Northumbria UniversityFaculty of Engineering and EnvironmentResearch ProposalAcademic Year: 2017-19Module: LD7028 – Research and Project Management Module tutor: Professor Hamid Jahankhani ??Title: “Implementation Of Artificial Intelligence in the NHS England” ?Student name: Noah Kyeyune – 16033996Programme: ?MSc Cyber SecuritySupervisor: Professor Hamid JahankhaniWord Count:Contents Aim.
……..
….
…….
…
….
……..
….
…
……..
…….
….
……
…….
…….
….
…
……..
……..
….
..3 Background…..
….
…
……..
…
……
…
……..
…
……
……
…
……..
……..
…
……..
……
…..
.3 Sources and Use of Knowledge …..
……
…
……
….
……..
……
……
……..
…….
……7 Scope.
……..
….
……
……
……..
…
……
…
…
…..
…..
……
…..
…
…
…….
……
……
…….
…..8Smart Objectives..
…..
……
……..
……
….
….
…….
…….
……..
….
…
…….
…..
…
….
…
…8Methodology…..
……..
….
….
…….
……..
….
…
……..
…….
……………………………….9Mind Map……………………………………………………………………….10Project Plan…………………………………………………………………………………..11Gantt Chart…………………………………………………………………………………..12Risk Log ………………………………………………………………………………………13Ethics Form…………………………………………………………………….. References…………………………………………………………………………………..15 AimThe main aim of this research is to discuss ways of how to implement Artificial Intelligence within the healthcare sector with emphasis on data privacy. Therefore, the research question is, “What’s the Impact Of Implementing Artificial Intelligence on data privacy in Healthcare/NHS England?BackgroundWhat is Artificial Intelligence?Artificial Intelligence (AI) refers to computer systems that think and act like humans, and think and act rationally. It is a combination of advanced general-purpose digital technologies that enable machines to do highly complex tasks effectively. AI is currently transforming modern life with new innovations like autonomous vehicles, smart home devices and automation in several other fields of life.Machine learning and artificial intelligence (AI) are being applied more broadly across industries and applications than ever before as computing power, data collection and storage capabilities increase. Human beings in real time cannot handle the vast trove of data held in various sectors of life. With machine learning and AI, the big beast of data could be dealt with in a very short space of time, which can help drive business and innovative progress to improve medical care, business operations and other spheres of life. AI could be game changer for modern and future generations. The application of AI in medicine has two main branches: virtual and physical. The virtual component is represented by Machine Learning, (also called Deep Learning) which is represented by mathematical algorithms that improve learning through experience. I have chosen to carry out this research to explore how to achieve the security triad of confidentiality, Integrity and availability of patient data while rolling out Artificial Intelligence. In order to deploy A.I securely in the healthcare sector such as UK’s NHS systems, a good framework with minimal sophistication must be engaged with this technology emphasizing security of patient data. Why implement Artificial Intelligence in Healthcare?Yves-Alexandre de Montjoye et al -2017, note that Artificial Intelligence (AI) has potential to fundamentally change the way we work, live, and interact. There are several opportunities of using A.I technologies such as identifying subtle signs of disease, clinical decision-making, introduction of complimentary medical technologies and automating elements of medical practice among others.The Current AI (IBM Watson) is capable of diagnosing life-threatening diseases and prescribes treatment. Currently we have robots being used in surgery, physiotherapy and many other medical fields.AI solutions are being developed to automate image analysis and diagnosis. This can help highlight areas of interest on a scan to a radiologist, to drive efficiency and reduce human error.AI solutions are being developed to identify new potential therapies from vast databases of information on existing medicines, which could be redesigned to target critical threats such as the Ebola virusBy analysing vast amounts of historic patient data, AI solutions can provide real-time support to clinicians to help identify at risk patients. A current focal point includes re-admission risks, and highlighting patients that have an increased chance of returning to hospital within 30 days of dischargeMultiple organisations are working on direct to patient solutions to triage and give advice via a voice or chat-based interaction. This provides quick, scalable access for basic questions and medical issues. This could help avoid unnecessary trips to the GP, reducing rising demand on primary healthcare providersHowever, to successfully launch Artificial Intelligence comes with several challenges and understanding the decisions AI algorithms make when using data. Consistent accuracy is important to preserve trust in the technology, but AI is still in its infancy. Whilst AI systems may have been trained on comprehensive datasets, in the clinical setting they may encounter data and scenarios that they have not been trained on, potentially making them less accurate and reliable and therefore risking patient safety.”We know that health data is personal and sensitive, so there are rightly strict rules in place about how and when it can be used or shared. We need to ensure that any new developments harness the power of data but that they do so responsibly and within the legal frameworks.”There are standards that shall be examined in order to assist in the deployment of A.I in the healthcare sector. These include the Information Governance Toolkit GDPR and Data Quality Metrics Index in the NHS UK.The Information Governance Toolkit adequately seeks to protect transfers / flows of information, organisations need to identify the transfers, risk assess the transfer methods and consider the sensitivity of the information being transferred. Transfers of all information (including personal information) must comply with professional standards and relevant legislation – (Data Protection Act 1998) which requires appropriate technical and organisational measures to be taken against unauthorised or unlawful processing of, and accidental loss or destruction of, or damage to, personal data. The Data Quality Maturity Index (DQMI) reports on the quality of provider data across seven national datasets. The use of both the IG toolkit and the DQMI is a way forward in designing an accepted framework for implementing A.I in the NHS while protecting patient/data privacy.In the case of GDPR, administrative data research, which is carried out on data routinely collected and held by public authorities, would be particularly affected as the sharing of de-identified data could constitute the unconsented disclosure of identifiable information. In addition, the development and deployment of modern privacy- enhancing technologies (PET), allowing data controllers to make data available in a safe and transparent way, will be key to unlocking the great potential of AI. Adrian Smalla et al (2017) in their research focused on socio-political and socio – cultural issues in the design and specification of the Role Access Control system, a precursor to the electronic health system in the NHS. But they noted the existence of several complicated overlapping systems and constraints in the policies of the NHS in the UK. They proposed a combination of process modeling and technology management. This approach could also be suitable with some tweaking in the implementation of Artificial intelligence in the NHS. It is important to note that it falls short of acknowledging the numerous NHS independent systems working in isolation of each other and the creation of new layers of security at each level.Professor Dame Wendy Hall et al (2017) in their review report, “Growing the artificial intelligence Industry in the UK”, point out that with A.I computers can learn and analyse information at a higher accuracy and speed than human beings.The main obstacle in all these recommendations and innovative systems is addressing data security of patient records. Many organisations such as DeepMind (Google) are pioneering and have launched many A.I applications for use in the healthcare sector, but their main obstacles are safety and privacy issues.A case study in point is an example where NHS patient data was provided without informed consent to Google to test an app designed to help monitor kidney disease. There was an argument that this violated UK regulations because patient records cannot be used without their explicit consent (according to the current IG Toolkit). However, this controversy has not slowed DeepMind’s continued and expanding partnership with NHS. The literature available in the field of Artificial Intelligence, which is still developing, shows that there are still lots of issues to solve if it is to be successfully implemented in the healthcare sector like the NHS UK. The successful future of AI requires us a new approach to data protection and examine what role information governance can play.It is important to note that the healthcare sector is a multidisciplinary focused service and has a number of stakeholders involved. Therefore, identifying a suitable framework for implementing A.I and a technological design that meets multiple government and industry standards, and implementing such a technology, is a challenging problem. My research shall focus on how the Information Governance toolkit in combination with the Data Quality Maturity Index (DQMI) and GDPR can be used as benchmarks for the protection of patient privacy while implementing Artificial Intelligence in the NHS UK.Having worked in the healthcare sector (NHS), I am interested in the introduction for A.I technologies to enhance patient safety and data privacy. This piece of work shall be a contribution to the development of A.I because of the focus of privacy issues and a rethinking of the current Information Governance Toolkit and Data Quality Maturity Index.Sources and use of Knowledge I look forward to have my work published in the International Journal of Medical Informatics Journal. This journal provides an international medium for dissemination of original results and interpretative reviews concerning the field of medical informatics. It also emphasizes the evaluation of systems in healthcare settings.The scope of this journal covers a wide variety of things including:Information systems, including national or international registration systems, hospital information systems, departmental and/or physician’s office systems, document handling systems, electronic medical record systems, standardization, systems integration etc.; Organizational, economic, social, clinical impact, ethical and cost-benefit aspects of IT applications in health care.Computer-aided medical decision support systems using heuristic, algorithmic and/or statistical methods as exemplified in decision theory, protocol development, artificial intelligence, etc. Educational computer based programs pertaining to medical informatics or medicine in general.This journal has a global authorship and readership with a very good rating. The following is its outstanding rating: – Journal Metrics Cite Score: 3.70 ?Impact Factor: 3.210 ?5 Year Impact Factor: 3.287 ?Source Normalized Impact per Paper (SNIP): 2.027 ?SCImago Journal Rank (SJR): 1.215 ScopeResearch shall dwell on information governance and its impact on implementing A.IData shall be obtained from open/public data setsThe NHS Toolkit shall be an important guide in all research undertakenExamine the role of NHS strategic Leadership in implementing A.ISmart ObjectivesSpecific Investigate and document primary and secondary data sources on implementation of A.I in the NHS UK (IG Tool Kit, Data Metric Quality Index, GDPR)Examine Impact of implementing Artificial Intelligence in NHS UK on Patient Confidentiality (Role Of Information Governance Toolkit, UK Anonymisation Network, National Data Guardian)Develop conceptual framework for security of patient data while implementing A.IAssess the Interdependence of business process systems and how they affect data security within the NHSMeasurable Gantt ChartAchievable Literature Review ResearchData Collection (Data sets are Publically available)Identify key variables (IG Toolkit, DMQI, DSP Toolkit, National Data Guardian)Relevant Artificial Intelligence is driving innovation in HealthcareTime – Bound 12 Weeks (Completion Date – January 2019)MethodologyThe work by Stuart Russell and Peter Norvig -2009 discusses the idea of an ‘Intelligent Agent’. They define A.I as the study of agents that receive percepts from the environment and perform actions.It is further asserted that each agent in the A.I realm implements a function that maps percept sequences to actions.The research shall be carried out using NHS data sets, which are publically available and the Data Coordination Board (DCB) provides assurance for all national data sets. There is also the Data Quality Maturity Index and the NHS Governance Toolkit. Primary and secondary sets of data shall be used to include policy papers, GDPR laws, journals and conference papers about Artificial Intelligence in the UK. I shall use a mixed method approach (both quantitative and qualitative) for my methodology.The Data Quality Maturity Index (DQMI) is a quarterly publication by NHS Digital (2018) and is intended to highlight the importance of data quality in the NHS. This framework is very instrumental in understanding the importance of data quality and integrity. The following are the datasets, which shall be used to examine the impact of implementing A.I in the NHS UK: -Accident and EmergencyAdmitted Patient CareChildren and Young People’s Health Services Diagnostic Imaging?Improving Access to Psychological Therapies Mental Health Services?Maternity Services?Outpatient The methodology is made of three analytical segments namely the Data Quality Maturity Index – a quantitative measure showing the importance of data quality be drawn from NHS datasets, GDPR and the NHS Governance Toolkit a qualitative measure for the protection of flow of information.All data used shall be completely anonymous and completely free of any personal information in accordance with the Data Protection Act (1998), UK Anonymous Network and ethical research practices.MIND MAP581464645275504431030109728000148844017145100505264626802860073855426216705744311613486511126225507960040327382680286-4921251459865 LegalNISTISO StandardsData Protection Act -1998GDPR (2018)Cyber EssentialsNHS Confidentiality Code of PracticeOther Regulatory Frameworks00 LegalNISTISO StandardsData Protection Act -1998GDPR (2018)Cyber EssentialsNHS Confidentiality Code of PracticeOther Regulatory Frameworks31407103710305 TechnologyDeepMind (VDA)BlockchainIoTIBM WatsonIntelligent AppsDigital TwinsImmersive Experience (AR;VR)0 TechnologyDeepMind (VDA)BlockchainIoTIBM WatsonIntelligent AppsDigital TwinsImmersive Experience (AR;VR)62826902409825 Data SecurityHackingHolomorphic EncryptionCyber Hygiene Data SecurityHackingHolomorphic EncryptionCyber Hygiene3761740286385 Ethical IssuesGDPRDSP Toolkit Ethical IssuesGDPRDSP Toolkit7348855182245ConfidentialityIntegrityAvailabilityAuthenticity0ConfidentialityIntegrityAvailabilityAuthenticity-11747553340 SocialSocietal AttitudesGender StereotypesGenerational Differences SocialSocietal AttitudesGender StereotypesGenerational Differences31870652021205Impact of Implementing A.I on Data Privacy in Healthcare (NHS –UK)00Impact of Implementing A.I on Data Privacy in Healthcare (NHS –UK)-6559553687445 ProcessesCollection of NHS Data SetsNational Data Guardian GuidelinesUK Anonymous NetworkData Quality Maturity IndexQuality Assurance ProcessesCollection of NHS Data SetsNational Data Guardian GuidelinesUK Anonymous NetworkData Quality Maturity IndexQuality Assurance59904923758028PeopleMedical ProfessionalsStrategic LeadershipIndividualsBusinessesThird PartiesPeopleMedical ProfessionalsStrategic LeadershipIndividualsBusinessesThird PartiesProject PlanTask Duration Start End Previous Task ResourcesPersonal Reading 22 Days 05/2018 12/2018 Research Journals, Trending studiesResearch PapersBooksConference PapersLiterature Search/ Journal Search 20 Days 05/2018 06/2018 Reading Data Banks, Online librariesBooksCollection of Data 10 Days 06/2018 07/2018 Data Retrieval NHS Data setsPublic RecordsDraft document 10 Days 08/10/2018 18/10/2018 Document Review SSRN, Colwiz, Labii, ZoteroEvaluation of Findings 17 Days 04/11/2018 21/11/2018 Data Analysis NVivo, ATLAS.ti, CAQDASProof Reading 4 Days 12/12/2018 16/12/2018 Document Draft LaptopSubmission of Project 1 Day 19/01/2019 19/01/2019 Document Binding Stationery, PrinterGANNT CHARTKey: X- axis – Duration (Months) Y-axis – Activities Risk LogKeyRisk types – F (Financial), T (Technology), P (People), E (Environmental), S (Security) Risk Type Risk Event Likelihood (1-10) Impact (1-10) Risk Value(1-100) Risk Monitoring Risk Management StrategyRisk Review date Risk owner Commentary E Late Submission2 10 40 Keep strict deadlines Ensure project schedule deadlines are met Weekly Myself Always keep monitoring all activities to stay on scheduleE Computer breakdown1 10 30 Disk clean ; Computer maintenance Ensure computer disks are cleaned and anti-virus protection enabled Weekly Myself Regular disk clean up and install regular updatesE Unforeseen catastrophes3 10 40 None Contingency Plan Weekly Myself Always have a contingency planE Missing Targets4 10 20 Keeping Regular RemindersStrict Adherence to task list Weekly Myself Self motivation to achieve all targetsT Inability to secure data2 10 30 Secure assurances from all data sources Advance planning and obtain written permissions Weekly Myself Always back up data and secure data sources in time 25 ; 50 Risk may be acceptable – more analysis required; 25 Low risk – no gains expected from extra work 75 Risk very high – urgent action required 50 ; 75 Risk high – action as soon as possibleReferencesRussell, S. and Norvig, P. (2018), Artificial Intelligence: A Modern Approach, online Aima.cs.berkeley.edu. Available at: http://aima.cs.berkeley.edu Accessed 12 May, 2018Hamet, Pavel and Johanne Tremblay, “Artificial intelligence in medicine.” Metabolism: clinical and experimental 69S (2017): S36-S40.NHS Digital. (2018), NHS Digital comment on Reform report into Artificial Intelligence in health care – NHS Digital, online Available at: https://digital.nhs.uk/news-and-events/latest-news/nhs-digital-comment-on-reform-report-into-artificial-intelligence-in-health-care Accessed 2 Jun. 2018Mourby, M., Mackey, E., Elliot, M., Gowans, H., Wallace, S., ; Bell, (2018). Are ‘pseudonymised’ data always-personal data? Implications of the GDPR for administrative data research in the UK, Computer Law ; Security Review, 34(2), 222-233, Doi: 10.1016/j.clsr.2018.01.002Small, A, ; Wainwright, D, (2018), Privacy and security of electronic patient records – Tailoring multimethodology to explore the socio-political problems associated with Role Based Access Control systems, European Journal Of Operational Research, 265(1), 344-360. Doi: 10.1016/j.ejor.2017.07.041Hall, W. ; Pesenti, J. (2018), Growing the artificial intelligence industry in the UK. Retrieved from https://www.gov.uk/government/publications/growing-the-artificial-intelligence-industry-in-the-ukDerrington, D. (2017). Google DeepMind NHS data deal was ‘legally inappropriate’. Retrieved from https://www.newscientist.com/article/2131256-google-deepmind-nhs-data-deal-was-legally-inappropriate/Riazul Islam, S, Daehan Kwak, Humaun Kabir, M., Hossain, M., ; Kyung-Sup Kwak (2015), The Internet of Things for Health Care: A Comprehensive Survey. IEEE Access, 3, 678-708. Doi: 10.1109/access.2015.2437951Informatics, I. (2018), International Journal of Medical Informatics, retrieved from https://www.journals.elsevier.com/international-journal-of-medical-informaticsSnell, E. (2018), Why Blockchain Technology Matters for Healthcare Security, retrieved from https://healthitsecurity.com/features/why-blockchain-technology-matters-for-healthcare-security Kobie, N. (2018), AI has no place in the NHS if patient privacy isn’t assured, Online Wired.co.uk. Available at: http://www.wired.co.uk/article/ai-healthcare-gp-deepmind-privacy-problems Accessed 6 May. 2018Margot, J. (2018), Margot James’ statement on AI Sector Deal, Online GOV.UK, and Available at: https://www.gov.uk/government/news/margot-james-statement-on-ai-sector-deal Accessed 1st May. 2018Cyran, M, (2018), Blockchain as a Foundation for Sharing Healthcare Data. Blockchain in Healthcare Today.Balthazar, P., Harri, P., Prater, A. and Safdar, N. (2018). Protecting Your Patients’ Interests in the Era of Big Data, Artificial Intelligence and Predictive Analytics, Journal of the American College of Radiology, 15(3), pp.580-586.Dreyer, K. and Allen, B. (2018). Artificial Intelligence in Health Care: Brave New World or Golden Opportunity? – Journal of the American College of Radiology, 15(4), pp.655-657.Jaulent, M., Leprovost, D., Charlet, J, and Choquet, R. (2018), Semantic interoperability challenges to process large amount of data perspectives in forensic and legal medicine – Journal of Forensic and Legal Medicine, 57, pp.19-23.Butterworth, M. (2018). The ICO and artificial intelligence: The role of fairness in the GDPR framework, Computer Law & Security Review, 34(2), pp.257-268.?erka, P., Grigien?, J. and Sirbikyt?, G. (2017), Is it possible to grant legal personality to artificial intelligence software systems? Computer Law & Security Review, 33(5), pp.685-699.
