AI and machine learning continue to redefine and reshuffle the healthcare landscape.

Puneet Pandit

June 27, 2018

5 Min Read
Reinventing the Healthcare Continuum with Machine Data Analytics
ColiN00B/Pixabay

As healthcare delivery organizations navigate the dual challenges of providing high-quality medical care while facing tougher cost control measures from insurance carriers and government entities, one strategy they are turning to is optimization of both expensive equipment, as well as the entire ecosystem in which these assets operate. This is no small task, as capital expenditures for these machines totaled more than $350 billion in 2016.[1]

Innovators in many industries have adopted asset optimization strategies for several years. Now, leaders in healthcare equipment manufacturing are turning to this approach by implementing a new generation of data management and predictive analytics solutions to increase the quality, consistency and efficiency of medical equipment in support of patient care. To accelerate adoption of asset optimization solutions, manufacturers are turning to data management and predictive analytics solution providers. These players bring years of experience and best practices gleaned from deployments in multiple industries in addition to the use of artificial intelligence (AI) applications powered by machine learning to drive business impact.

Today’s medical equipment systems produce increasing amounts of complex machine data that require more advanced data transformation solutions. These solutions include deeper root cause analysis, predictive analytics, machine learning, AI and other high value support applications. Manufacturers can now provide insights far beyond basic diagnostics to help healthcare providers anticipate and address equipment failures and maintenance. Today’s industry standard of 95% to 96% machine uptime can be increased to more than 99.5%, resulting in millions of dollars per year in reclaimed revenue from operational budgets.

AI applications powered by machine learning models are being used to predict part failures in expensive imaging modalities, revolutionizing the landscape of how equipment maintenance is performed today by in-house support staff at healthcare providers, independent service organizations (ISOs), and by the OEMs themselves. For example, replacing a CT scanner's X-ray tube today is more an art than a science. It is reliant on several ad hoc data inputs based on age of the machine, number of scans performed, image quality rendered, and other subjective factors. Without proper diagnostics on machine data signals, many companies end up replacing tubes under the gun to ensure machine uptime at any cost.

This move also transforms the relationship between OEMs and healthcare providers. Manufacturers have the potential to become advocates for operational effectiveness, helping providers deploy best practices throughout the delivery of services. For example, data collected by a medical imaging machine can determine that the room in which the machine operates is too warm or too cool, and automatically reset the temperature. The machine can also analyze how long each imaging test takes and recommend additional training for operators who take excessive time in completing the test.

Because manufacturers collect data uploaded from machines and equipment deployed at many healthcare providers, they are able to collect larger and more relevant data sets with which to conduct analytics. From a predictive analytics perspective, this is a critical point. A large equipment manufacturer is able to offer predictive insights based on a dataset from several thousand machines deployed in a wide range of operating environment around the world, versus insights from 10 or 20 machines deployed at a specific facility. This leads to deeper, more accurate insights that contribute to enhanced operational effectiveness.

Working with a data transformation and predictive analytics solution provider can also result in a dramatic increase to the equipment provider’s dataset. Many machines may lack an adequate number of sensors to provide a complete picture of the equipment’s performance. It is not practical to install these sensors retroactively or for a healthcare delivery provider to wait for the next generation of equipment. Leading solution providers have the ability to integrate machine log data into the comprehensive data set, enabling manufacturers to obtain the high-quality predictive insights they need.

New advances in data management and analytics have powered this transformation. The most significant advancements include the ability to:

  • Ingest, parse and analyze structured, semi-structured and non-structured complex log and sensor data.

  • Deploy rapidly, which reduces implementation and staff costs.

  • Manage complex log data to uncover more variables for analysis than traditional solutions.

  • Integrate multiple data types to analyze increasingly complex applications.

  • Enable machine learning and predictive analytics on complex log data.

The evolution of manufacturers to provide enhanced analytics that contribute to improved operational effectiveness also enables introduction of new services. Among these:

  • “Equipment as a service” – Instead of purchasing imaging equipment, a healthcare provider might purchase x thousand scans per month on a subscription basis.

  • Automating the spare parts and service inventory and delivery system for routine maintenance and repairs.

  • Implementing fact-based versus time-interval based maintenance through predictive support to become aware in advance of potential machine failure.

  • Providing managed services through which the manufacturer continuously monitors machine performance and can often execute repairs remotely, without the healthcare provider’s involvement.

This also assists healthcare delivery providers through a current difficult demographic challenge. Older engineers are retiring at a rapid pace, often taking their institutional knowledge gained over decades of work with them. Equipment manufacturers can utilize their data and predictive analytics to capture this knowledge and seamlessly share the experiences and insights with a new generation of engineers, dramatically reducing costs associated with training and onboarding.

As healthcare equipment manufacturers re-imagine their role in providing products and services to healthcare providers, providers must also re-evaluate their relationships with manufacturers and how they can contribute to enhanced operational effectiveness. Innovative equipment manufacturers in partnership with innovative data transformation and predictive analytics solutions providers can help healthcare delivery providers navigate today’s asset optimization as well as human resources challenges.

[1] Harbor Research, “Data Transformation Drives Machine Intelligence in Healthcare,” April 2018, page 4

About the Author(s)

Puneet Pandit

Puneet Pandit is the Founder and CEO of Glassbeam, the premier machine data analytics company. With more than 20 years of global IT experience Puneet founded Glassbeam in 2009 with the mission of bringing structure and meaning to complex data generated from any connected machine in the Industrial IoT industry.  With a strong focus on medical devices, Glassbeam’s next generation cloud-based platform is designed to transform, analyze, and build Artificial Intelligence applications from multi-structured logs, delivering powerful solutions on customer support and product intelligence for companies.

Sign up for the QMED & MD+DI Daily newsletter.

You May Also Like