List of questions

How can machine learning help improve the quality of bespoke software solutions, for biomarker adoption in the pharmaceutical industry, whilst reducing time to market of the technical solution and associated drug trial(s)?

This question originates from a combination of interests from the Almac Diagnostics Informatics and Bioinformatics teams along with the Almac Information Services Systems Delivery team. Diagnostics goal is to try to enable biomarker adoption within the pharmaceutical industry. With the increase in volume, complexity and types of data generated for biomarker analyses, we are interested in how machine learning could be used to advance this. Other factors to consider are to try and collect large volumes (molecular and other meta data e.g. clinical history) and analyse that data together to try and advance biomarker uptake. As we cannot achieve this solely by eye, we need machine learning techniques to address this.

In an increasingly complex world of molecular diagnostics being used to inform therapeutic decision making, how can we leverage machine learning to ensure that physicians test the right patients at the right time?

Just this month we have seen reports of IBM Watson having a challenging time with its Watson Oncology product designed to assist in cancer treatment decisions. Is there an alternative way to leverage real world data to better understand testing and response to treatment?

1904 (GSK): How do the external and commercial research organisations collaborate with QUB and other Northern Irish academic institutions to access health data?

1925 (INI): How do we address the ethical and communication issues in the use of personal health data to ensure better health outcomes?

1924 (Straylight Consulting): How can Northern Ireland truly differentiate with AI?

1904 (GSK): Northern Ireland is one of only a few UK areas with fully integrated electronic care records (NIECR). In addition to this, the Honest Broker Service (HBS) provides health and social care data from Data Warehouses held within the Business Services Organisation (BSO) to the DHSSPS and HSC organisations. HBS/NIECR present interesting resources for external and commercial research organisations however uncertainty around getting access to healthcare data in NI creates a challenge to investment and collaboration in NI. This workshop requires the participation of several stakeholders across NI in order to provide new angles to the challenge and an open discussion on how all stakeholders can work together to develop potential solutions. We are inviting academia, industry, INI and the Trust to discuss possible ways to tackle the issue raised together.

Building on this issue, this workshop will also tackle additional challenges for NI;
1925 (INI): This challenge focuses on how health departments and governments around the world ensure that personal health data is used effectively and how that conversation takes place at a national level. How do we avoid the issues already encountered e.g. Google DeepMind and Royal Free London NHS Foundation Trust (https://link.springer.com/article/10.1007/s12553-017-0179-1)
and of course the misuse of data etc. ( Cambridge Analytica etc.)

1924 (Straylight Consulting): With countries around the world formulating AI strategies and in many cases allocating large budgets to advancing AI, what can NI do to establish itself as relevant and impactful.

1904 (GSK): How do the external and commercial research organisations collaborate with QUB and other Northern Irish academic institutions to access health data?

1925 (INI): How do we address the ethical and communication issues in the use of personal health data to ensure better health outcomes?

1924 (Straylight Consulting): How can Northern Ireland truly differentiate with AI?

1904 (GSK): Northern Ireland is one of only a few UK areas with fully integrated electronic care records (NIECR). In addition to this, the Honest Broker Service (HBS) provides health and social care data from Data Warehouses held within the Business Services Organisation (BSO) to the DHSSPS and HSC organisations. HBS/NIECR present interesting resources for external and commercial research organisations however uncertainty around getting access to healthcare data in NI creates a challenge to investment and collaboration in NI. This workshop requires the participation of several stakeholders across NI in order to provide new angles to the challenge and an open discussion on how all stakeholders can work together to develop potential solutions. We are inviting academia, industry, INI and the Trust to discuss possible ways to tackle the issue raised together.

Building on this issue, this workshop will also tackle additional challenges for NI;
1925 (INI): This challenge focuses on how health departments and governments around the world ensure that personal health data is used effectively and how that conversation takes place at a national level. How do we avoid the issues already encountered e.g. Google DeepMind and Royal Free London NHS Foundation Trust (https://link.springer.com/article/10.1007/s12553-017-0179-1)
and of course the misuse of data etc. ( Cambridge Analytica etc.)

1924 (Straylight Consulting): With countries around the world formulating AI strategies and in many cases allocating large budgets to advancing AI, what can NI do to establish itself as relevant and impactful.

What is the true potential for AI in combating loneliness and providing companionship and assistance for elderly people living alone

Loneliness in the elderly is an epidemic and is now seen as the biggest risk to health for that age group. How do reconnect those isolated at home with the world to improve their quality of life and reduce the cost of their care.

How can we use AI and bioinformatics to move to a model of pre-emptive rather than reactive care - particularly for the elderly

recent studies such as "Janet's Story" (attached) show the dramatic savings possible with optimal frail care pathways versus the reactive model. How can we use tech to bring this about?

How can machine learning impact diagnostic applications?

We are already aware of some applications for AI in the diagnostic field but there is enormous potential for further developments. We currently collect a variety of data from our trials that encompasses not only biomarker data but also clinical and sociodemographic data. How can AI use this data to assist diagnosis? How can AI expedite adoption of new technology by the healthcare service?

How do we update/develop our software solutions to meet regulatory requirements around patient consent?

Implementing big data analytics in healthcare:
What are the current guidelines regarding patient consent and do we see regulations changing dramatically in the near future?

1904 (GSK): How do the external and commercial research organisations collaborate with QUB and other Northern Irish academic institutions to access health data?

1925 (INI): How do we address the ethical and communication issues in the use of personal health data to ensure better health outcomes?

1924 (Straylight Consulting): How can Northern Ireland truly differentiate with AI?

1904 (GSK): Northern Ireland is one of only a few UK areas with fully integrated electronic care records (NIECR). In addition to this, the Honest Broker Service (HBS) provides health and social care data from Data Warehouses held within the Business Services Organisation (BSO) to the DHSSPS and HSC organisations. HBS/NIECR present interesting resources for external and commercial research organisations however uncertainty around getting access to healthcare data in NI creates a challenge to investment and collaboration in NI. This workshop requires the participation of several stakeholders across NI in order to provide new angles to the challenge and an open discussion on how all stakeholders can work together to develop potential solutions. We are inviting academia, industry, INI and the Trust to discuss possible ways to tackle the issue raised together.

Building on this issue, this workshop will also tackle additional challenges for NI;
1925 (INI): This challenge focuses on how health departments and governments around the world ensure that personal health data is used effectively and how that conversation takes place at a national level. How do we avoid the issues already encountered e.g. Google DeepMind and Royal Free London NHS Foundation Trust (https://link.springer.com/article/10.1007/s12553-017-0179-1)
and of course the misuse of data etc. ( Cambridge Analytica etc.)

1924 (Straylight Consulting): With countries around the world formulating AI strategies and in many cases allocating large budgets to advancing AI, what can NI do to establish itself as relevant and impactful.

What opportunities exist for hardware acceleration in Omics, what are the drivers, specifically which data-processing operations need acceleration and what will be required for deployment of cloud based acceleration applications?

We are interested in identifying processing loads for acceleration that have widest application which will provide the best business opportunities.

We are interested in identifying applications that cannot be delivered without acceleration technologies

Our interest is algorithm acceleration. We are specifically interested in identifying the technologies that would be required to deliver this solution and identify the opportunities for acceleration eg encryption, database acceleration . machine vision and learning