Oliver J. Gibson

Temporally constrained ICA: an application to artifact rejection in electromagnetic brain signal analysis

Independent component analysis (ICA) is a technique which extracts statistically independent components from a set of measured signals. The technique enjoys numerous applications in biomedical signal analysis in the literature, especially in the analysis of electromagnetic (EM) brain signals. Standard implementations of ICA are restrictive mainly due to the square mixing assumption-for signal recordings which have large numbers of channels, the large number of resulting extracted sources makes the subsequent analysis laborious and highly subjective. There are many instances in neurophysiological analysis where there is strong a priori information about the signals being sought; temporally constrained ICA (cICA) can extract signals that are statistically independent, yet which are constrained to be similar to some reference signal which can incorporate such a priori information. We demonstrate this method on a synthetic dataset and on a number of artifactual waveforms identified in multichannel recordings of EEG and MEG. cICA repeatedly converges to the desired component within a few iterations and subjective analysis shows the waveforms to be of the expected morphologies and with realistic spatial distributions. This paper shows that cICA can be applied with great success to EM brain signal analysis, with an initial application in automating artifact extraction in EEG and MEG.

A systematic review of telemedicine interventions to support blood glucose self-monitoring in diabetes

Aims  To evaluate evidence for feasibility, acceptability and cost‐effectiveness of diabetes telemedicine applications.

Chemotherapy side-effect management using mobile phones

Colorectal cancer is a major health problem in developed countries, accounting for a significant proportion of deaths in the population. Advances in chemotherapy treatment have led to therapy being delivered in the home-setting, which presents challenges in ensuring that treatment-related side-effects are detected and reported to clinical staff in an appropriate time-frame. A telemedicine system has been developed using a mobile-phone platform to allow patients to complete symptom diaries which trigger alerts paged to their nurse in the event of severe side-effects. Six patients used this system for two cycles of oral chemotherapy. Two cases of moderate symptoms deteriorating to more severe conditions were observed, and individual self-care and treatment advice were presented to these patients.

Development of a real-time smartphone solution for the management of women with or at high risk of gestational diabetes.

Background: Gestational diabetes mellitus (GDM) is defined as new onset or recognition of glucose intolerance in pregnancy. Evidence supports tight blood glucose regulation to prevent adverse maternal and fetal outcomes. Finger-prick blood glucose (BG) testing with frequent clinic review remains the most common method of managing diabetes in pregnancy. The prevalence of GDM is rising globally, pressuring resource-limited services. Objectives: We have developed an intuitive, interactive, reliable, and accurate management system to record BG measurements and deliver management of GDM remotely. Methods: Following an initial scoping phase, a prototype software application was developed using an Android smartphone with BG meter linkage via Bluetooth. A custom website was built for clinician review of the data transmitted by the smartphone. After system refinement, further evaluation was undertaken for usability and reliability in a 48-patient service development project. Results: Women used the system for an average of 13.1 weeks. In all, 19 686 BG measures were transmitted, 98.6% of which had a meal tag. A total of 466 text messages were transmitted. A mean of 30 BG readings per woman per week were transmitted, and 85% of women submitted the minimum requirement of 18 readings per week. Discussion: We have developed a novel, real-time, smartphone-based BG monitoring management system that allows clinician review of real-time patient-annotated BG results. Results indicate high usage and excellent compliance by women. Conclusion: Robust clinical, economic, and satisfaction evaluations are required. To address these requirements, we are currently conducting a randomized controlled pilot trial.

Stepwise Self-Titration of Oral Glucose-Lowering Medication Using a Mobile Telephone-Based Telehealth Platform in Type 2 Diabetes: A Feasibility Trial in Primary Care

Background: Telehealth-supported clinical interventions may improve diabetes self-management. We explored the feasibility of stepwise self-titration of oral glucose-lowering medication guided by a mobile telephone-based telehealth platform for improving glycemic control in type 2 diabetes. Methods: We recruited 14 type 2 diabetes patients to a one-year feasibility study with 1:1 randomization. Intervention group patients followed a stepwise treatment plan for titration of oral glucose-lowering medication with self-monitoring of glycemia using real-time graphical feedback on a mobile telephone and remote nurse monitoring using a Web-based tool. We carried out an interim analysis at 6 months. Results: We screened 3476 type 2 diabetes patients; 94% of the ineligible did not meet the eligibility criteria for hemoglobin A1c (HbA1c) or current treatment. Mean (standard deviation) patient age at baseline was 58 (11) years, HbA1c was 65 (12) mmol/mol (8.1% [1.1%]), body mass index was 32.9 (6.4) kg/m2, median [interquartile range (IQR)] diabetes duration was 2.6 (0.6 to 4.7) years, and 10 (71%) were men. The median (IQR) change in HbA1c from baseline to six months was −10 (−21 to 3) mmol/mol (−0.9% [-1.9% to 0%]) in the intervention group and −5 (−13 to 6) mmol/mol (−0.5% [-1.2% to 0.6%]) in the control group. Six out of seven intervention group patients and four out of seven control group patients changed their oral glucose-lowering medication (p = .24). Conclusions: Self-titration of oral glucose-lowering medication in type 2 diabetes with self-monitoring and remote monitoring of glycemia is feasible, and further studies using adapted recruitment strategies are required to evaluate whether it improves clinical outcomes.

Acceptability and User Satisfaction of a Smartphone-Based, Interactive Blood Glucose Management System in Women With Gestational Diabetes Mellitus

Background: The increase in gestational diabetes mellitus (GDM) is challenging maternity services. We have developed an interactive, smartphone-based, remote blood glucose (BG) monitoring system, GDm-health. Aims: The objective was to determine women’s satisfaction with using the GDm-health system and their attitudes toward their diabetes care. Methods: In a service development program involving 52 pregnant women (September 2012 to June 2013), BG was monitored using GDm-health from diagnosis until delivery. Following birth, women completed a structured questionnaire assessing (1) general satisfaction, (2) equipment issues, and (3) relationship with the diabetes care team. Responses were scored on a 7-point Likert-type scale. Reliability and validity of the questionnaire were assessed using statistical methods. Results: Of 52 women, 49 completed the questionnaire; 32 had glucose tolerance test confirmed GDM (gestation at recruitment 29 ± 4 weeks (mean ± SD), and 17 women previous GDM recommended for BG monitoring (18 ± 6 weeks). In all, 45 of 49 women agreed their care was satisfactory and the best for them, 47 of 49 and 43 of 49 agreed the equipment was convenient and reliable respectively, 42 of 49 agreed GDm-health fitted into their lifestyle, and 46 of 49 agreed they had a good relationship with their care team. Written comments supported these findings, with very positive reactions from the majority of women. Cronbach’s alpha was .89 with factor analysis corresponding with question thematic trends. Conclusions: This pilot demonstrates that GDm-health is acceptable and convenient for a large proportion of women. Effects on clinical and economic outcomes are currently under investigation in a randomized trial (clinicaltrials.gov NCT01916694).

Trial protocol to compare the efficacy of a smartphone-based blood glucose management system with standard clinic care in the gestational diabetic population

Introduction The prevalence of gestational diabetes mellitus (GDM) is rising in the UK. Good glycaemic control improves maternal and neonatal outcomes. Frequent clinical review of patients with GDM by healthcare professionals is required owing to the rapidly changing physiology of pregnancy and its unpredictable course. Novel technologies that allow home blood glucose (BG) monitoring with results transmitted in real time to a healthcare professional have the potential to deliver good-quality healthcare to women more conveniently and at a lower cost to the patient and the healthcare provider compared to the conventional face-to-face or telephone-based consultation. We have developed an integrated GDm-health management system and aim to test the impact of using this system on maternal glycaemic control, costs, patient satisfaction and maternal and neonatal outcomes compared to standard clinic care in a single large publicly funded (National Health Service (NHS)) maternity unit. Methods and analysis Women with confirmed gestational diabetes in a current pregnancy are individually randomised to either the GDm-health system and half the normal clinic visits or normal clinic care. Primary outcome is mean BG in each group from recruitment to delivery calculated, with adjustments made for number of BG measurements, proportion of preprandial and postprandial readings and length of time in study, and compared between the groups. The secondary objective will be to compare the two groups for compliance to the allocated BG monitoring regime, maternal and neonatal outcomes, glycaemic control using glycated haemoglobin (HbA1c) and other BG metrics, and patient attitudes to care assessed using a questionnaire and resource use. Ethics and dissemination Thresholds for treatment, dietary advice and clinical management are the same in both groups. The results of the study will be published in a peer-reviewed journal and disseminated electronically and in print. Trial registration number NCT01916694; Pre-results.

Personalized alerts for patients with COPD using pulse oximetry and symptom scores

Chronic Obstructive Pulmonary Disease (COPD) is a progressive chronic disease, predicted to become the third leading cause of death by 2030. COPD patients are at risk of sudden and acute worsening of symptoms, reducing the patients quality of life and leading to hospitalization. We present the results of a pilot study with 18 COPD patients using an m-Health system, based on a tablet computer and pulse oximeter, for a period of six months. For prioritizing patients for clinical review, a data-driven approach has been developed which generates personalized alerts using the electronic symptom diary, pulse rate, blood oxygen saturation, and respiratory rate derived from oximetry data. This work examines the advantages of multivariate novelty detection over univariate approaches and shows the benefit of including respiratory rate as a predictor.

GDm-Health: a pilot study demonstrating the feasibility of mobile phone assisted treatment advice and medication adjustment for women with gestational diabetes

To describe change in self‐reported diet and plasma vitamin C, and to examine associations between change in diet and cardiovascular disease risk factors and modelled 10‐year cardiovascular disease risk in the year following diagnosis of Type 2 diabetes.

Pregnancy physiology pattern prediction study (4P study): protocol of an observational cohort study collecting vital sign information to inform the development of an accurate centile-based obstetric early warning score.

Introduction Successive confidential enquiries into maternal deaths in the UK have identified an urgent need to develop a national early warning score (EWS) specifically for pregnant or recently pregnant women to aid more timely recognition, referral and treatment of women who are developing life-threatening complications in pregnancy or the puerperium. Although many local EWS are in use in obstetrics, most have been developed heuristically. No current obstetric EWS has defined the thresholds at which an alert should be triggered using evidence-based normal ranges, nor do they reflect the changing physiology that occurs with gestation during pregnancy. Methods and analysis An observational cohort study involving 1000 participants across three UK sites in Oxford, London and Newcastle. Pregnant women will be recruited at approximately 14 weeks’ gestation and have their vital signs (heart rate, blood pressure, respiratory rate, oxygen saturation and temperature) measured at 4 to 6-week intervals during pregnancy. Vital signs recorded during labour and delivery will be extracted from hospital records. After delivery, participants will measure and record their own vital signs daily for 2 weeks. During the antenatal and postnatal periods, vital signs will be recorded on an Android tablet computer through a custom software application and transferred via mobile internet connection to a secure database. The data collected will be used to define reference ranges of vital signs across normal pregnancy, labour and the immediate postnatal period. This will inform the design of an evidence-based obstetric EWS. Ethics and dissemination The study has been approved by the NRES committee South East Coast–Brighton and Sussex (14/LO/1312) and is registered with the ISRCTN (10838017). All participants will provide written informed consent and can withdraw from the study at any point. All data collected will be managed anonymously. The findings will be disseminated in international peer-reviewed journals and through research conferences.