In Healthcare, Automation Is Not Just About Efficiency

Print Friendly, PDF & Email

In this special guest feature, Carla Leibowitz, Head of Strategy and Marketing at Arterys, discusses how deep learning tools can aid physicians in determining a patient’s condition more quickly and accurately and what promise this holds for personalized care. Before Arterys, Carla focused on strategic projects for Global 500 companies around precision medicine and SaaS a team lead at Bain & Co. Prior, she led the innovation group at Kyphon/Medtronic Spine and Biologics and helped teach the BioDesign class at Stanford University. She holds 16 patents for medical device technologies. Carla earned a B.S. from MIT as well an M.S. and an MBA from Stanford University.

Widespread adoption of artificial intelligence (AI) tools for healthcare is an exciting prospect, but has led to concerns that radiologists may one day be replaced by advanced technologies. Although we believe in the disruptive potential of AI, the radiologist’s job is far too complex and vital to patient care to be replaced. Just like Excel did not replace the analyst, we believe advanced software will make radiologists far more informed and efficient, allowing them to see more patients and spend more time on assessing the most difficult cases.

With the recent availability of medical imaging analytics cloud software with deep learning for cardiac MRI, many are asking whether the sole value of cloud infrastructure and precise automation is efficiency. It is true that the zero-footprint nature of cloud computing can significantly lower enterprise IT costs and that automation produces faster results. While these gains are important in a healthcare system with ever-increasing costs, we believe that the true value of precise automation is improvements to patient care. Automation can bring consistency where it matters most: in ensuring patient assessments are accurate to inform the optimal course of treatment.

Today, interpretation of medical images is often qualitative, and can vary significantly between clinicians. In the most difficult cases, even the same individual might read the same exam differently the second time. Furthermore, quantification of anatomy can be extremely tedious and time consuming. Lastly, these measurements are often shortcuts for what would otherwise take hours to measure in a specific study. For example, the standard measurement for many tumors consists of the largest diameter of the mass on a single slice, whereas most physicians would rather know the volume of the lesion.

Using Big Data and advanced analytics, we can now develop tools that not only automate the tedious process of quantifying images, but can also make these measurements far more consistent- no matter the experience level of the radiologist, or the type of scanner on which the image was acquired. This capability enables significantly more reliable assessments of whether there have been changes in the anatomy as a patient responds to therapy or as their disease progresses. The importance of this cannot be overestimated. An early assessment of a therapy’s efficacy can mean the difference between life and death, and greatly reduces time and money wasted on ineffective, expensive therapies.

Long term, once we can rely on these measurements, we can start using the data collected over time to develop predictive tools for treatment options. A growing, accessible database will help clinicians make informed decisions about which treatments are better suited for each patient, and to better identify patients  at  high risk of poor outcomes. While this is already underway in way the field of genomics, imaging has lagged behind. That is where the power of the cloud comes in. With sophisticated technology that protects patient identity, we can aggregate data from around the world – data once locked inside hospitals – to advance our understanding of disease and response to therapy.

The idea of value based care is fairly simple – it is about using our healthcare resources to maximize outcomes and minimize waste. What makes this so difficult today is that there is no standard system to measure patients consistently and collect enough of this information to develop new, optimized care pathways. Advanced AI technologies for imaging have the ability to solve this problem. In the short term, large gains will come from increased efficiency and more accurate measurements, but the most valuable benefit of cloud, big data and automation in imaging will be to help physicians determine the best course of action for their patients quickly and accurately.

 

Sign up for the free insideBIGDATA newsletter.

 

Speak Your Mind

*

Comments

  1. Automation is transforming every industry that it comes into contact with. It greatly improves efficiency, productivity, and minimizes mistakes by reducing the need for human intervention. The same holds true even for healthcare. Hospitals, as well as outpatient care centers, deal with a considerable load of operational tasks. The bigger the organization, the greater the workload. Healthcare automation can ease that load and boost the efficiency of workflows.

    Until automation was rolled out a few years ago, people had to call up the hospital, wait for the line to free up, have the available time slots told to them, and then finalize a time. On the other end of the line, a member of the staff would have had to attend each call, check a calendar for availability and inform patients. If a patient couldn’t attend for any reason, they had to call again and go through the whole process all over again.

    However, the growth of automation in medical workflows has led to the development of applications that make the whole appointment process simpler. Using these, people can just log in to a website or browse their smartphone to check for available appointment slots 24/7 and select one with just a few clicks. Moreover, if any patient wants to reschedule, he or she can do so without hassles. This functionality eases the burden on the staff at the provider organization, and even reduces staffing requirements.