Noise suppression of proton magnetic resonance spectroscopy improves paediatric brain tumour classification

Proton magnetic resonance spectroscopy (1H-MRS) is increasingly used for clinical brain tumour diagnosis, but suffers from limited spectral quality. This retrospective and comparative study aims at improving paediatric brain tumour classification by performing noise suppression on clinical 1H-MRS. Eighty-three/forty-two children with either an ependymoma (ages 4.6 ± 5.3/9.3 ± 5.4), a medulloblastoma (ages 6.9 ± 3.5/6.5 ± 4.4), or a pilocytic astrocytoma (8.0 ± 3.6/6.3 ± 5.0), recruited from four centres across England, were scanned with 1.5T/3T short-echo-time point-resolved spectroscopy. The acquired raw 1H-MRS was quantified by using Totally Automatic Robust Quantitation in NMR (TARQUIN), assessed by experienced spectroscopists, and processed with adaptive wavelet noise suppression (AWNS). Metabolite concentrations were extracted as features, selected based on multiclass receiver operating characteristics, and finally used for identifying brain tumour types with supervised machine learning. The minority class was oversampled through the synthetic minority oversampling technique for comparison purposes. Post-noise-suppression 1H-MRS showed significantly elevated signal-to-noise ratios (P < .05, Wilcoxon signed-rank test), stable full width at half-maximum (P > .05, Wilcoxon signed-rank test), and significantly higher classification accuracy (P < .05, Wilcoxon signed-rank test). Specifically, the cross-validated overall and balanced classification accuracies can be improved from 81% to 88% overall and 76% to 86% balanced for the 1.5T cohort, whilst for the 3T cohort they can be improved from 62% to 76% overall and 46% to 56%, by applying Naïve Bayes on the oversampled 1H-MRS. The study shows that fitting-based signal-to-noise ratios of clinical 1H-MRS can be significantly improved by using AWNS with insignificantly altered line width, and the post-noise-suppression 1H-MRS may have better diagnostic performance for paediatric brain tumours.

Characterisation of paediatric brain tumours by their MRS metabolite profiles

1H-magnetic resonance spectroscopy (MRS) has the potential to improve the noninvasive diagnostic accuracy for paediatric brain tumours. However, studies analysing large, comprehensive, multicentre datasets are lacking, hindering translation to widespread clinical practice. Single-voxel MRS (point-resolved single-voxel spectroscopy sequence, 1.5 T: echo time [TE] 23-37 ms/135-144 ms, repetition time [TR] 1500 ms; 3 T: TE 37-41 ms/135-144 ms, TR 2000 ms) was performed from 2003 to 2012 during routine magnetic resonance imaging for a suspected brain tumour on 340 children from five hospitals with 464 spectra being available for analysis and 281 meeting quality control. Mean spectra were generated for 13 tumour types. Mann-Whitney U-tests and Kruskal-Wallis tests were used to compare mean metabolite concentrations. Receiver operator characteristic curves were used to determine the potential for individual metabolites to discriminate between specific tumour types. Principal component analysis followed by linear discriminant analysis was used to construct a classifier to discriminate the three main central nervous system tumour types in paediatrics. Mean concentrations of metabolites were shown to differ significantly between tumour types. Large variability existed across each tumour type, but individual metabolites were able to aid discrimination between some tumour types of importance. Complete metabolite profiles were found to be strongly characteristic of tumour type and, when combined with the machine learning methods, demonstrated a diagnostic accuracy of 93% for distinguishing between the three main tumour groups (medulloblastoma, pilocytic astrocytoma and ependymoma). The accuracy of this approach was similar even when data of marginal quality were included, greatly reducing the proportion of MRS excluded for poor quality. Children's brain tumours are strongly characterised by MRS metabolite profiles readily acquired during routine clinical practice, and this information can be used to support noninvasive diagnosis. This study provides both key evidence and an important resource for the future use of MRS in the diagnosis of children's brain tumours.

Using digital tools in clinical, health and social care research: a mixed-methods study of UK stakeholders

OBJECTIVE

The COVID-19 pandemic accelerated changes to clinical research methodology, with clinical studies being carried out via online/remote means. This mixed-methods study aimed to identify which digital tools are currently used across all stages of clinical research by stakeholders in clinical, health and social care research and investigate their experience using digital tools.

DESIGN

Two online surveys followed by semistructured interviews were conducted. Interviews were audiorecorded, transcribed and analysed thematically.

SETTING, PARTICIPANTS

To explore the digital tools used since the pandemic, survey participants (researchers and related staff (n=41), research and development staff (n=25)), needed to have worked on clinical, health or social care research studies over the past 2 years (2020-2022) in an employing organisation based in the West Midlands region of England (due to funding from a regional clinical research network (CRN)). Survey participants had the opportunity to participate in an online qualitative interview to explore their experiences of digital tools in greater depth (n=8).

RESULTS

Six themes were identified in the qualitative interviews: 'definition of a digital tool in clinical research'; 'impact of the COVID-19 pandemic'; 'perceived benefits/drawbacks of digital tools'; 'selection of a digital tool'; 'barriers and overcoming barriers' and 'future digital tool use'. The context of each theme is discussed, based on the interview results.

CONCLUSIONS

Findings demonstrate how digital tools are becoming embedded in clinical research, as well as the breadth of tools used across different research stages. The majority of participants viewed the tools positively, noting their ability to enhance research efficiency. Several considerations were highlighted; concerns about digital exclusion; need for collaboration with digital expertise/clinical staff, research on tool effectiveness and recommendations to aid future tool selection. There is a need for the development of resources to help optimise the selection and use of appropriate digital tools for clinical research staff and participants.

Metabolite profiles of medulloblastoma for rapid and non-invasive detection of molecular disease groups

BACKGROUND

The malignant childhood brain tumour, medulloblastoma, is classified clinically into molecular groups which guide therapy. DNA-methylation profiling is the current classification 'gold-standard', typically delivered 3-4 weeks post-surgery. Pre-surgery non-invasive diagnostics thus offer significant potential to improve early diagnosis and clinical management. Here, we determine tumour metabolite profiles of the four medulloblastoma groups, assess their diagnostic utility using tumour tissue and potential for non-invasive diagnosis using in vivo magnetic resonance spectroscopy (MRS).

METHODS

Metabolite profiles were acquired by high-resolution magic-angle spinning NMR spectroscopy (MAS) from 86 medulloblastomas (from 59 male and 27 female patients), previously classified by DNA-methylation array (WNT (n = 9), SHH (n = 22), Group3 (n = 21), Group4 (n = 34)); RNA-seq data was available for sixty. Unsupervised class-discovery was performed and a support vector machine (SVM) constructed to assess diagnostic performance. The SVM classifier was adapted to use only metabolites (n = 10) routinely quantified from in vivo MRS data, and re-tested. Glutamate was assessed as a predictor of overall survival.

FINDINGS

Group-specific metabolite profiles were identified; tumours clustered with good concordance to their reference molecular group (93%). GABA was only detected in WNT, taurine was low in SHH and lipids were high in Group3. The tissue-based metabolite SVM classifier had a cross-validated accuracy of 89% (100% for WNT) and, adapted to use metabolites routinely quantified in vivo, gave a combined classification accuracy of 90% for SHH, Group3 and Group4. Glutamate predicted survival after incorporating known risk-factors (HR = 3.39, 95% CI 1.4-8.1, p = 0.025).

INTERPRETATION

Tissue metabolite profiles characterise medulloblastoma molecular groups. Their combination with machine learning can aid rapid diagnosis from tissue and potentially in vivo. Specific metabolites provide important information; GABA identifying WNT and glutamate conferring poor prognosis.

FUNDING

Children with Cancer UK, Cancer Research UK, Children's Cancer North and a Newcastle University PhD studentship.

The Internet of Medical Things: Opportunities, Benefits, Challenges and Concerns

In this narrative review, we investigate the potential opportunities and benefits, as well as the challenges and concerns of integrating the Internet of Things in healthcare. The opportunities include enhanced patient monitoring and management, improved efficiency and resource utilization, personalized and precision medicine, empowering patients and promoting self-management, and data-driven decision-making, while the challenges include security and privacy risks, interoperability and integration, regulatory and compliance issues, ethical considerations and impact on healthcare professionals and patients. These challenges must be carefully weighed against the benefits before deployment of the IoMT-enabled services.

How Interoperability Challenges Are Addressed in Healthcare IoT Projects

The rapid development and implementation of Internet of Medical Things has made interoperability a serious challenge. In this scoping review, we provide an overview of the interoperability challenge, as reported in the health literature, and highlight the proposed solutions. After searching between January 2018 and June 2023 in Compendex via Engineering Village and PubMed, we found 18 publications. The interoperability challenges identified were device heterogeneity, system heterogeneity, data standardization, security and safety, system and architecture standard, system and workflow integration and regulatory and compliance requirements. Solutions included ontology approaches, conceptual semantic frameworks, improved standards, design of middleware, and using blockchain technology.

Platform-agnostic electrochemical sensing app and companion potentiostat

Electrochemical sensing is ubiquitous in a number of fields ranging from biosensing, to environmental monitoring through to food safety and battery or corrosion characterisation. Whereas conventional potentiostats are ideal to develop assays in laboratory settings, they are in general, not well-suited for field work due to their size and power requirements. To address this need, a number of portable battery-operated potentiostats have been proposed over the years. However, most open source solutions do not take full advantage of integrated circuit (IC) potentiostats, a rapidly evolving field. This is partly due to the constraining requirements inherent to the development of dedicated interfaces, such as apps, to address and control a set of common electrochemical sensing parameters. Here we propose the PocketEC, a universal app that has all the functionalities to interface with potentiostat ICs through a user defined property file. The versatility of PocketEC, developed with an assay developer mindset, was demonstrated by interfacing it, via Bluetooth, to the ADuCM355 evaluation board, the open-source DStat potentiostat and the Voyager board, a custom-built, small footprint potentiostat based around the LMP91000 chip. The Voyager board is presented here for the first time. Data obtained using a standard redox probe, Ferrocene Carboxylic Acid (FCA) and a silver ion assay using anodic stripping multi-step amperometry were in good agreement with analogous measurements using a bench top potentiostat. Combined with its Voyager board companion, the PocketEC app can be used directly for a number of wearable or portable electrochemical sensing applications. Importantly, the versatility of the app makes it a candidate of choice for the development of future portable potentiostats. Finally, the app is available to download on the Google Play store and the source codes and design files for the PocketEC app and the Voyager board are shared via Creative Commons license (CC BY-NC 3.0) to promote the development of novel portable or wearable applications based on electrochemical sensing.

Clinical Decision Support Systems for Brain Tumour Diagnosis and Prognosis: A Systematic Review

CDSSs are being continuously developed and integrated into routine clinical practice as they assist clinicians and radiologists in dealing with an enormous amount of medical data, reduce clinical errors, and improve diagnostic capabilities. They assist detection, classification, and grading of brain tumours as well as alert physicians of treatment change plans. The aim of this systematic review is to identify various CDSSs that are used in brain tumour diagnosis and prognosis and rely on data captured by any imaging modality. Based on the 2020 preferred reporting items for systematic reviews and meta-analyses (PRISMA) protocol, the literature search was conducted in PubMed and Engineering Village Compendex databases. Different types of CDSSs identified through this review include Curiam BT, FASMA, MIROR, HealthAgents, and INTERPRET, among others. This review also examines various CDSS tool types, system features, techniques, accuracy, and outcomes, to provide the latest evidence available in the field of neuro-oncology. An overview of such CDSSs used to support clinical decision-making in the management and treatment of brain tumours, along with their benefits, challenges, and future perspectives has been provided. Although a CDSS improves diagnostic capabilities and healthcare delivery, there is lack of specific evidence to support these claims. The absence of empirical data slows down both user acceptance and evaluation of the actual impact of CDSS on brain tumour management. Instead of emphasizing the advantages of implementing CDSS, it is important to address its potential drawbacks and ethical implications. By doing so, it can promote the responsible use of CDSS and facilitate its faster adoption in clinical settings.

Interoperable E-Health System Using Structural and Semantic Interoperability Approaches in CAREPATH

Technical and semantic interoperability are broadly used components of interoperability technology in healthcare. Technical Interoperability provides interoperability interfaces to enable data exchange within different healthcare systems, despite any underlying heterogeneity. Semantic interoperability make different healthcare systems understand and interpret the meaning of the data that is exchanged, by using and mapping standardized terminologies, coding systems, and data models to describe the concept and structure of data. We propose a solution using Semantic and Structural Mapping techniques within CAREPATH; a research project designed to develop ICT solutions for the care management of elderly multimorbid patients with mild cognitive impairment or mild dementia. Our technical interoperability solution supplies a standard-based data exchange protocol to enable information exchange between local care systems and CAREPATH components. Our semantic interoperability solution supplies programmable interfaces, in order to semantically mediate different clinical data representation formats and incorporating data format and terminology mapping features. The solution offers a more reliable, flexible and resource efficient method across EHRs.

MedSecurance Project: Advanced Security-for-Safety Assurance for Medical Device IoT (IoMT)

The MedSecurance project focus on identifying new challenges in cyber security with focus on hardware and software medical devices in the context of emerging healthcare architectures. In addition, the project will review best practice and identify gaps in the guidance, particularly the guidance stipulated by the medical device regulation and directives. Finally, the project will develop comprehensive methodology and tooling for the engineering of trustworthy networks of inter-operating medical devices, that shall have security-for-safety by design, with a strategy for device certification and certifiable dynamic network composition, ensuring that patient safety is safeguarded from malicious cyber actors and technology "accidents".

Applying machine learning classifiers to automate quality assessment of paediatric dynamic susceptibility contrast (DSC-) MRI data

OBJECTIVE

Investigate the performance of qualitative review (QR) for assessing dynamic susceptibility contrast (DSC-) MRI data quality in paediatric normal brain and develop an automated alternative to QR.

METHODS

1027 signal-time courses were assessed by Reviewer 1 using QR. 243 were additionally assessed by Reviewer 2 and % disagreements and Cohen's κ (κ) were calculated. The signal drop-to-noise ratio (SDNR), root mean square error (RMSE), full width half maximum (FWHM) and percentage signal recovery (PSR) were calculated for the 1027 signal-time courses. Data quality thresholds for each measure were determined using QR results. The measures and QR results trained machine learning classifiers. Sensitivity, specificity, precision, classification error and area under the curve from a receiver operating characteristic curve were calculated for each threshold and classifier.

RESULTS

Comparing reviewers gave 7% disagreements and κ = 0.83. Data quality thresholds of: 7.6 for SDNR; 0.019 for RMSE; 3 s and 19 s for FWHM; and 42.9 and 130.4% for PSR were produced. SDNR gave the best sensitivity, specificity, precision, classification error and area under the curve values of 0.86, 0.86, 0.93, 14.2% and 0.83. Random forest was the best machine learning classifier, giving sensitivity, specificity, precision, classification error and area under the curve of 0.94, 0.83, 0.93, 9.3% and 0.89.

CONCLUSION

The reviewers showed good agreement. Machine learning classifiers trained on signal-time course measures and QR can assess quality. Combining multiple measures reduces misclassification.

ADVANCES IN KNOWLEDGE

A new automated quality control method was developed, which trained machine learning classifiers using QR results.

Assessment of the Effectiveness, Socio-Economic Impact and Implementation of a Digital Solution for Patients with Advanced Chronic Diseases: The ADLIFE Study Protocol

Due to population ageing and medical advances, people with advanced chronic diseases (ACD) live longer. Such patients are even more likely to face either temporary or permanent reduced functional reserve, which typically further increases their healthcare resource use and the burden of care on their caregiver(s). Accordingly, these patients and their caregiver(s) may benefit from integrated supportive care provided via digitally supported interventions. This approach may either maintain or improve their quality of life, increase their independence, and optimize the healthcare resource use from early stages. ADLIFE is an EU-funded project, aiming to improve the quality of life of older people with ACD by providing integrated personalized care via a digitally enabled toolbox. Indeed, the ADLIFE toolbox is a digital solution which provides patients, caregivers, and health professionals with digitally enabled, integrated, and personalized care, supporting clinical decisions, and encouraging independence and self-management. Here we present the protocol of the ADLIFE study, which is designed to provide robust scientific evidence on the assessment of the effectiveness, socio-economic, implementation, and technology acceptance aspects of the ADLIFE intervention compared to the current standard of care (SoC) when applied in real-life settings of seven different pilot sites across six countries. A quasi-experimental trial following a multicenter, non-randomized, non-concurrent, unblinded, and controlled design will be implemented. Patients in the intervention group will receive the ADLIFE intervention, while patients in the control group will receive SoC. The assessment of the ADLIFE intervention will be conducted using a mixed-methods approach.

"It depends on the people!" - A qualitative analysis of contextual factors, prior to the implementation of digital health innovations for chronic condition management, in a German integrated care network

Objective

Integrated care and digital health technology interventions are promising approaches to coordinate services for people living with chronic conditions, across different care settings and providers. The EU-funded ADLIFE project intends to provide digitally integrated personalized care to improve and maintain patients' health with advanced chronic conditions. This study conducted a qualitative assessment of contextual factors prior to the implementation of the ADLIFE digital health platforms at the German pilot site. The results of the assessment are then used to derive recommendations for action for the subsequent implementation, and for evaluation of the other pilot sites.

Methods

Qualitative interviews with healthcare professionals and IT experts were conducted at the German pilot site. The interviews followed a semi-structured interview guideline, based on the HOT-fit framework, focusing on organizational, technological, and human factors. All interviews were audio recorded, transcribed, and subsequently analysed following qualitative content analysis.

Results

The results of the 18 interviews show the interviewees' high openness and motivation to use new innovative digital solutions, as well as an apparent willingness of cooperation between different healthcare professionals. Challenges include limited technical infrastructure and large variability of software to record health data, lacking standards and interfaces.

Conclusions

Considering contextual factors on different levels is critical for the success of implementing innovations in healthcare and the transfer into other settings. In our study, the HOT-fit framework proved suitable for assessing contextual factors, when implementing IT innovations in healthcare. In a next step, the methodological approach will be transferred to the six other European pilot sites, participating in the project, for a cross-national assessment of contextual factors.

Informatics Opportunities and Challenges in Medical Imaging: A Journey

The role of the field of informatics in medical imaging is vital; novel or adapted informatics' core methods can be employed to realise innovative information processing and engineering of medical images. As such, imaging informatics can assist in the interpretation of image-based, clinically recorded evidence. This, in turn, leads to the generation of associated actionable knowledge to achieve precision medicine practice. The discipline of informatics has the power to transform data to useful clinical information patterns of observable evidence and, subsequently to generate actionable knowledge in terms of diagnosis, prognosis, and disease management. This paper presents the author's personal viewpoint and distinct contributions to innovations in the acquisition and collection of imaging data; storage, retrieval, and management of imaging information objects; quantitative analysis, classification, and dissemination of imaging observable evidence.

Multifunctional Composite Hydrogels for Bacterial Capture, Growth/Elimination, and Sensing Applications

Hydrogels are cross-linked networks of hydrophilic polymer chains with a three-dimensional structure. Owing to their unique features, the application of hydrogels for bacterial/antibacterial studies and bacterial infection management has grown in importance in recent years. This trend is likely to continue due to the rise in bacterial infections and antimicrobial resistance. By exploiting their physicochemical characteristics and inherent nature, hydrogels have been developed to achieve bacterial capture and detection, bacterial growth or elimination, antibiotic delivery, or bacterial sensing. Traditionally, the development of hydrogels for bacterial/antibacterial studies has focused on achieving a single function such as antibiotic delivery, antibacterial activity, bacterial growth, or bacterial detection. However, recent studies demonstrate the fabrication of multifunctional hydrogels, where a single hydrogel is capable of performing more than one bacterial/antibacterial function, or composite hydrogels consisting of a number of single functionalized hydrogels, which exhibit bacterial/antibacterial function synergistically. In this review, we first highlight the hydrogel features critical for bacterial studies and infection management. Then, we specifically address unique hydrogel properties, their surface/network functionalization, and their mode of action for bacterial capture, adhesion/growth, antibacterial activity, and bacterial sensing, respectively. Finally, we provide insights into different strategies for developing multifunctional hydrogels and how such systems can help tackle, manage, and understand bacterial infections and antimicrobial resistance. We also note that the strategies highlighted in this review can be adapted to other cell types and are therefore likely to find applications beyond the field of microbiology.

A Distributed Cancer Care Model with a Technology-Driven Hub-and-Spoke and further Spoke Hierarchy: Findings from a Pilot Implementation Programme in Kerala, India

BACKGROUND

The technology enabled distributed model in Kerala is based on an innovative partnership model between Karkinos Healthcare and private health centers. The model is designed to address the barriers to cancer screening by generating demand and by bringing together the private health centers and service providers at various levels to create a network for continued care. This paper describes the implementation process and presents some preliminary findings.  Methods: The model follows the hub-and-spoke and further spoke framework. In the pilot phases, from July 2021 to December 2021, five private health centers (partners) collaborated with Karkinos Healthcare across two districts in Kerala. Screening camps were organized across the districts at the community level where the target groups were administered a risk assessment questionnaire followed by screening tests at the spoke hospitals based on a defined clinical protocol. The screened positive patients were examined further for confirmatory diagnosis at the spoke centers. Patients requiring chemotherapy or minor surgeries were treated at the spokes. For radiation therapy and complex surgeries the patients were referred to the hubs.

RESULTS

A total of 2,459 individuals were screened for cancer at the spokes and 299 were screened positive. Capacity was built at the spokes for cancer surgery and chemotherapy. A total of 189 chemotherapy sessions and 17 surgeries were performed at the spokes for cancer patients. 70 patients were referred to the hub.

CONCLUSION

Initial results demonstrate the ability of the technology Distributed Cancer Care Network (DCCN) system to successfully screen and detect cancer and to converge the actions of various private health facilities towards providing a continuum of cancer care. The lessons learnt from this study will be useful for replicating the process in other States.

An Integrated Care Platform System (C3-Cloud) for Care Planning, Decision Support, and Empowerment of Patients With Multimorbidity: Protocol for a Technology Trial

Background

There is an increasing need to organize the care around the patient and not the disease, while considering the complex realities of multiple physical and psychosocial conditions, and polypharmacy. Integrated patient-centered care delivery platforms have been developed for both patients and clinicians. These platforms could provide a promising way to achieve a collaborative environment that improves the provision of integrated care for patients via enhanced information and communication technology solutions for semiautomated clinical decision support.

Objective

The Collaborative Care and Cure Cloud project (C3-Cloud) has developed 2 collaborative computer platforms for patients and members of the multidisciplinary team (MDT) and deployed these in 3 different European settings. The objective of this study is to pilot test the platforms and evaluate their impact on patients with 2 or more chronic conditions (diabetes mellitus type 2, heart failure, kidney failure, depression), their informal caregivers, health care professionals, and, to some extent, health care systems.

Methods

This paper describes the protocol for conducting an evaluation of user experience, acceptability, and usefulness of the platforms. For this, 2 “testing and evaluation” phases have been defined, involving multiple qualitative methods (focus groups and surveys) and advanced impact modeling (predictive modeling and cost-benefit analysis). Patients and health care professionals were identified and recruited from 3 partnering regions in Spain, Sweden, and the United Kingdom via electronic health record screening.

Results

The technology trial in this 4-year funded project (2016-2020) concluded in April 2020. The pilot technology trial for evaluation phases 3 and 4 was launched in November 2019 and carried out until April 2020. Data collection for these phases is completed with promising results on platform acceptance and socioeconomic impact. We believe that the phased, iterative approach taken is useful as it involves relevant stakeholders at crucial stages in the platform development and allows for a sound user acceptance assessment of the final product.

Conclusions

Patients with multiple chronic conditions often experience shortcomings in the care they receive. It is hoped that personalized care plan platforms for patients and collaboration platforms for members of MDTs can help tackle the specific challenges of clinical guideline reconciliation for patients with multimorbidity and improve the management of polypharmacy. The initial evaluative phases have indicated promising results of platform usability. Results of phases 3 and 4 were methodologically useful, yet limited due to the COVID-19 pandemic.

Trial Registration

ClinicalTrials.gov NCT03834207; https://clinicaltrials.gov/ct2/show/NCT03834207

International Registered Report Identifier (IRRID)

RR1-10.2196/21994

A Method for the Classification of Digital Health Architectures as Medical Devices; a Digital Health Research Perspective

It is typical for many digital health research projects to develop IT architectures that will implement integrated care services that may also deliver interventions. As part of compliance with the requirements of the regulation, the components that are considered as a medical device will need to be classified to a medical device category. This is often seen as task that may increase the business risk and a major barrier of the project, particularly during the earlier stages when not all information is available. The paper offers a method assisting with classification of such architectures in the context of the Medical Devices Rregulation, offering a structured way to identifying how the initial deliverables of a project can be used to provide assurance to the justification of the classification.

CAREPATH: Developing Digital Integrated Care Solutions for Multimorbid Patients with Dementia

CAREPATH project is focusing on providing an integrated solution for sustainable care for multimorbid elderly patients with dementia or mild cognitive impairment. The project has a digitally enhanced integrated patient-centered care approach clinical decision and associated intelligent tools with the aim to increase patients' independence, quality of life and intrinsic capacity. In this paper, the conceptual aspects of the CAREPATH project, in terms of technical and clinical requirements and considerations, are presented.

Electronic Patient Reported Outcomes for Miscarriage Research in Tommy's Net

The Tommy's National Centre for Miscarriage Research aims to support the diagnosis and treatment for couples suffering from recurrent miscarriage. Tommy's Net is an electronic data gathering tool, collecting miscarriage data and links with hospital Clinical Information System databases. The gathering of patient reported data is an important aspect, especially as data relating to pregnancy and miscarriage events are often left unreported.

METHODS

Both traditional paper-based and electronic patient reported outcome (ePRO) solutions have been explored to improve response rates, minimize data redundancy and reduce burden on staff. Popular ePRO survey solutions have been compared, including REDCap, SurveyMonkey, Qualtrics and LimeSurvey.

RESULTS

LimeSurvey was selected as the most appropriate solution as it provided self-hosting capability, SMS integration and ease of use.

CONCLUSION

We have implemented a LimeSurvey based ePRO system for collection of baseline and follow-up data for participants on the Tommy's study.

Structural and Semantic Mapping of Application Programming Interfaces

Modern healthcare providers rely upon Electronic Healthcare Records (EHR) systems to record patient data inside their own organization. Some healthcare providers share this data to facilitate patient care with other providers. Medical devices and healthcare providers can use differing standards of recording healthcare information. The Structural and Semantic Mapper Proxy API solution offers a practical way to tackles the issues of Structural and Semantic mapping of Application Programing Interfaces (API) in a healthcare context to enable connection of all existing systems to a healthcare providers EHR creating a single source of truth regarding the treatment of patients and enabling healthcare providers to bridge the gap between external EHR systems.

Digital Inequalities in Cancer Care Delivery in India: An Overview of the Current Landscape and Recommendations for Large-Scale Adoption

Introduction

COVID-19 pandemic has caused major disruptions to delivery of various cancer care services as efforts were put to control the outbreak of the pandemic. Although the pandemic has highlighted the inadequacies of the system but has also led to emergence of a new cancer care delivery model which relies heavily on digital mediums. Digital health is not only restricted to virtual dissemination of information and consultation but has provided additional benefits ranging from support to cancer screening, early and more accurate diagnosis to increasing access to specialized care. This paper evaluates the challenges in the adoption of digital technologies to deliver cancer care services and provides recommendation for large-scale adoption in the Indian healthcare context.

Methods

We performed a search of PubMed and Google Scholar for numerous terms related to adoption of digital health technologies for cancer care during pandemic. We also analyze various socio-ecological challenges—from individual to community, provider and systematic level—for digital adoption of cancer care service which have existed prior to pandemic and lead to digital inequalities.

Results

Despite encouraging benefits accruing from the adoption of digital health key challenges remain for large scale adoption. With respect to user the socio-economic characteristics such as age, literacy and socio-cultural norms are the major barriers. The key challenges faced by providers include regulatory issues, data security and the inconvenience associated with transition to a new system.

Policy Summary

For equitable digital healthcare, the need is to have a participatory approach of all stakeholders and urgently addressing the digital divide adequately. Sharing of health data of public and private hospitals, within the framework of the Indian regulations and Data Protection Act, is critical to the development of digital health in India and it can go a long way in better forecasting and managing cancer burden.

Bottom-Up Natural Language Processing Based Evaluation of the Fitness of UMLS as a Semantic Source for a Computer Interpretable Guidelines Ontology

BACKGROUND

CIGs languages consist of approach specific concepts. More widely used concepts, such as those in UMLS are not typically used.

OBJECTIVE

An evaluation of UMLS concept sufficiency for CIG definition.

METHOD

A popular guideline is mapped to UMLS concepts with NLP. Results are reviewed to evaluate gaps, and appropriateness.

RESULTS

A significant number of the guideline text mapped to UMLS concepts.

CONCLUSIONS

The approach has shown promise and highlighted further challenges.

IMG-10. Determining brain tumour grade non-invasively using a simplified MRI perfusion protocol: single-bolus, leakage-corrected dynamic susceptibility-contrast MRI

INTRODUCTION: Perfusion is associated with grade and survival in children’s brain tumours. Dynamic susceptibility-contrast (DSC-) MRI measures perfusion non-invasively, estimating relative cerebral blood volume (rCBV). We previously showed significant differences between pre-treatment rCBV in low- and high-grade tumours in a multicentre study. Contrast agent leakage from tumour vessels during acquisition affects rCBV accuracy. A contrast agent pre-bolus can be given but this can be challenging in a clinical environment, introducing variability. Alternatively, a single bolus can be administered with leakage correction applied when processing the data. We investigated pre-treatment rCBV values in a multicentre study without pre-bolus administration. METHODS: Thirty-six patients underwent pre-treatment DSC-MRI scans at 2 centres on 4 different scanners. Protocols were variable. Pixel-by-pixel contrast agent concentration time courses were analysed. Maps of uncorrected (rCBVuncorr) and leakage-corrected rCBV (rCBVcorr) were produced. Whole-tumour regions-of-interest were defined and median whole-tumour DSC-MRI parameters calculated. Patients subsequently underwent surgery / biopsy. Tumours were classified and graded. RESULTS: Twelve tumours were classified as low-grade; 24 as high-grade. Median whole-tumour rCBVuncorr was significantly higher in high-grade tumours than in low-grade tumours (1.628 vs -0.167, p&lt;0.001). Median rCBV significantly increased in low-grade tumours following leakage correction (-0.167 to 1.072, p=0.007); there was no significant change for high-grade tumours. Using the median rCBVuncorr of 1.19 to differentiate between low- and high-grade tumours resulted in sensitivity and specificity of 75% and 100%, respectively; median 1.45 for rCBVcorr resulted in sensitivity and specificity of 67% and 100%, respectively. DISCUSSION: DSC-MRI measures of perfusion can distinguish between low- and high-grade paediatric brain tumours. Contrast agent leakage correction is essential for accurate measurement of rCBV. This is robust across multiple centres despite using multiple protocols. CONCLUSION: Pre-treatment multicentre perfusion MRI acquired with a single-bolus and contrast agent leakage correction can aid with differentiating between high- and low-grade paediatric brain tumours.

Metabolite selection for machine learning in childhood brain tumour classification

MRS can provide high accuracy in the diagnosis of childhood brain tumours when combined with machine learning. A feature selection method such as principal component analysis is commonly used to reduce the dimensionality of metabolite profiles prior to classification. However, an alternative approach of identifying the optimal set of metabolites has not been fully evaluated, possibly due to the challenges of defining this for a multi-class problem. This study aims to investigate metabolite selection from in vivo MRS for childhood brain tumour classification. Multi-site 1.5 T and 3 T cohorts of patients with a brain tumour and histological diagnosis of ependymoma, medulloblastoma and pilocytic astrocytoma were retrospectively evaluated. Dimensionality reduction was undertaken by selecting metabolite concentrations through multi-class receiver operating characteristics and compared with principal component analysis. Classification accuracy was determined through leave-one-out and k-fold cross-validation. Metabolites identified as crucial in tumour classification include myo-inositol (P < 0.05, AUC = 0 . 81 ± 0 . 01 ), total lipids and macromolecules at 0.9 ppm (P < 0.05, AUC = 0 . 78 ± 0 . 01 ) and total creatine (P < 0.05, AUC = 0 . 77 ± 0 . 01 ) for the 1.5 T cohort, and glycine (P < 0.05, AUC = 0 . 79 ± 0 . 01 ), total N-acetylaspartate (P < 0.05, AUC = 0 . 79 ± 0 . 01 ) and total choline (P < 0.05, AUC = 0 . 75 ± 0 . 01 ) for the 3 T cohort. Compared with the principal components, the selected metabolites were able to provide significantly improved discrimination between the tumours through most classifiers (P < 0.05). The highest balanced classification accuracy determined through leave-one-out cross-validation was 85% for 1.5 T 1 H-MRS through support vector machine and 75% for 3 T 1 H-MRS through linear discriminant analysis after oversampling the minority. The study suggests that a group of crucial metabolites helps to achieve better discrimination between childhood brain tumours.

Dynamic susceptibility-contrast magnetic resonance imaging with contrast agent leakage correction aids in predicting grade in pediatric brain tumours: a multicenter study

BACKGROUND

Relative cerebral blood volume (rCBV) measured using dynamic susceptibility-contrast MRI can differentiate between low- and high-grade pediatric brain tumors. Multicenter studies are required for translation into clinical practice.

OBJECTIVE

We compared leakage-corrected dynamic susceptibility-contrast MRI perfusion parameters acquired at multiple centers in low- and high-grade pediatric brain tumors.

MATERIALS AND METHODS

Eighty-five pediatric patients underwent pre-treatment dynamic susceptibility-contrast MRI scans at four centers. MRI protocols were variable. We analyzed data using the Boxerman leakage-correction method producing pixel-by-pixel estimates of leakage-uncorrected (rCBVuncorr) and corrected (rCBVcorr) relative cerebral blood volume, and the leakage parameter, K2. Histological diagnoses were obtained. Tumors were classified by high-grade tumor. We compared whole-tumor median perfusion parameters between low- and high-grade tumors and across tumor types.

RESULTS

Forty tumors were classified as low grade, 45 as high grade. Mean whole-tumor median rCBVuncorr was higher in high-grade tumors than low-grade tumors (mean ± standard deviation [SD] = 2.37±2.61 vs. -0.14±5.55; P<0.01). Average median rCBV increased following leakage correction (2.54±1.63 vs. 1.68±1.36; P=0.010), remaining higher in high-grade tumors than low grade-tumors. Low-grade tumors, particularly pilocytic astrocytomas, showed T1-dominant leakage effects; high-grade tumors showed T2*-dominance (mean K2=0.017±0.049 vs. 0.002±0.017). Parameters varied with tumor type but not center. Median rCBVuncorr was higher (mean = 1.49 vs. 0.49; P=0.015) and K2 lower (mean = 0.005 vs. 0.016; P=0.013) in children who received a pre-bolus of contrast agent compared to those who did not. Leakage correction removed the difference.

CONCLUSION

Dynamic susceptibility-contrast MRI acquired at multiple centers helped distinguish between children's brain tumors. Relative cerebral blood volume was significantly higher in high-grade compared to low-grade tumors and differed among common tumor types. Vessel leakage correction is required to provide accurate rCBV, particularly in low-grade enhancing tumors.

A method for machine learning generation of realistic synthetic datasets for validating healthcare applications

Digital health applications can improve quality and effectiveness of healthcare, by offering a number of new tools to users, which are often considered a medical device. Assuring their safe operation requires, amongst others, clinical validation, needing large datasets to test them in realistic clinical scenarios. Access to datasets is challenging, due to patient privacy concerns. Development of synthetic datasets is seen as a potential alternative. The objective of the paper is the development of a method for the generation of realistic synthetic datasets, statistically equivalent to real clinical datasets, and demonstrate that the Generative Adversarial Network (GAN) based approach is fit for purpose. A generative adversarial network was implemented and trained, in a series of six experiments, using numerical and categorical variables, including ICD-9 and laboratory codes, from three clinically relevant datasets. A number of contextual steps provided the success criteria for the synthetic dataset. A synthetic dataset that exhibits very similar statistical characteristics with the real dataset was generated. Pairwise association of variables is very similar. A high degree of Jaccard similarity and a successful K-S test further support this. The proof of concept of generating realistic synthetic datasets was successful, with the approach showing promise for further work.

P10 Rapid capture of uropathogenic bacteria and on-chip determination of antimicrobial resistance

Background

Urinary tract infection (UTI) is one of the most common bacterial infections responsible for increased annual incidence of antimicrobial resistance (AMR) cases. Clinical diagnosis of UTI AMR relies heavily on conventional urine culture and antibiotic susceptibility testing (AST) which has a turnaround time of ∼3 days. Often, irrespective of the infection status, antibiotics are prescribed to patients even before the test results are available, leading to non-judicious use of antibiotics. Over the years, several technologies have been developed for the rapid detection and diagnosis of UTI AMR, however, most of them are limited to traditional microbiological techniques and large laboratory equipment that are not readily available in low-to-middle income countries (LMICs). To address these diagnostic limitations, we are developing a rapid and affordable UTI-AMR diagnostic microfluidic device that is clinical friendly aimed at improving UTI management and AMR stewardship.

Results

Our device enables the flow of a large volume of urine specimens for the capture/enrichment of uropathogenic bacteria and determination of AST via a porous membrane that is augmented with a multifunctional polymer-based material. Important objectives for the development of UTI AMR diagnostic microfluidic device are: (i) development of a multifunctional polymer-based material; and (ii) validation of UTI AMR diagnostic device. We have successfully developed a polysaccharide-based platform to (i) selectively capture uropathogenic bacteria from urine specimen by immobilizing concanavalin A (con A) lectin as bacterial capture agent on the polymer surface via chemical modification; (ii) encapsulate and release bacterial nutrient media and antibiotics for AST; and (iii) detect AST via encapsulation of bacterial growth indicator. In addition, we have also determined the development of methacrylate-based and acrylamide-based synthetic polymer-based material for our application. Further, we have demonstrated the uniform augmentation of the polysaccharide-based polymer onto porous membrane via dip-coating technique for on-chip bacterial capture/enrichment and AST in fluid (urine) flow conditions. The porous membrane is a conducting material which enables us to perform electrochemical measurements such as impedance spectroscopy that accelerates the detection process of antibiotic susceptibility. As a proof-of-concept, we have determined the capture of biosafety level I Escherichia coli expressing kanamycin resistance gene on chemically surface modified polysaccharide-based polymer containing con A and the antibiotic susceptibility of captured bacteria against different antibiotics with and without the porous membrane. We have quantitatively determined the limit of detection of E. coli on multifunctional polysaccharide-based polymer material.

Conclusions

The utility of the UTI AMR microfluidic device in clinical settings enables clinicians to make informed decisions on the most appropriate antibiotic for treatment in less than a day. Integration of impedance spectroscopy will further accelerate the detection by significantly reducing the time of detection. Further, the device allows for off-chip analysis by retrieving the captured uropathogenic bacteria to perform high throughput sequencing for identifying AMR genetic determinants. Therefore, with the ability to selectively capture uropathogenic bacteria and determine AST in a short time, our technology has the potential to overcome some of the current limitations in UTI AMR diagnostics.

Quantitative assessment of pregnancy outcome following recurrent miscarriage clinic care: a prospective cohort study

OBJECTIVES

To measure pregnancy outcome following attendance at a recurrent miscarriage service and identify factors that influence outcome.

DESIGN

Prospective, observational electronic cohort study.

SETTING

Participants attending a specialist recurrent miscarriage clinic, with a history of two or more pregnancy losses. 857 new patients attended over a 30-month period and were invited to participate. Participant data were recorded on a bespoke study database, 'Tommy's Net'.

PARTICIPANTS

777 women consented to participate (90.7% of new patients). 639 (82%) women continued within the cohort, and 138 were lost to follow-up. Mean age of active participants was 34 years for women and 37 years for partners, with a mean of 3.5 (1-19) previous pregnancy losses. Rates of obesity (maternal: 23.8%, paternal: 22.4%), smoking (maternal:7.4%, paternal: 19.4%) and alcohol consumption (maternal: 50%, paternal: 79.2%) were high and 55% of participants were not taking folic acid.

OUTCOME MEASURES

Biannual collection of pregnancy outcomes, either through prompted self-reporting, or existing hospital systems.

RESULTS

639 (82%) women were followed up. 404 (83.4%) reported conception and 106 (16.6%) reported no pregnancy, at least 6 months following registration. Of those that conceived, 72.8% (294/404) had a viable pregnancy. Maternal smoking and body mass index (BMI) over 30 were significantly higher in those who did not conceive (p=0.001) CONCLUSIONS: Tommy's Net provides a secure electronic repository on data for couples with recurrent pregnancy loss and associated outcomes. The study identified that subfertility, as well as repeated miscarriage, maternal BMI and smoking status, contributed to failure to achieve live birth. Study findings may enable comparison of clinic outcomes and inform the development of a personalised holistic care package.

Added value of magnetic resonance spectroscopy for diagnosing childhood cerebellar tumours

¹ H-magnetic resonance spectroscopy (MRS) provides noninvasive metabolite profiles with the potential to aid the diagnosis of brain tumours. Prospective studies of diagnostic accuracy and comparisons with conventional MRI are lacking. The aim of the current study was to evaluate, prospectively, the diagnostic accuracy of a previously established classifier for diagnosing the three major childhood cerebellar tumours, and to determine added value compared with standard reporting of conventional imaging. Single-voxel MRS (1.5 T, PRESS, TE 30 ms, TR 1500 ms, spectral resolution 1 Hz/point) was acquired prospectively on 39 consecutive cerebellar tumours with histopathological diagnoses of pilocytic astrocytoma, ependymoma or medulloblastoma. Spectra were analysed with LCModel and predefined quality control criteria were applied, leaving 33 cases in the analysis. The MRS diagnostic classifier was applied to this dataset. A retrospective analysis was subsequently undertaken by three radiologists, blind to histopathological diagnosis, to determine the change in diagnostic certainty when sequentially viewing conventional imaging, MRS and a decision support tool, based on the classifier. The overall classifier accuracy, evaluated prospectively, was 91%. Incorrectly classified cases, two anaplastic ependymomas, and a rare histological variant of medulloblastoma, were not well represented in the original training set. On retrospective review of conventional MRI, MRS and the classifier result, all radiologists showed a significant increase (Wilcoxon signed rank test, p < 0.001) in their certainty of the correct diagnosis, between viewing the conventional imaging and MRS with the decision support system. It was concluded that MRS can aid the noninvasive diagnosis of posterior fossa tumours in children, and that a decision support classifier helps in MRS interpretation.

Hazards for the Implementation and Use of Artificial Intelligence Enabled Digital Health Interventions, a UK Perspective

BACKGROUND

Artificial Intelligence (AI) has seen an increased application within digital healthcare interventions (DHIs). DHIs use entails challenges about their safety assurance. Exacerbated by regulatory requirements, in the UK, this places the onus of safety assurance not only on the manufacturer, but also on the operator of a DHI. Clinical Safety claims and evidencing safe implementation and use of AI-based DHIs require expertise, to understand and act to control or mitigate risk. Current health software standards, regulation, and guidance do not provide the insight necessary for safer implementation.

OBJECTIVE

To interpret published guidance and policy related to AI and justify clinical safety assurance of DHIs.

METHOD

Assessment of UK health regulation policy, standards, and AI institution insights, utilizing a published Hazard Assessment framework, to structure safety justifications, and articulate hazards relating to AI-based DHIs.

RESULTS

AI enabled DHI hazard identification, relating to implementation and use within healthcare delivery organizations.

CONCLUSION

By application of the method, we postulate that UK research of AI DHIs highlighted issues that may affect safety, in need of consideration to justify safety of a DHI.

Effect of mobile health interventions in increasing utilization of Maternal and Child Health care services in developing countries: A scoping review

Background

Mobile health (mHealth) technology is being used predominantly in low- and middle-income countries. Developing countries with low level of investment in health infrastructure can augment existing capacity by adopting low-cost affordable technology. The aim of the review was to summarize the available evidence on mHealth interventions that aimed at increasing the utilization of Maternal and Child Health (MCH) care services. Further, this review investigated the barriers which prevent the use of mHealth among both health care workers as well as beneficiaries.

Methodology

A scoping review of literature was undertaken using the five-stage framework developed by Arksey and O’Malley. The articles published between 1990 and 2021 were retrieved from three databases (PubMed, Cochrane Reviews, and Google Scholar) and grey literature for this review. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist was followed to present the findings.

Result

A total of 573 studies were identified. After removing duplicates, studies not related to mHealth and MCH and publications of systematic reviews and protocols for studies, a total of 28 studies were selected for review. The study design of the research articles which appeared during the search process were mostly observational, cross-sectional, and randomized controlled trials (RCTs). We have classified the studies into four categories based on the outcomes for which the mHealth intervention was implemented: MCH care services, child immunization, nutrition services, and perceptions of stakeholders toward using technology for improving MCH outcomes.

Conclusion

This brief review concludes that mHealth interventions can improve access to MCH services. However, further studies based on large sample size and strong research design are recommended.

Magnetic resonance image-based brain tumour segmentation methods: A systematic review

Background

Image segmentation is an essential step in the analysis and subsequent characterisation of brain tumours through magnetic resonance imaging. In the literature, segmentation methods are empowered by open-access magnetic resonance imaging datasets, such as the brain tumour segmentation dataset. Moreover, with the increased use of artificial intelligence methods in medical imaging, access to larger data repositories has become vital in method development.

Purpose

To determine what automated brain tumour segmentation techniques can medical imaging specialists and clinicians use to identify tumour components, compared to manual segmentation.

Methods

We conducted a systematic review of 572 brain tumour segmentation studies during 2015-2020. We reviewed segmentation techniques using T1-weighted, T2-weighted, gadolinium-enhanced T1-weighted, fluid-attenuated inversion recovery, diffusion-weighted and perfusion-weighted magnetic resonance imaging sequences. Moreover, we assessed physics or mathematics-based methods, deep learning methods, and software-based or semi-automatic methods, as applied to magnetic resonance imaging techniques. Particularly, we synthesised each method as per the utilised magnetic resonance imaging sequences, study population, technical approach (such as deep learning) and performance score measures (such as Dice score).

Statistical tests

We compared median Dice score in segmenting the whole tumour, tumour core and enhanced tumour.

Results

We found that T1-weighted, gadolinium-enhanced T1-weighted, T2-weighted and fluid-attenuated inversion recovery magnetic resonance imaging are used the most in various segmentation algorithms. However, there is limited use of perfusion-weighted and diffusion-weighted magnetic resonance imaging. Moreover, we found that the U-Net deep learning technology is cited the most, and has high accuracy (Dice score 0.9) for magnetic resonance imaging-based brain tumour segmentation.

Conclusion

U-Net is a promising deep learning technology for magnetic resonance imaging-based brain tumour segmentation. The community should be encouraged to contribute open-access datasets so training, testing and validation of deep learning algorithms can be improved, particularly for diffusion- and perfusion-weighted magnetic resonance imaging, where there are limited datasets available.

Digitally enabled flash glucose monitoring for inpatients with COVID-19: Feasibility and pilot implementation in a teaching NHS Hospital in the UK

BACKGROUND

COVID-19 placed significant challenges on healthcare systems. People with diabetes are at high risk of severe COVID-19 with poor outcomes. We describe the first reported use of inpatient digital flash glucose monitoring devices in a UK NHS hospital to support management of people with diabetes hospitalized for COVID-19.

METHODS

Inpatients at University Hospitals Coventry & Warwickshire (UHCW) NHS Trust with COVID-19 and diabetes were considered for digitally enabled flash glucose monitoring during their hospitalization. Glucose monitoring data were analysed, and potential associations were explored between relevant parameters, including time in hypoglycaemia, hyperglycaemia, and in range, glycated haemoglobin (HbA1c), average glucose, body mass index (BMI), and length of stay.

RESULTS

During this pilot, digital flash glucose monitoring devices were offered to 25 inpatients, of whom 20 (type 2/type 1: 19/1; mean age: 70.6 years; mean HbA1c: 68.2 mmol/mol; mean BMI: 28.2 kg/m2) accepted and used these (80% uptake). In total, over 2788 h of flash glucose monitoring were recorded for these inpatients with COVID-19 and diabetes. Length of stay was not associated with any of the studied variables (all p-values >0.05). Percentage of time in hyperglycaemia exhibited significant associations with both percentage of time in hypoglycaemia and percentage of time in range, as well as with HbA1c (all p-values <0.05). The average glucose was significantly associated with percentage of time in hypoglycaemia, percentage of time in range, and HbA1c (all p-values <0.05).

DISCUSSION

We report the first pilot inpatient use of digital flash glucose monitors in an NHS hospital to support care of inpatients with diabetes and COVID-19. Overall, there are strong arguments for the inpatient use of these devices in the COVID-19 setting, and the findings of this pilot demonstrate feasibility of this digitally enabled approach and support wider use for inpatients with diabetes and COVID-19.

Combining multi-site magnetic resonance imaging with machine learning predicts survival in pediatric brain tumors

Brain tumors represent the highest cause of mortality in the pediatric oncological population. Diagnosis is commonly performed with magnetic resonance imaging. Survival biomarkers are challenging to identify due to the relatively low numbers of individual tumor types. 69 children with biopsy-confirmed brain tumors were recruited into this study. All participants had perfusion and diffusion weighted imaging performed at diagnosis. Imaging data were processed using conventional methods, and a Bayesian survival analysis performed. Unsupervised and supervised machine learning were performed with the survival features, to determine novel sub-groups related to survival. Sub-group analysis was undertaken to understand differences in imaging features. Survival analysis showed that a combination of diffusion and perfusion imaging were able to determine two novel sub-groups of brain tumors with different survival characteristics (p < 0.01), which were subsequently classified with high accuracy (98%) by a neural network. Analysis of high-grade tumors showed a marked difference in survival (p = 0.029) between the two clusters with high risk and low risk imaging features. This study has developed a novel model of survival for pediatric brain tumors. Tumor perfusion plays a key role in determining survival and should be considered as a high priority for future imaging protocols.

Association between glycosylated haemoglobin and outcomes for patients discharged from hospital with diabetes: A health informatics approach

Aims/Objectives

Extensive research considers associations between inpatient glycaemic control and outcomes during hospital admission; this cautions against overly tight glycaemic targets. Little research considers glycaemic control following hospital discharge. This is despite a clear understanding that people with diabetes are at increased risk of negative outcomes, following discharge. We evaluate absolute and relative Hba1c values, and frequency of Hba1c monitoring, on readmission and mortality rates for people discharged from hospital with diabetes.

Methods

All discharges (n = 46,357) with diabetes from a major tertiary referral centre over 3 years were extracted, including biochemistry data. We conducted an evaluation of association between Hba1c, mortality and readmission, statistical significance and standardised Cohen's D effect size calculations.

Results

399 patients had a Hba1c performed during their admission. 3,138 patients had a Hba1c within 1 year of discharge. Mean average Hba1c for readmissions was 57.82 vs 60.39 for not readmitted (p = 0.009, Cohen's D 0.28). Mean average number of days to Hba1c testing in readmitted was 97 vs 113 for those not readmitted (p = 0.00006, Cohen's D 0.39). Further evaluation of mortality outcomes, cohorts of T1DM and T2DM and association of relative change in Hba1c was performed.

Conclusions

Lower Hba1c values following discharge from hospital are significantly associated with increased risk of readmission, as is a shorter duration until testing. Similar patterns observed for mortality. Findings particularly prominent for T1DM. Further research needed to consider underlying causation and design of appropriate risk stratification models.

Identifying Myocardial Infarction Using Hierarchical Template Matching-Based Myocardial Strain: Algorithm Development and Usability Study

Background

Myocardial infarction (MI; location and extent of infarction) can be determined by late enhancement cardiac magnetic resonance (CMR) imaging, which requires the injection of a potentially harmful gadolinium-based contrast agent (GBCA). Alternatively, emerging research in the area of myocardial strain has shown potential to identify MI using strain values.

Objective

This study aims to identify the location of MI by developing an applied algorithmic method of circumferential strain (CS) values, which are derived through a novel hierarchical template matching (HTM) method.

Methods

HTM-based CS H-spread from end-diastole to end-systole was used to develop an applied method. Grid-tagging magnetic resonance imaging was used to calculate strain values in the left ventricular (LV) myocardium, followed by the 16-segment American Heart Association model. The data set was used with k-fold cross-validation to estimate the percentage reduction of H-spread among infarcted and noninfarcted LV segments. A total of 43 participants (38 MI and 5 healthy) who underwent CMR imaging were retrospectively selected. Infarcted segments detected by using this method were validated by comparison with late enhancement CMR, and the diagnostic performance of the applied algorithmic method was evaluated with a receiver operating characteristic curve test.

Results

The H-spread of the CS was reduced in infarcted segments compared with noninfarcted segments of the LV. The reductions were 30% in basal segments, 30% in midventricular segments, and 20% in apical LV segments. The diagnostic accuracy of detection, using the reported method, was represented by area under the curve values, which were 0.85, 0.82, and 0.87 for basal, midventricular, and apical slices, respectively, demonstrating good agreement with the late-gadolinium enhancement–based detections.

Conclusions

The proposed applied algorithmic method has the potential to accurately identify the location of infarcted LV segments without the administration of late-gadolinium enhancement. Such an approach adds the potential to safely identify MI, potentially reduce patient scanning time, and extend the utility of CMR in patients who are contraindicated for the use of GBCA.

Mobile consulting as an option for delivering healthcare services in low-resource settings in low- and middle-income countries: A mixed-methods study

OBJECTIVE

Remote or mobile consulting is being promoted to strengthen health systems, deliver universal health coverage and facilitate safe clinical communication during coronavirus disease 2019 and beyond. We explored whether mobile consulting is a viable option for communities with minimal resources in low- and middle-income countries.

METHODS

We reviewed evidence published since 2018 about mobile consulting in low- and middle-income countries and undertook a scoping study (pre-coronavirus disease) in two rural settings (Pakistan and Tanzania) and five urban slums (Kenya, Nigeria and Bangladesh), using policy/document review, secondary analysis of survey data (from the urban sites) and thematic analysis of interviews/workshops with community members, healthcare workers, digital/telecommunications experts, mobile consulting providers, and local and national decision-makers. Project advisory groups guided the study in each country.

RESULTS

We reviewed four empirical studies and seven reviews, analysed data from 5322 urban slum households and engaged with 424 stakeholders in rural and urban sites. Regulatory frameworks are available in each country. Mobile consulting services are operating through provider platforms (n = 5-17) and, at the community level, some direct experience of mobile consulting with healthcare workers using their own phones was reported - for emergencies, advice and care follow-up. Stakeholder willingness was high, provided challenges are addressed in technology, infrastructure, data security, confidentiality, acceptability and health system integration. Mobile consulting can reduce affordability barriers and facilitate care-seeking practices.

CONCLUSIONS

There are indications of readiness for mobile consulting in communities with minimal resources. However, wider system strengthening is needed to bolster referrals, specialist services, laboratories and supply chains to fully realise the continuity of care and responsiveness that mobile consulting services offer, particularly during/beyond coronavirus disease 2019.

IMG-06. PREDICTING SURVIVAL FROM PERFUSION AND DIFFUSION MRI BY MACHINE LEARNING

INTRODUCTION

Magnetic Resonance Imaging (MRI) is routinely used in the assessment of children’s brain tumours. Reduced diffusion and increased perfusion on MRI are commonly associated with higher grade but there is a lack of quantitative data linking these parameters to survival. Machine learning is increasingly being used to develop diagnostic tools but its use in survival analysis is rare. In this study we combine quantitative parameters from diffusion and perfusion MRI with machine learning to develop a model of survival for paediatric brain tumours. METHOD: 69 children from 4 centres (Birmingham, Liverpool, Nottingham, Newcastle) underwent MRI with diffusion and perfusion (dynamic susceptibility contrast) at diagnosis. Images were processed to form ADC, cerebral blood volume (CBV) and vessel leakage correction (K2) parameter maps. Parameter mean, standard deviation and heterogeneity measures (skewness and kurtosis) were calculated from tumour and whole brain and used in iterative Bayesian survival analysis. The features selected were used for k-means clustering and differences in survival between clusters assessed by Kaplan-Meier and Cox-regression.

RESULTS

Bayesian analysis revealed the 5 top features determining survival to be tumour volume, ADC kurtosis, CBV mean, K2 mean and whole brain CBV mean. K-means clustering using these features showed two distinct clusters (high- and low-risk) which bore significantly different survival characteristics (Hazard Ratio = 5.6).

DISCUSSION AND CONCLUSION

Diffusion and perfusion MRI can be used to aid the prediction of survival in children’s brain tumours. Tumour perfusion played a particularly important role in predicting survival despite being less routinely measured than diffusion.

A systematic review considering risk factors for mortality of patients discharged from hospital with a diagnosis of diabetes

AIM

To identify known risk factors for mortality for adult patients, discharged from hospital with diabetes.

METHOD

The systematic review was based on the PRISMA protocol. Studies were identified through EMBASE & MEDLINE databases. The inclusion criteria were papers that were published over the last 6 years, in English language, and focused on risk factors of mortality in adult patients with diabetes, after they were discharged from hospitals. This was followed by data extraction "with quality assessment and semi-quantitative synthesis according to PRISMA guidelines".

RESULTS

There were 35 studies identified, considering risk factors relating to mortality for patients, discharged from hospital with diabetes. These studies are distributed internationally. 48 distinct statistically significant risk factors for mortality can be identified. Risk factors can be grouped into the following categories; demographic, socioeconomic, lifestyle, patient medical, inpatient stay, medication related, laboratory results, and gylcaemic status. These risk factors can be further divided into risk factors identified in generalized populations of patients with diabetes, compared to specific sub-populations of people with diabetes.

CONCLUSION

A relatively small number of studies have considered risk factors relating to mortality for patients, discharged from hospital with a diagnosis of diabetes. Mortality is an important outcome, when considering discharge from hospital with diabetes. However, there has only been limited consideration within the research literature.

Application of standardised effect sizes to hospital discharge outcomes for people with diabetes

BACKGROUND

Patients with diabetes are at an increased risk of readmission and mortality when discharged from hospital. Existing research identifies statistically significant risk factors that are thought to underpin these outcomes. Increasingly, these risk factors are being used to create risk prediction models, and target risk modifying interventions. These risk factors are typically reported in the literature accompanied by unstandardized effect sizes, which makes comparisons difficult. We demonstrate an assessment of variation between standardised effect sizes for such risk factors across care outcomes and patient cohorts. Such an approach will support development of more rigorous risk stratification tools and better targeting of intervention measures.

METHODS

Data was extracted from the electronic health record of a major tertiary referral centre, over a 3-year period, for all patients discharged from hospital with a concurrent diagnosis of diabetes mellitus. Risk factors selected for extraction were pre-specified according to a systematic review of the research literature. Standardised effect sizes were calculated for all statistically significant risk factors, and compared across patient cohorts and both readmission & mortality outcome measures.

RESULTS

Data was extracted for 46,357 distinct admissions patients, creating a large dataset of approximately 10,281,400 data points. The calculation of standardized effect size measures allowed direct comparison. Effect sizes were noted to be larger for mortality compared to readmission, as well as for being larger for surgical and type 1 diabetes cohorts of patients.

CONCLUSIONS

The calculation of standardised effect sizes is an important step in evaluating risk factors for healthcare events. This will improve our understanding of risk and support the development of more effective risk stratification tools to support patients to make better informed decisions at discharge from hospital.

Generation of Realistic Synthetic Validation Healthcare Datasets Using Generative Adversarial Networks

BACKGROUND

Assurance of digital health interventions involves, amongst others, clinical validation, which requires large datasets to test the application in realistic clinical scenarios. Development of such datasets is time consuming and challenging in terms of maintaining patient anonymity and consent.

OBJECTIVE

The development of synthetic datasets that maintain the statistical properties of the real datasets.

METHOD

An artificial neural network based, generative adversarial network was implemented and trained, using numerical and categorical variables, including ICD-9 codes from the MIMIC III dataset, to produce a synthetic dataset.

RESULTS

The synthetic dataset, exhibits a correlation matrix highly similar to the real dataset, good Jaccard similarity and passing the KS test.

CONCLUSIONS

The proof of concept was successful with the approach being promising for further work.

A Method of Justifying Confidence in the Safety of Digital Health Interventions

Digital health interventions (DHIs) enable improvements in health strategy and address health system challenges. The World Health Organization provides a formal classification for DHIs. However, safety claims, about such interventions, vary in quality and are often vague as to how they are communicated between technical, clinical experts and stakeholders. By combining the classifications with a method of safety analysis and justification, we postulate confidence in the safety of digital technology. Confidence is resulting from the application of the framework to the DHI, using defined health system challenges. The framework and derived safety justifications can be applied to any DHI. It can serve as guideline for health strategy, regulatory and standards based compliance.

Safety Justification of Healthcare Applications Using Synthetic Datasets

BACKGROUND

Increasing numbers of intelligent healthcare applications are developed by analysing big data, on which they are trained. It is necessary to assure that such applications will be safe for patients; this entails validation against datasets. But datasets cannot be shared easily, due to privacy, and consent issues, resulting in delaying innovation. Realistic Synthetic Datasets (RSDs), equivalent to the real datasets, are seen as a solution to this.

OBJECTIVE

To develop the outline for safety justification of an application, validated with an RSD, and identify the safety evidence the RSD developers will need to generate.

METHOD

Assurance case argument development approaches were used, including high level data related risk identification.

RESULT

An outline of the justification of such applications, focusing on the contribution of the RSD.

CONCLUSIONS

Use of RSD will require specific arguments and evidence, which will affect the adopted methods. Mutually supporting arguments can result in a compelling justification.

Mortality statistics in England and Wales: the SARS-CoV-2 paradox

Objective

To analyse mortality statistics in the United Kingdom during the initial phases of the severe acute respiratory coronavirus 2 (SARS-CoV-2) pandemic and to understand the impact of the pandemic on national mortality.

Methods

Retrospective review of weekly national mortality statistics in the United Kingdom over the past 5 years, including subgroup analysis of respiratory mortality rates.

Results

During the early phases of the SARS-CoV-2 pandemic in the first months of 2020, there were consistently fewer deaths per week compared with the preceding 5 years. This pattern was not observed at any other time within the past 5 years. We have termed this phenomenon the “SARS-CoV-2 paradox.” We postulate potential explanations for this seeming paradox and explore the implications of these data.

Conclusions

Paradoxically, but potentially importantly, lower rather than higher weekly mortality rates were observed during the early stages of the SARS-CoV-2 pandemic. This paradox may have implications for current and future healthcare utilisation. A rebound increase in non-SARS-CoV-2 mortality later this year might coincide with the peak of SARS-CoV-2 admissions and mortality.

Localisation, Personalisation and Delivery of Best Practice Guidelines on an Integrated Care and Cure Cloud Architecture: The C3-Cloud Approach to Managing Multimorbidity

BACKGROUND

C3-Cloud is an integrated care ICT infrastructure offering seamless patient-centered approach to managing multimorbidity, deployed in three European pilot sites. Challenge: The digital delivery of best practice guidelines unified for multimorbidity, customized to local practice, offering the capability to improve patient personalization and benefit.

METHOD

C3-Cloud has adopted a co-production approach to developing unified multimorbidity guidelines, by collating and reconciling best practice guidelines for each condition. Clinical and technical teams at pilot sites and the C3-Cloud consortium worked in tandem to create the specification and technical implementation.

RESULTS

C3-Cloud offers CDSS for diabetes, renal failure, depression and congenital heart failure, with over 300 rules and checks that deliver four best practice guidelines in parallel, customized for each pilot site.

CONCLUSIONS

The process provided a traceable, maintainable and audited digitally delivered collated and reconciled guidelines.

Evaluation of patient perception towards dynamic health data sharing using blockchain based digital consent with the Dovetail digital consent application: A cross sectional exploratory study

Background

New patient-centric integrated care models are enabled by the capability to exchange the patient’s data amongst stakeholders, who each specialise in different aspects of the patient’s care. This requires a robust, trusted and flexible mechanism for patients to offer consent to share their data. Furthermore, new IT technologies make it easier to give patients more control over their data, including the right to revoke consent. These characteristics challenge the traditional paper-based, single-organisation-led consent process. The Dovetail digital consent application uses a mobile application and blockchain based infrastructure to offer this capability, as part of a pilot allowing patients to have their data shared amongst digital tools, empowering patients to manage their condition within an integrated care setting.

Objective

To evaluate patient perceptions towards existing consent processes, and the Dovetail blockchain based digital consent application as a means to manage data sharing in the context of diabetes care.

Method

Patients with diabetes at a General Practitioner practice were recruited. Data were collected using focus groups and questionnaires. Thematic analysis of the focus group transcripts and descriptive statistics of the questionnaires was performed.

Results

There was a lack of understanding of existing consent processes in place, and many patients did not have any recollection of having previously given consent. The digital consent application received favourable feedback, with patients recognising the value of the capability offered by the application. Patients overwhelmingly favoured the digital consent application over existing practice.

Conclusions

Digital consent was received favourably, with patients recognising that it addresses the main limitations of the current process. Feedback on potential improvements was received. Future work includes confirmation of results in a broader demographic sample and across multiple conditions.