Med-Tech Innovation Conference 25-26th April 2018

Report by Stephen Ayre, Library Services Manager, George Eliot Hospital,

I recently attended the Med-Tech Innovation Conference at the Ricoh Arena in Coventry. As the main focus of this event was medical devices, it was out of my previous experience. What particularly interested me was the horizon scanning.

Karen Taylor spoke on the November 2017 Deloitte report The future awakens: Life Sciences and Health Care Predictions 2022. Deloitte’s six predictions for 2022 are:

1             Quantified self

Individuals will be better informed about their conditions. They want best treatments and to stay healthy. They are becoming true consumers. This trend is facilitated by the Internet of medical things. An example is OurPath for diabetes management.

  1. The culture in health care is transformed by digital technologies

Smart technology will make hospitals more efficient to better monitor patients and reduce length of stay, leading to smaller hospitals and fewer beds. First government funded digital hospital opened in 2017.

Patients know best: patient. Controlled health records systems are being developed.

Drones will deliver medicines; robots will diagnose.

  1. The life sciences industry is industrialised

Pharma are using machine learning to develop drugs, leading to faster development and cheaper drugs. Real-time inventory management with hospitals will become possible. Digital contract management will emerge.

  1. Data is the new health care currency

There has already been a vast increase in data. Artificial intelligence (AI) will be used to analyse it. Only 0.5 of data is ever analysed. DeepMind Health is being used to monitor vital signs for deterioration. Other applications:  genomics, & using data to control costs.

  1. The future of medicine is here and now

Preventive, predictive, personalised, participatory medicine has already started e.g. CAR T-cell therapy: precision medicine, Combination therapy is being looked at for dementia (No new dementia drug since 2003).

  1. New entrants are disrupting health care

Companies like Apple, Amazon are creating systems. Philips, what was once a lighting company, is emerging as a health company. UK model more resistant to new entrants, but in places like China new systems can be built from scratch, e.g. partnership with IBM for tackling cancer.

Another interesting presentation was from Max Jones of General Electric on the development of artificial intelligence. He presented a Gartner hype cycle on emerging technologies for 2017.

There is a shortage of radiologists in the UK, with a 17% vacancy rate. There is a 4% error rate and more pressure leads to more errors.  AI may provide a solution: it is predicted that AI will be able to read mammograms and chest X-rays by 2020. By 2027 CT of the head, chest and abdomen, MR of the head, knee and shoulder, and ultrasound of the liver, thyroid and carotids will be automated, and within 15 to 20 years nearly all imaging will be interpreted by learning algorithms.

Already Visual Vitals can tell your vital signs by videoing your face and comparing it with a databases. Another system (CliX Enrich can interpret notes from doctors and assign them to patients (discharge letters and handover notes are more informative than hard data). The BVI App can calculate body volume index from photos.

Humans plus machines is the way forward. Humans need to do best what AIs can’t do.

There was a note of caution from Mike Hannay of the Academic Health Sciences Network. Spreading of innovation is hard when there is resistance, e.g. smoking was proved to be harmful in 1954, but smoking rates still high decades later. The average uptake of health innovations is 3.5%. Many clinicians are focuses on immediate survival, Trusts on immediate costs. Compare this to spread of smartphones.

Debbie Chinn of NHS Digital spoke on “Clinical risk management for health informatics.” There is a need for education and training and guidance to underpin standards and regulations. On e-LfH there is Clinical Risk Management training. They have produced a Digital tools library.

Christian Subbe of Bangor University talked about increasing safety. Safety 1 looks at errors. Safety 2 looks at things that went right. Safety 3 looks to create redundancy in networks, e.g. giving data to patients who have a direct interest in safety, like pilot in plane.

Liz Ashall-Payne of Orcha spoke about the Mobile health landscape. Thera are more smartphones on the planet than toothbrushes. There are currently 326,000 health apps, with 5 million downloads per day. Of these, 28% cover mental health, 16% diabetes, 11% heart. The major blockers to wider adoption are awareness, accessibility and trust, e.g. Suicide prevention app promoting suicide. Prank apps; ratings of apps can be gamed. 50% of apps have fewer than 500 downloads. 2/3 of apps zombie apps, I.e. Not maintained. Orcha analyses and categorises apps. Orcha creates local targeted app library, e.g. Lancashire STP, Salford Royal FT. Professional recommendations are key to promoting use. Orcha match apps by conditions, technical preferences, features, demographics. Gaps in the market are childhood obesity and bullying, and the next thing needing to be tackled is interoperability.

This was a conference that is outside normal librarian networks, but was valuable for the insight it gave. It struck me that the rise of Artificial Intelligence is something that Librarians as a profession need to face. If an AI can interpret medical images and notes, then surely an AI will be able to do literature searches. A question to ponder: what is our specific human contribution as librarians?

2 thoughts on “Med-Tech Innovation Conference 25-26th April 2018

  1. Thanks for this.

    “Humans plus machines is the way forward. Humans need to do best what AIs can’t do.” chimes with something I read over the weekend looking at Kasparov versus chess computers. Centaur chess sees the human working with the AI and generally beating the AI alone

Leave a Reply

Your email address will not be published. Required fields are marked *