The algorithm was deployed on 14 patients comprising 1762800 framesof new, untrained data. There was a strong correlation between the number ofcaregivers in the room and the standardized movement of the patient (p < 0.0001) withmore caregivers associated with more movement. There was a significant difference incaregiver activity throughout the day (p < 0.05), HDU vs. ICU status (p < 0.05), delirious vs. non delirious patients (p < 0.05), and intubated vs. not intubated patients(p < 0.05). Caregiver activity was lowest between 0400 and 0800 (average .71 ± .026caregivers per hour) with statistically significant differences in activity compared to 0800-2400 (p < 0.05). Caregiver activity was highest between 1200 and 1600 (1.02 ± .031 caregivers per hour) with a statistically significant difference in activity comparedto activity from 1600 to 0800 (p < 0.05). The three most dominant predictors of workeractivity were patient motion (Standardized Dominance 78.6%), Mechanical Ventilation(Standardized Dominance 7.9%) and Delirium (Standardized Dominance 6.2%). Ambient Intelligence could potentially be used to derive a single standardized metricthat could be applied to patients to illustrate their overall workload. This could be usedto predict workflow demands for better staff deployment, monitoring of caregiver workload, and potentially as a tool to predict burnout.
Publisher: Wiley
Date: 23-01-2023
DOI: 10.1111/ANS.18248
Publisher: Wiley
Date: 30-07-0088
DOI: 10.1111/BJU.16154
Publisher: Ovid Technologies (Wolters Kluwer Health)
Date: 31-08-2023
Publisher: American Medical Association (AMA)
Date: 19-11-2021
Publisher: MDPI AG
Date: 21-02-2021
DOI: 10.3390/S21041495
Abstract: Infrared thermography for camera-based skin temperature measurement is increasingly used in medical practice, e.g., to detect fevers and infections, such as recently in the COVID-19 pandemic. This contactless method is a promising technology to continuously monitor the vital signs of patients in clinical environments. In this study, we investigated both skin temperature trend measurement and the extraction of respiration-related chest movements to determine the respiratory rate using low-cost hardware in combination with advanced algorithms. In addition, the frequency of medical examinations or visits to the patients was extracted. We implemented a deep learning-based algorithm for real-time vital sign extraction from thermography images. A clinical trial was conducted to record data from patients on an intensive care unit. The YOLOv4-Tiny object detector was applied to extract image regions containing vital signs (head and chest). The infrared frames were manually labeled for evaluation. Validation was performed on a hold-out test dataset of 6 patients and revealed good detector performance (0.75 intersection over union, 0.94 mean average precision). An optical flow algorithm was used to extract the respiratory rate from the chest region. The results show a mean absolute error of 2.69 bpm. We observed a computational performance of 47 fps on an NVIDIA Jetson Xavier NX module for YOLOv4-Tiny, which proves real-time capability on an embedded GPU system. In conclusion, the proposed method can perform real-time vital sign extraction on a low-cost system-on-module and may thus be a useful method for future contactless vital sign measurements.
Publisher: Wiley
Date: 10-08-2022
DOI: 10.1111/ANAE.15834
Abstract: We performed a systematic review and meta‐analysis to identify, classify and evaluate the body of evidence on novel wearable and contactless devices that measure heart rate, respiratory rate and oxygen saturations in the clinical setting. We included any studies of hospital inpatients, including sleep study clinics. Eighty‐four studies were included in the final review. There were 56 studies of wearable devices and 29 of contactless devices. One study assessed both types of device. A high risk of patient selection and rater bias was present in proportionally more studies assessing contactless devices compared with studies assessing wearable devices (p = 0.023 and p 0.0001, respectively). There was high but equivalent likelihood of blinding bias between the two types of studies (p = 0.076). Wearable device studies were commercially available devices validated in acute clinical settings by clinical staff and had more real‐time data analysis (p = 0.04). Contactless devices were more experimental, and data were analysed post‐hoc. Pooled estimates of mean (95%CI) heart rate and respiratory rate bias in wearable devices were 1.25 (−0.31–2.82) beats.min ‐1 (pooled 95% limits of agreement −9.36–10.08) and 0.68 (0.05–1.32) breaths.min ‐1 (pooled 95% limits of agreement −5.65–6.85). The pooled estimate for mean (95%CI) heart rate and respiratory rate bias in contactless devices was 2.18 (3.31–7.66) beats.min ‐1 (pooled limits of agreement −6.71–10.88) and 0.30 (−0.26–0.87) breaths.min ‐1 (pooled 95% limits of agreement −3.94–4.29). Only two studies of wearable devices measured S p O 2 these reported mean measurement biases of 3.54% (limits of agreement −5.65–11.45%) and 2.9% (−7.4–1.7%). Heterogeneity was observed across studies, but absent when devices were grouped by measurement modality and reference standard. We conclude that, while studies of wearable devices were of slightly better quality than contactless devices, in general all studies of novel devices were of low quality, with small ( 100) patient datasets, typically not blinded and often using inappropriate statistical techniques. Both types of devices were statistically equivalent in accuracy and precision, but wearable devices demonstrated less measurement bias and more precision at extreme vital signs. The statistical variability in precision and accuracy between studies is partially explained by differences in reference standards.
Publisher: Elsevier BV
Date: 10-2021
No related grants have been discovered for David Chen.