No description available.

This thesis explores the contactless estimation of people’s vital signs. We designed two camera-based systems and applied object detection algorithms to locate the regions of interest where vital signs are estimated. With the development of Deep Learning, Convolutional Neural Network (CNN) model has many applications in the real world nowadays. We applied the CNN based frameworks to the different types of camera based systems and improve the efficiency of the contactless vital signs estimation. In the field of medical healthcare, contactless monitoring draws a lot attention in the recent years because the wide use of different sensors. However most of the methods are still in the experimental phase and have never been used in real applications. We were interested in monitoring vital signs of patients lying in bed or sitting around the bed at a hospital. This required using sensors that have range of 2 to 5 meters. We developed a system based on the depth camera for detecting people’s chest area and the radar for estimating the respiration signal. We applied a CNN based object detection method to locate the position of the subject lying in the bed covered with blanket. And the respiratory-like signal is estimated from the radar device based on the detected subject’s location. We also create a manually annotated dataset containing 1,320 depth images. In each of the depth image the silhouette of the subject’s upper body is annotated, as well as the class. In addition, a small subset of the depth images also labeled four keypoints for the positioning of people’s chest area. This dataset is built on the data collected from the anonymous patients at the hospital which is substantial. Another problem in the field of human vital signs monitoring is that systems seldom contain the functions of monitoring multiple vital signs at the same time. Though there are few attempting to work on this problem recently, they are still all prototypes and have a lot limitations like shorter operation distance. In this application, we focused on contactless estimating subjects’ temperature, breathing rate and heart rate at different distances with or without wearing the mask. We developed a system based on thermal and RGB camera and also explore the feasibility of CNN based object detection algorithms to detect the vital signs from human faces with specifically defined RoIs based on our thermal camera system. We proposed the methods to estimate respiratory rate and heart rate from the thermal videos and RGB videos. The mean absolute difference (MAE) between the estimated HR using the proposed method and the baseline HR for all subjects at different distances is 4.24 ± 2.47 beats per minute, the MAE between the estimated RR and the reference RR for all subjects at different distances is 1.55 ± 0.78 breaths per minute.

Thesis (Ph. D. in Biomedical Engineering)--Vanderbilt University, May 2006. / Title from title screen. Includes bibliographical references.

No description available.

La mesure cardiaque sans contact réalisée grâce aux méthodes de photopléthysmographie sans contact est un domaine de recherche très actif. Depuis l'introduction en 2010 d'une nouvelle méthode de mesure avec des capteurs optiques d'entrée de gamme (webcam PC), les travaux de recherche se sont multipliés. Ainsi, on observe une plus grande diversité des méthodes proposées afin de réaliser la mesure. Egalement, la précision de la mesure a grandement progressé et les scénarios et possibilités d'usage de la technologie sont aujourd'hui très nombreux. Au cœur de ce processus de mesure, la segmentation dans l'image de la ou les zones d'intérêt est une étape clé. Nous proposons dans cette thèse une méthode innovante afin de réaliser la mesure photoplethysmographique sans contact en identifiant implicitement les zones de peau vivante dans la vidéo. Nous avons montré que notre approche permet d'améliorer la qualité de la mesure en favorisant les zones dans l'image où le signal est de plus grande qualité. Afin de rendre possible l'intégration de notre solution, nous avons proposé une nouvelle méthode de segmentation en superpixels, nommée IBIS, qui permet de réduire la complexité algorithmique de cette étape du traitement. Ce faisant, nous avons démontré la faisabilité de l'intégration de notre solution au sein d'une plateforme embarquée. Les différentes méthodes ont été évaluées au travers de plusieurs expérimentations afin de valider leurs performances. Notre méthode de segmentation en superpixels est comparée aux méthodes de l'état de l'art tandis que nous avons implémenté plusieurs des méthodes de mesure du signal photoplethysmographique afin de discuter de l'impact de notre approche sur la qualité de la mesure photoplethysmographique. Que ce soit pour la segmentation en superpixels ou pour l'estimation du rythme cardiaque sans contact, nous avons montré une importante plus-value de nos méthodes comparées à celles disponibles dans la littérature. Les différents travaux présentés dans ce document ont été valorisés au travers de publications en conférences et revue. / Heart-rate estimation performed with remote photoplethysmography is a very active research field. Since pioneer works in 2010, which demonstrated the feasibility of the measure with low-grade consumers’ camera (webcam), the number of scientific publications have increased significantly in the domain. Hence, we observe a multiplication of the methods in order to retrieve the photoplethysmographic signal which has led to an increased precision and quality of the heart-rate estimation. Region of interest segmentation is a key step of the processing pipeline in order to maximize the quality of the measured signal. We propose a new method to perform remote photoplethysmographic measurement using an implicit living skin identification method. Hence, we have shown that our approach lead to an improvement in both quality of the signal measured and precision of the heart-rate estimation by favoring more contributive area. As we are working with hardware integration constraint, we propose a new superpixels segmentation method which requires significantly less computation power than state of the art methods by reducing the algorithmic complexity of this step. Moreover, we have demonstrated the integration and real time capabilities by implementing our solution to an embedded device. All of our proposed method have been evaluated through different experimentations. Our new segmentation method, called IBIS, have been compared to state of the art methods to quantify the quality of the produced segmentation. To quantify the impact of our approach on the quality of the photoplethysmographic measure, we have implemented and compared state of the art methods with our proposed method. For both the superpixels segmentation and remote heart-rate estimation, our methods have shown great results and advantages compared to state of the art ones. Our works have been reviewed by the scientific community through several conference presentations and journal publications.

Copy of author's previously published work inserted. Includes bibliographical references (leaves 273-283). Aims to identify if the current practice of post-operative vital sign collection detects complications in the first 24 hours after the patient has returned to the general ward setting using a combination of methods within a triangulated approach to data collection.

Includes bibliographical references. Electronic publication: Fulltext available in PDF format; abstract in HTML format. The nursing practice of monitoring patients in the post-operative (PO) phase upon returning to the general ward setting has traditionally consisted of the systematic collection of vital signs and observation of other aspects of the patient's recovery. For the most part the primary focus of this monitoring has been the detection of post-operative complications. There is a need for more substantive evidence to support an appropriate frequency of post-operative observation. The aim of this research was to identify if the current practice of PO vital sign collection detects PO complications in the first 24 hours after the patient has returned to the general ward setting. Due to the complex world in which nurses practice the research was undertaken using a combination of methods within a triangulated approach to collect data. Electronic reproduction.[Australia] :Australian Digital Theses Program,2001.

Bakgrund: Modified Early Warning Score (MEWS) är ett bedömningsinstrument för utvärdering av patienters vitalparametrar. Verktygets funktion är att genom strukturerade observationer identifiera patienter med sviktande vitala funktioner, och således kunna vidta åtgärder innan patienten avlider eller kräver intensivvård. Nyttan av MEWS inom onkologisk heldygnsvård är än så länge dokumenterat i mycket liten utsträckning. En studie från 2012 indikerar att MEWS inte är ett effektivt verktyg för att förutse försämringar hos onkologiska patienter.   Syfte: Att undersöka om patienter inom onkologisk vård uppvisar MEWS 0-17 poäng veckan innan dödsfall. Vidare ämnar författarna vilka specifika parametrar inom MEWS som orsakade förhöjda totalpoäng, samt om några skillnader föreligger mellan män och kvinnor. Metod: Retrospektiv journalgranskningsstudie där dokumentation av MEWS-bedömningar granskades i 70 patientjournaler. Resultat: Undersökningsgruppen hade ett genomsnittligt MEWS på 2,99 under den sista levnadsveckan. Av de enskilda variablerna inom MEWS var andningsfrekvens den som oftast gav poäng ≥1. Det genomsnittliga värdet av MEWS steg från 3,13 poäng det sjunde sista till 8 poäng det sista levnadsdygnet, men på grund av den låga frekvensen av mätningar under det sista levnadsdygnet kan inga slutsatser dras från resultatet. Inga signifikanta skillnader mellan män och kvinnor har identifierats. Låg förekomst av registrerade MEWS hos avlidna onkologpatienter orsakade ett stort initialt bortfall.   Slutsats: Studien indikerar att MEWS används i liten utsträckning på onkologiska patienter innan dödsfall. Det är dock inte möjligt att med detta underlag uttala sig om huruvida MEWS är ett relevant verktyg inom den onkologiska vården, och vidare forskning måste därför göras. / Background: Modified Early Warning Score, MEWS, is a scoring system (0-17 points) for assessment of patients’ vital signs. The function of MEWS is to detect deteriating patients at an early stage, and being able to put in adequate treatment before their physical condition worsens. The benefits from using MEWS on oncology patients have so far been researched to small extent. A study from 2012 indicates that MEWS is not an effective tool for detecting deteriation in oncology patients.   Objective: To research if oncology patients have scores 0-17 on MEWS measurements the week before they are deceased. The writers also intend to research which certain parameters within MEWS caused higher overall scores, and lastly whether there are any differences in scores between men and women.   Methods: A retrospective review of medical records was performed on a total of 104 MEWS measurements belonging to 70 deceased patients. Results: The main results show an average MEWS of 2,99 points during the last week of the patients’ lives. Respiratory rate was the variable within MEWS to cause elevated scores (≥1) most often. The average MEWS increased from 3,13 points the seventh day before death to 8 points the last day before death, but due to the low frequency of measurements from the last day of the patients’ lives, it is not possible to make any assumptions based on these results. Furthermore, the initial loss of patients meeting the inclusion criteria was substantial due to low prevalence of registered MEWS.   Conclusion: The results indicate that MEWS is not used on oncology patients to a great extent. It is not, however, possible to determine whether MEWS is a relevant assessment tool in care of oncology patients, and further research is therefore needed.

Thesis (Ph.D.)--University of Adelaide, Dept. of Clinical Nursing, 2003. / Copy of author's previously published work inserted. Includes bibliographical references (leaves 273-283). Also available electronically.

Thesis (Ph.D.)--University of Adelaide, Dept. of Clinical Nursing, 2003. / Includes bibliographical references. Also available in a print form.