ABSTRACT: Background: Malaria is endemic throughout Ghana and continues to be a major public health concern especially among pregnant women and children under the age of five. The Ministry of Health (MoH) estimates that over the past ten years, there have been 2-3 million cases of malaria each year, representing 40 percent of outpatient cases, while severe malaria accounts for 33-36 percent of in-patients. Malaria also accounts for 25 percent of the deaths in children under the age of five (GHS, 2001). Correct identification of risk factors could focus interventions at reducing malaria mortality in children. Demographic Surveillance System (DSS) sites have been established and they generate high quality population based longitudinal health and demographic data. The DSS conduct Verbal Autopsies to determine probable causes of death. Objective: This study examines factors affecting childhood malaria mortality in Northern Ghana, using longitudinal data collected by the Navrongo DSS during the period 1995- 2000. It deals especially with the role of socioeconomic factors (mother’s education, family wealth index based on the possessions and housing characteristics and residence, and possession of bed net) and the demographic characteristics (child’s sex and age, and mother’s age). Design: Secondary data analysis of longitudinal data collected by the Navrongo Health Research Centre. Multinomial logistic regression was used to compare the relative risk in three groups of children i.e. those who died of Malaria and those who died of other causes to those who survived as base. Results: Overall, for the deaths due to malaria, older children (1-5years) had a higher risk (RRR 1.4, 95%CI 1.25-1.57 P <0.0001) of dying compared to the infants. Equally, children born of older mothers (maternal age at birth of child >30 years) had a higher risk (RRR 1.28, 95%CI 1.15-1.42 P <0.0001). However, maternal education and residence had a protective effect, with children born of mothers who had some education (RRR 0.79, 95%CI 0.67-0.93 P=0.004) and residing in urban area (RRR 0.61, 95%CI 0.46-0.82 P=0.001) having a lower risk. Similarly, those children whose families are in the highest wealth index had a lower risk (RRR 0.76, 95%CI 0.63-0.91 P=0.003). Interestingly, the same factors were associated with deaths occurring due to other causes, but with varying degree of association. Whereas sex of child was not associated with malaria deaths, being female offered a lower risk of dying from other causes (RRR 0.9, 95%CI 0.84-0.98 P=0.017). It was observed that children in the older age group (1-5 years) were at higher risk of dying (RRR 1.14, 95%CI 1.05-1.25 P=0.002) just as those born of older mothers (RRR 1.16, 95%CI 1.07-1.26 P <0.0001). Even in this group, maternal education (RRR 0.87, 95%CI 0.76-0.98 P=0.023), a higher wealth index (RR 0.87, 95%CI 0.77-0.99 P=0.032 and RRR 0.63 95%CI 0.54-0.73 P <0.0001 for the two highest categories of wealth indices respectively), and area of residence (RRR 0.67, 95%CI 0.55-0.83 P <0.0001) offered a reduction in the risk of dying. Conclusion: The study identified the risk factors (age and sex of the child and mother’s age, maternal education, wealth and residence of the family) associated with malaria mortality and other causes of death in childhood in northern Ghana and this should help formulate cost effective interventions such as health education.

Neonatal septicaemia and nosocomial infection are major causes of morbidity in neonatal intensive care units (NICU) in both developed and developing countries. This study documents infective episodes at two resource disparate NICUs; the Grantley Stable Neonatal Unit (GSNU), Royal Brisbane and Women’s Hospital (RBWH), Australia 1997-2006 and Danang NICU, Vietnam. The specific aims are: 1. To compare the incidence of neonatal septicaemia at RBWH and Danang NICU. 2. To compare of the epidemiological characteristics of nosocomial infections in the two NICUs. 3. To study risk factors associated with nosocomial infection in the two NICUs (including staff numbers, infant numbers and nursing workload). 4. To review published guidelines on healthcare physical environments and staffing levels and to survey the GSNU & Danang NICU in relation to these. 5. To develop a prospective surveillance system to monitor infection episodes in Danang NICU 6. To recommend strategies for the control of neonatal sepsis in Danang NICU. The GSNU at the RBWH, Australia is a 71 bed neonatal care facility with an established infection surveillance system in a well resource environment. Danang NICU, Vietnam provides care to a high risk neonatal population with no established infection surveillance system in a developing country with more limited resources. This study found the GSNU had a low rate of neonatal septicaemia, during the ten year period from 1997 to 2006 with 253 babies (2.1%) diagnosed with septicaemia. In Danang NICU, in the year 2007 alone, there were 52 cases (2.9%) of septicaemia documented but the incidence of neonatal infection is likely to have underestimated due to (i) the failure to collect blood culture before commencing antibiotic treatment. (ii) Difficulties associated with data retrieval from maternal and neonatal records and (iii) lack of a systematic surveillance system with prospectively collected data. The organisms causing early onset and late onset septicaemia were substantially different in the two NICUs. Group B streptococcus (GBS) and Escherichia coli were the major organisms causing early onset sepsis (EOS) at the GSNU, accounting for 37.8% and 29.7% respectively, whilst coagulase-negative staphylococcus (CONS) was the commonest organism (25.7%) for late onset sepsis (LOS). In Danang NICU, Klebsiella spp was the major pathogen and was responsible for 75% of neonatal septicaemia in both early and late onset sepsis. The isolation of Klebsiella spp in a high proportion of infants at less than 48 hours may suggest nosocomial or maternal route of acquisition. The rate of GBS infections in Danang Hospital, Vietnam, is hard to estimate and interpret due to limited microbiologic data. The anti-serum used to classify the Streptococcal types is not readily available in Danang Hospital. In some cases, organisms that are isolated are suspected to be GBS however, this is not confirmed microbiologically. The study identifies a high prevalence of neonatal septicaemia in Danang NICU. It identifies a close relationship between rates of infection and poor nursing care practices and limited resources. The standard care system used at the GSNU as well as the international literature was studied in relation to infection control practices. This was then used to develop the following recommendations that specifically target a reduction in infection rates in Danang NICU: 1. Implement an infection surveillance system 2. Report regularly from the database and embed discussion of the data into unit management policy 3. Revise neonatal and pathology practice in relation to blood cultures 4. Use shorter duration of antibiotics course based on culture results 5. Modify care practices that expose infants to cross infection 6. Identify specific staff with particular responsibilities in maintenance of standards

This thesis explores the role mobile sensor's mobility model and how it affects surveillance system performance in term of area coverage and detection effectiveness. Several algorithms which are categorized into three types, namely, fully coordinated mobility, fully random mobility and emergent mobility models are discussed with their advantages and limitations. A multi-agent platform to organize mobile sensor nodes, control nodes and actor nodes was implemented. It demonstrated great flexibility and was favourable for its distributed, autonomous and cooperative problem-solving characters. Realistic scenarios which are based on three KheperaIII mobile robots and a model which mimics Waterloo regional airport were used to examine the implementation platform and evaluate performance of different mobility algorithms. Several practical issues related to software configurations and interface library were addressed as by-products. The experimental results from both simulation and real platform show that the area coverage and the detection effectiveness vary with applying different mobility models. Fully coordinated model's super efficiency comes with carefully task planning and high requirements of sensor navigational accuracy. Fully random model is the least efficient in area coverage and detection because of the repetitive searching of each sensor and among sensors. A self-organizing algorithm named anti-flocking which mimics solitary animal's social behaviour was first proposed. It works based on quite simple rules for achieving purposeful coordinated group action without explicit global control. Experimental results demonstrate its attractive target detection efficiency in term of both detection rate and detection time while providing desirable features such as scalability, robustness and adaptivity. From the simulation results, the detection rate of the anti-flocking model increases by 36.5% and average detection time decreases by 46.2% comparing with the fully random motion model. The real platform results also reflect the superior performance improvement.

This thesis explores the role mobile sensor's mobility model and how it affects surveillance system performance in term of area coverage and detection effectiveness. Several algorithms which are categorized into three types, namely, fully coordinated mobility, fully random mobility and emergent mobility models are discussed with their advantages and limitations. A multi-agent platform to organize mobile sensor nodes, control nodes and actor nodes was implemented. It demonstrated great flexibility and was favourable for its distributed, autonomous and cooperative problem-solving characters. Realistic scenarios which are based on three KheperaIII mobile robots and a model which mimics Waterloo regional airport were used to examine the implementation platform and evaluate performance of different mobility algorithms. Several practical issues related to software configurations and interface library were addressed as by-products. The experimental results from both simulation and real platform show that the area coverage and the detection effectiveness vary with applying different mobility models. Fully coordinated model's super efficiency comes with carefully task planning and high requirements of sensor navigational accuracy. Fully random model is the least efficient in area coverage and detection because of the repetitive searching of each sensor and among sensors. A self-organizing algorithm named anti-flocking which mimics solitary animal's social behaviour was first proposed. It works based on quite simple rules for achieving purposeful coordinated group action without explicit global control. Experimental results demonstrate its attractive target detection efficiency in term of both detection rate and detection time while providing desirable features such as scalability, robustness and adaptivity. From the simulation results, the detection rate of the anti-flocking model increases by 36.5% and average detection time decreases by 46.2% comparing with the fully random motion model. The real platform results also reflect the superior performance improvement.

Human behavior analysis is an important challenge in many domains, such as surveillance systems, video content retrieval, human interactive systems, medical diagnosis, etc. With the increasing needs of public safety, intelligent surveillance system becomes an activating issue in computer vision and related research fields. In this thesis we present a method to analyze human behavior in a video sequence with depth information obtained from the depth camera. When interested actions are detected in the scene, the system will trigger alarm information. Contour line and Delaunay triangulation are used to establish human posture model. By traversing the triangulation meshes with the depth first search, we obtain the spanning tree with the depth information, and then construct human posture model with this spanning tree. Posture sequence from video sequence with corresponding posture models can be obtained, and then the posture sequences is clustered into key posture sequence. By querying the key posture sequence, the system can recognize human behavior in real-time and inform users immediately when interested actions detected. Experimental results show that the system is accurate and robust for human behavior recognition.

As our society gets increasingly mobile, it is becoming commonplace for residences to remain vacant for a significant amount of time every day. Unfortunately, emergencies can occur during those time, which may require immediate mitigatory action. This project proposes an approach that allows the resident to be notified of such emergencies and to perform mitigatory actions, even when she is hundreds of miles away. Our infrastructure includes three components: (1) programmable sensor devices to detect emergency situations; (2) a Web service hosted in the resident's home computer to send a notification to the smartphone of the user; and (3) a smartphone app that communicates with this Web service to notify the user, and provides a interface for the user to perform any mitigatory action. We develop a prototype system for detecting fire and intrusion emergencies. Our prototype system uses two sunSPOTs as sensors, an iRobot Create® as a mitigatory device, an Android app for user notification. / text

Thesis (M.A. (Population and Reproductive Health Research))--Mahidol University, 2003.

Thesis (M.A. (Population and Reproductive Health Research))--Mahidol University, 2003.

Thesis (M.A. (Population and Reproductive Health Research))--Mahidol University, 2003.

Thesis (M.A. (Population and Reproductive Health Research))--Mahidol University, 2003.