Respiratory Rate (RR) is a vital health indicator, especially in infant monitoring, where early detection of abnormalities or variabilities in RR is crucial. Traditionally, the respiratory rate is extracted using contact-based methods, which, although reliable, can be quite intrusive and stressful for long-term monitoring. This study explores the potential of real-time remote RR monitoring on inexpensive hardware, by comparing three motion-based methods of extracting RR from RGB-camera feed: Pixel Intensity Changes (PIC), Optical Flow (OF), and Eulerian Video Magnification (EVM). The three algorithms were benchmarked using the public AIR-125 dataset, which features videos of infants in various positions, with a focus on their accuracy and computational intensity. The results show that the PIC algorithm slightly outperformed the other two algorithms in both accuracy and computational complexity. However, none of the algorithms managed to replicate the performance of the study which initially proposed the dataset as a benchmark.

The measurement of the heart rate (HR) is of vitalimportance in modern medicine. Advancements in medical technology have resulted in a myriad of techniques to measure and analyze these bio-signals, and the advent of telemedicine and the post-COVID-19 world has placed greater emphasis on contact-free measurement tools.

Previous works have explored several methods of contact-free heart rate measurement. Remote Photo-Plethysmography(rPPG) measures HR from RGB camera streams by detecting and analyzing the frequency of “micro-blushes” in the skin corresponding to pulse; this method can also be used to estimate SpO2. Eulerian video magnification instead attempts to detectsubtle micro-movements caused by the pulse to measure HR aswell as RR. Ballistocardiography tracks longitudinal movements of feature points to estimate HR. However, there is little research on applying these (typically computationally-intensive) algorithms in the context of real-time, low-performance embedded systems.

This paper evaluates the computational requirements of algorithms used in extracting bio-signals using an RGB camera. It focuses on rPPG, a camera based method for extracting heart rate from video. Several rPPG implementations are tested on standard hardware and benchmarked on real-time performance and memory. The experiments were conducted on publicly available datasets. Additionally, region-of-interest selection algorithms are also compared. These results are of much use in developing embedded devices for remote monitoring of bio-signals and provide some insight into algorithms viable for use in real-time contexts.

Heart rate (HR) is a critical indicator of an individual’s health, serving as a key metric for detecting potential cardiac issues. This paper explores a method for real-time heart rate measurement using RGB camera footage, aimed at general health monitoring. The proposed method utilizes a convolutional neural network (CNN) to generate a 3D mesh of the subjects’ facial features. The movement over time of the points in this mesh is used to compute a signal that captures the small pulsatile movements corresponding to the mechanical motion of blood being pumped through the veins. This signal is filtered, and motion sources are separated using principal component analysis (PCA). The most periodic component, the one with the highest frequency, is considered to correspond to the heart rate, and it’s frequency is used to estimate the heart rate. The proposed method is tested using the ECG-Fitness dataset, characterized by challenging environmental conditions such as significant subject motion and dim lighting conditions. Experimental results demonstrate the method’s capability for real-time applications, though further enhancements are needed to improve robustness under difficult environmental conditions.

This thesis report, one of a set of two reports, describes a novel way to detect incidents that could occur in the daily life of the elderly. Unlike most systems already implemented in this field, this system does not use any wearable (positioning) sensors and works off an Single Board Computer (SBC).Independent of both of these systems is a system for reassurance to alleviate distress.

In the past decades, the workload and the pressure on medical personnel has been growing to an unprecedented peak. This is partly due to the ageing population and the increasing capabilities to be independent at an older age, increasing the age people enter nursing homes. This paper focuses on a novel way to detect incidents that could occur in the daily life of the elderly. Unlike most systems already proposed by others, there will be no use of wearable positioning sensors and the system is implemented on an Single Board Computer (SBC). This thesis report is one of a set of two reports discussing the final implementation of the system.

Every year, nearly five million people undergo knee, making it one of the most common orthopaedic procedures. Sport-related injuries and age-related diseases are the major causes that lead to a knee operation. To recover mobility, stability and muscle strength, the person must go through a post-surgical rehabilitation program.
The price of these physiotherapy sessions increases the already high medical cost of the treatment. Additionally, it is a long and repetitive process that might make them lose motivation and discourage them before finishing the whole program. It is also inaccessible for some people, who live far away from a hospital or rehabilitation centre or in areas where the public healthcare system is saturated. As a result, many people abandon their rehabilitation programs and never recover the pre-injury state.
In this project, we propose the use of a game controlled using Inertial Measurement Units to improve the rehabilitation process. The game would reduce the costs of the rehabilitation, make it accessible for everybody and increase the motivation of the patients.

This thesis describes the design and implementation of a subsystem within a larger medical monitoring system. This subsystem will make bidirectional audio communication possible between a patient and their caregivers or family members. The patient in question can be premature babies that have to stay inside of incubators and elderly people that need to be monitored for their health.
The designed subsystem is tasked with creating an audio stream that goes bidirectionally between a speaker with a handsfree functionality and the front end via a server. On the server the audio from the patient should be stored on a hard drive. Whether the audio communication should be stopped or started and whether the audio should be stored will be controlled by the front-end user. Initially the goal was to implement the solution on a Pine singleboard computer (SBC), but due to unforeseen delays of the delivery of the necessary Pine SBC, it was decided that a proof of concept would be developed. The future work in this thesis will discuss the portability of the solution to the Pine SBC.

This report details research, design, and implementation of a distance measuring subsystem for a pre existing patient monitoring platform. This platform is used to monitor vitals and the general situation regarding the elderly, people in care facilities, and babies in NICU’s (Neonatal Intensive Care Units). Every monitoring unit contains a GPS, a touch screen, and a camera. The product is also connected to a Bluetooth speaker for audio in- and output. Everything is designed in such a way that no technical knowledge is required to interact with it. The distance measurement subsystem needs to perform a distance measurement based on optical technology. Research revealed that an accurate and low-cost system can be implemented which is based on images. In the end a single camera feature-based method was implemented. A stereo vision solution was also looked into.

This project aims to explore the possibility of monitoring the vitals of the neonate remotely and without placing any sensors or electrodes on the subjects body using Thermal Imaging.

Non-contact infrared thermal imaging provides a non-invasive, safe and accurate means for screening subjects. COVID-19 (a disease caused by the Severe Acute Respiratory Syndrome Corona Virus 2 (SARS-CoV-2)) has changed how health care workers (HCW) interact with patients. Despite the usage of personal protective equipment (PPE), social distancing protocols, and unique facilities to screen individuals, the increase of infection rates among HCW is high. This has enormous consequences as HCW cannot provide direct patient care, thereby decreasing the workforce’s availability during the pandemic and has fuelled the urgent need for a non-invasive IR system to screen for potential cases of infections. Existing remote screening systems are susceptible to shortcomings caused by location and technique used for measurement, distance-ambient temperature associated inaccuracies and inefficient integration of hardware and software modules, invalidating their utility in the current situation. During this project, a non-invasive screening prototype was developed that mitigates these shortcomings by using an adaptive measurement method with automatic environmental compensation for screening elevated skin temperature and respiratory rate. The prototype is tested and validated in a controlled environment.

Neonatal sepsis is a dangerous non-specific disease in babies, especially neonate/newborns. It is one of the leading causes of neonate mortality rate, because of the difficulty to diagnose, leading to late or false treatment. Previous research has found the promising feature of artificial intelligence or machine learning in solving the problem. After analysing hours of the electronic health record data available, they are able to diagnose sepsis condition on neonates. However, the accuracy and time needed before diagnosis are still concerning considering the risk of mistreated or late diagnosed sepsis cases. In this research, machine learning and thermal imaging technology is used to explore the possibility of predicting sepsis. 57 thermal videos from 26 babies are processed to track the highest skin temperature visible to the thermal camera. The temperature data then is utilized to train and test several machine learning models for predicting sepsis cases. Support Vector Machine (SVM) was found to be the best sepsis predictor using time-series variation of the temperature data as the feature. The model needs 10-30 minutes of thermal recording, 19 minutes in average, to predict sepsis and achieved 82\% accuracy. Simulation also shows the high possibility in increasing the accuracy when more data/thermal videos are available to train the model. High accuracy model with fast reacting sepsis prediction could help doctors precisely treat septic neonates in timely manner, decreasing the mortality rate for sepsis cases.

This report describes the design and implementation of an interface subsystem for a fall detection system using a pressure based floor sensor. The goal of the fall detection system is to detect and alarm when an elderly person has fallen. The subsystem covers communication between hardware and software, using the WiFi protocol MQTT. The communication is fast enough so that it does not limit other subsystem performance and allows for high expandability to accommodate the hardware’s modular design. The interface module receives a probability of fall occurrence from the separate processing module, from which an average is taken over a 30 second time window to avoid false positives. Alarming is done through the use of text-to-speech voice API to make automated phone calls that retrieve user input to confirm or deny that help is on the way. A habit tracker proof of concept is provided that could improve the accuracy of fall detection by checking if the client is showing anomalies during a given day. A user interface is implemented using the Kivy environment. This environment has its own language which allows for separation of functionality and layout, which keeps the interface elements separate and leads to ease of use. Overall, this results in a strong alternative for fall detection that could be used to improve the time an elderly person can live at home safely without the need to move to a nursing home.

This report describes the design and implementation of a fall detection algorithm for a fall detection system using a pressure based floor sensor. The goal of the system is to detect falls and alarm the relevant personnel when an elderly person has fallen. The fall detection algorithm has a strong connection with the interface subsystem, which uses the algorithm as a function. The interface subsystem supplies matrices containing the raw sensor values of the pressure floor. The algorithm has been divided into multiple sub-algorithms. First, pre-processing: data linearization was applied on the raw sensor values and the sensor matrix was processed such that an image formed that looked like the real world scenario. Second, image processing techniques were applied to detect contours. Contours were being tracked through time, and being grouped. The characteristics of the contours and groups were used to classify falls. Tests have been done to validate the behaviour of the algorithm, from which an average false negative ratio of 30% was achieved in a time window of 30 seconds. The created prototype proves that image processing is a viable tool for detecting falls with the use of a pressure-based floor sensor. Overall, this results in a strong alternative for fall detection that could be used to improve the time an elderly person can live at home safely without the need to move to a nursing home

This report is part of a bigger project to design a fall detection system for elderly and describes the hardware of the system. The importance of this research is that elderly can live longer in their home by alerting the correct instances in case of an accident. The hardware consists of a 1x2 meter sensor grid below the surface of the floor. The sensor grid uses the piezoresistivity of linqstat to measure the pressure and the location of the pressured point on the floor. An ESP8266 microprocessor combined with 74HC595 shift registers and a 74HC4067 demultiplexer is used to gather the data and send it to a server. The design is modular allowing easy installation in differently sized or shaped rooms. Each individual sensor grid has an update frequency of 14Hz and uses wifi to connect to a common server. Each module has an average power consumption of 0,74W and a peak power of 1,6W.

This report describes the implementation of a distributed system for an initially standalone application that is to be used in the medical sector. This new system aims to provide scalability and reliability.