Augmented Intelligence
Identifying and utilising tools that can improve our services.
How can AI add value to Medica?
When we talk about AI in Medica we are talking about Augmented Intelligence, rather than Artificial Intelligence. While Artificial Intelligence is the creation of machines to work and react like humans, Augmented Intelligence is using those same machines to complement/enhance the human worker. Our current focus is to identify tools which can improve the service for our patients and the reporting experience for reporters. The key areas where AI can enhance our services are:
Reduced turnaround times
Improved clinical accuracy
Reduce cost
Assurance read
Improved data quality
Prioritisation based on clinical urgency
What is Medica doing with AI?
We have been keeping pace with developments in Radiology AI for several years and at the end of 2019 we identified a number of providers who had ‘real world’ data to evidence the impact their product could make. In March 2020, we announced a strategic partnership with Qure.ai, to work together on the deployment of AI as a tool for service improvement and to support the reporting process.
The partnership covers two distinct algorithms:
qER: Commercially available CT Head product which analyses non contrast studies and flags up possible abnormalities and gives a worklist prioritisation output allowing us to flag that this it is urgent in other products.
qFLAG: A new product we will work on and develop together to allow us to identify the body parts and study types for each exam that is analysed.
Progress to date
We have developed the necessary clinical, technical and operational expertise and invested in our technical infrastructure to support the delivery of the AI. Investment in these areas has been done not only to support the Qure.ai project, but also to provide a foundation for adoption of more tools in the future. This includes:
- Rigorous clinical governance oversight
- Adaptive technical hosting solutions with flexible integration options
- Evaluated in ‘real world’ scenarios by a selected pool of reporters
- Workflow optimisation team
With the Augmented Intelligence project team and infrastructure in place we ran an evaluation of qER with a selected pool of NightHawk reporters. This allowed us to check that the performance correlated to the published data and to optimise the workflow in the Medica environment. The feedback was positive with some of the key findings being:
- Very low level of false negatives
- Acceptable level of false positives (with the reason for the false positive being obvious e.g. shunt)
- Good buy-in from reporters involved in the project
- The algorithm allowed the Nighthawk allocation team to consider prioritise on potential clinical need rather than ‘First In First Out’
Following the successful pilot, the tool went into full production in Q4 2020. We are gathering quantitative and qualitive feedback on an ongoing basis and will provide an update on the impact in the future.
The economics of AI
- For healthcare providers and Medica, the AI pricing model vendors adopt must be affordable in the context of the local healthcare economy
- Where Medica can achieve sufficient ROI through efficiency or quality gains there will be no additional charge to our clients
- Where a tool is does not deliver sufficient ROI, but a client wants access to the benefits this will be sold as a ‘paid for’ enhancement.
- Our investment in our FutureTech platform facilitates the rapid adoption of new algorithms from a variety of vendors on a pay as you go basis
- We will continue to explore opportunities to jointly develop tools and ownership of intellectual property with AI vendors.
An award winning integration
Medica’s augmented intelligence project to introduce AI into the Acute Workflow, was shortlisted in the APM project management awards 2021. The awards have global reach and are open to projects from all countries and organisations. They are widely regarded as the most prestigious and sought after recognition for excellence within the project management profession.
The project was recognised as the winner in the technology category and its importance to patient care led the judges to award it overall project of the year. A great testament to the dedication of the Medica team and the project management style
