Share this post:
As healthcare costs continue to rise, governments, healthcare providers and insurers must find more efficient ways to maintain quality and improve outcomes for patients.
They’re moving away from traditional reliance on acute and primary care delivered by specialist physicians at hospitals and clinics toward a model in which patients and their families manage their conditions at home reducing the need for expensive hospital visits.
Huge advances have made in-home care much more practical, says Philip Daffas, CEO and managing director of ePAT Technologies, which specializes in innovative pain management solutions. A digital blood pressure monitor costs just a few dollars. Automated glucose monitors and insulin pumps help patients keep blood sugar levels in check.
But not everything is quite that easy. Pain, the so-called “fifth vital sign” has always been difficult to measure. Infants, dementia sufferers and many other vulnerable patients may be unable to tell their physicians about the pain they are feeling. Today, pain assessment is a specialist field, requiring the skills and experience of experts to ensure accurate results.
Scientists from Curtin University in Perth, Western Australia, one of the world’s leading pain management research centers, founded ePat to combine cutting-edge academic research with a new generation of cognitive computing technologies, such as machine learning and artificial intelligence (AI). One way they’re doing that is by analyzing facial movements linked to pain.
A robust architecture
Maker of AI emotion recognition software nViso and ePAT began developing an app that uses a smartphone camera to record a ten-second video of a patient’s face. It then analyzes the images using nViso’s sophisticated facial analysis algorithms. If it recognizes any muscle movements that indicate pain, it notes them.
Next, the caregiver uses the app to answer questions resulting in a total of 49 pain indicators, such as how the patient is moving, and how they are vocalizing their pain. Finally, the app calculates an overall pain score for the patient, and syncs the results with a central database in the cloud.
The design of the device needed to embrace “offline first” principles: it had to work without a data connection in case a user is temporarily unable to connect to the internet. At the same time, it needed to send data to the cloud when a connection is re-established so the patient’s data is stored and protected.
IBM Cloudant offered replication that made it easier to store data locally on the device and sync it with the central cloud data store. It’s also a fully managed service, which means no worries about database management.
Prototype to production
A number of major healthcare organizations in Australia are trying out the ePat app, and it has passed its initial validation and implementation studies with flying colors.
Yet the current version of the app is just a first step on the journey, according to Tim Llewellynn, CEO and co-founder of nViso. The companies plan to expand AI in to other areas too, using IBM Watson cognitive services to augment capabilities.
For example, Watson Visual Recognition could be used to scan barcodes on medication bottles so patients know they’re administering medication in the right amounts at the right times.
Lean more about IBM Cloud healthcare solutions.