Medical Device Uses AI To Predict Lung Cancer
AI for medical devices


When you think of medical devices, what comes to mind? A purewick catheter? A hernia mesh? An EKG machine? All are definitely common examples, but you may be wondering if software with a medical application falls under that category. Well, they absolutely do, as software powers the healthcare world. The FDA classifies it under software as a medical device. Here are just a few examples:

  • Apps to view scan results on a computer
  • An app that collects sleep data through a patient’s smartphone
  • Software that compiles radiation imaging to create 3D models.

Now as we venture further into the 2020s, some of these medical devices are incorporating AI. These AI applications are accelerating progress in the healthcare industry faster than we could’ve ever imagined. 

One of the most recent and impressive ways AI has worked with medical devices is in predicting diseases in a patient before they happen. We’ll discuss this groundbreaking technology as well as how it may impact the information flow regarding medical devices, and how your organization can benefit.

AI-Powered Medical Devices For The Future – Enter Sybil

Around 50 years ago, the first CT (computed tomography) scanners were installed in hospitals across the US. Fast-forward to 2010 and these medical devices evolved to Low-dose computed tomography (LDCT). Today, healthcare professionals use LCDT to screen patients for heavy smokers over 50 for lung cancer. In the US, however, less than 10% of the people who should be screened actually are. Additionally, lung cancer diagnoses among light smokers and non-smokers are rising at an alarming rate. So if researchers continue focusing mainly on heavy smokers, they may unintentionally create a wider gap between the people screened and the people who need screening.

Researchers at MIT have trained and validated an AI model they call Sybil. So how does Sybil factor into this issue regarding lung cancer screenings? Well, the researchers trained Sybil on tens of thousands of LCDT images with the goal of accurately predicting lung cancer in patients. The result was astounding. Sybil was able to predict patients’ lung cancer development up to 6 years in advance. At the moment, Sybil is only for research purposes. However, if it maintains its traction, Sybil and future AI-powered medical devices may assist in clinics all over the world.

How Sybil & Future Healthcare-Oriented AI Models Impact MDR Reporting?

Many of us in the healthcare industry are enthralled by the direct life-saving potential of AI in medical device software. However, the indirect lifesaving potential of this software via the information flow in the healthcare field through MDR reporting is also exciting.

How does this evolution in software as medical devices affect the medical device adverse event report (MDR) process? Well, proper MDR reporting ultimately relies on information and drawing connections. If medical devices cause instant injuries like burning, cutting, etc. they should be reported to the FDA within 5 to 30 days.

What if, however, a certain medical device can cause lung cancer? Think about the time enough information surfaces for researchers, the FDA, or any other party to draw connections between the device and the harm it’s inflicting on patients. This unfortunately delays the information necessary to intervene for years. By the time the FDA or the device’s manufacturer issues a recall, it may have already harmed if not killed potentially hundreds of thousands of patients.

In this case, Sybil could intervene early by detecting the signs of lung cancer up to 6 years in advance. Sybil can then predict illness in enough patients to draw connections between devices and instances of lung cancer. This in turn would not only allow those affected to be treated before cancer even starts, but provide the information necessary to justify recalling the device. With this information available so much sooner, healthcare organizations, manufacturers, and the FDA  could dramatically accelerate their recall process, preventing untold damage.

How To Leverage This Trend To Optimize Organization & User Safety

This technology will likely become more prevalent in the healthcare industry. As this trend continues, MDR reports for medical devices causing delayed damage can start flowing into FDA databases much sooner. Does your organization rely on medical devices? If so, why not take advantage of this acceleration in the information flow to take proactive measures?

Device Events is an intuitive software tool that utilizes natural language processing algorithms to help deliver clear and comprehensive metrics and reports on the millions of complex medical device adverse event reports (MDRs) and recalls filed with the FDA. So naturally, if AI helps speed up MDR reporting across the industry, then using Device Events software to gather data as quickly as it comes out is crucial. Taking advantage of this shift in the industry can help your organization protect itself from delayed adverse events from defective medical devices.

Are you ready to leverage the future? Contact us to learn more.