By Katie Haley
Artificial Intelligence (AI) has been evolving in the healthcare field for years, and manufacturers are advertising their AI advancements now more than ever. Healthcare providers are curious about emerging technology and are actively exploring the best ways to implement these solutions. One of the most compelling appeals is the ability to achieve quicker, more precise diagnoses using AI assistance built into medical software and devices; a capability that has been successfully implemented across many areas of care delivery, most notably in diagnostic radiology.
Companies have also been expanding AI solutions to diagnose issues with the medical equipment. Some manufacturers are recommending predictive maintenance (PdM) to anticipate potential points of failure within a device before an unplanned outage occurs, while others offer real-time, large-scale troubleshooting support tools.
The goal in any clinical environment is to provide continuous, uninterrupted care for patients. HTM departments are dedicated to supporting this mission by minimizing equipment downtime. For example, if a CT or MRI scanner experiences a prolonged interruption, patient visits may be delayed or canceled. AI-driven PdM algorithms are being leveraged to increase equipment uptime and reduce costs by decreasing the frequency of emergency repairs.
USE CASES
GE HealthCare offers a product called OnWatch Predict for its CT and MRI systems. It is a continuous monitoring platform that tracks critical components of the imaging equipment, compares performance data against a validated virtual model and uses AI technology to predict failures before they occur. GE estimates an average of 22% reduction in unplanned downtime, resulting in approximately 99% equipment uptime when using the OnWatch programs.
Aquant offers AI assistance for medical device repair and proactive maintenance. Field service engineers and technicians use the platform for real-time AI-driven support when troubleshooting and resolving medical equipment issues. Aquant estimates a 28% reduction in service costs and fourfold improvement in resolution speed.
Oxmaint AI integrates device telemetry directly into its maintenance platform. Internet of Things (IoT) sensors installed on the medical devices send data continuously to the system, where it is analyzed by AI. If a discrepancy is detected, such as an abnormal device temperature, amperage reading, or other monitored parameters, the system flags the unit for review. This approach provides constant monitoring of entire device fleets within a facility.
POTENTIAL RISKS, PATIENT SAFETY & LIABILITY
Many companies are now offering AI solutions as a resource for HTM teams, and PdM technology is becoming increasingly prominent in the industry. However, adoption has been met with resistance. Some HTM professionals are unfamiliar with AI-driven platforms and may not fully trust their recommendations. Others have expressed concern that widespread AI implementation could threaten job security. Beyond these organizational considerations, there are genuine technical and clinical risks associated with using AI to diagnose medical equipment failure. Some potential risks include:
• Inaccurate Predictions: Even though AI models undergo extensive testing prior to release, there remains a risk of incorrect recommendations or missed critical findings during analysis. No algorithm is infallible, and decisions made on the basis of inaccurate AI guidance can have real consequences for equipment reliability and patient safety.
• Cybersecurity Exposures: Cybersecurity risk is a significant vulnerability with AI-driven maintenance platforms. Protected Health Information (PHI) and Personally Identifiable Information (PII) may be transmitted through sensors and incorporated into AI algorithms. If an algorithm is manipulated by a malicious actor, the consequences for a facility could be severe. These systems are also susceptible to data breaches and leaks, which can expose both patient information and medical equipment to harm.
REDUCED EQUIPMENT OVERSIGHT
When AI is used as the primary evaluator of potential equipment issues, there is a risk that components or conditions that would be readily apparent to an experienced technician may be overlooked. It is imperative that human oversight remain a central part of equipment management. AI should augment, not replace, the judgment of skilled HTM professionals.
• Lack of Available Data: For AI to function as intended, it requires substantial volumes of historical data to anticipate failure points. Facilities with limited data histories or newer equipment inventories may find that AI-driven predictions are less reliable until sufficient data has been accumulated.
ECRI’s #1 concern for patient safety in 2026 was the use of AI technology in healthcare organizations. While this determination focused primarily on clinical diagnoses, the implications extend directly to medical device management. HTM departments must ensure that equipment is reliable and functioning correctly to protect patient safety. When medical devices are in use in a clinical setting, they must be validated as safe, and that validation remains the responsibility of the HTM technician. An AI program will not sign an inspection notice, work order, calibration certificate or any other form of verification. The technician will. This distinction reinforces a critical principle: humans must use AI as guidance, not as a replacement for professional judgment.
AI is at the forefront of medical equipment marketing, and there are many factors to consider when evaluating whether these advancements are appropriate for a particular hospital or facility. Key considerations include the extent of manufacturer testing, usability and relevance to the clinical environment, user acceptance and funding. Funding must be considered not only for initial procurement, but also any costs that go into supporting and maintaining the system past the OEM’s end-of-life designation.
While emerging technologies can assist in day-to-day operations, it is important to monitor AI-driven processes with careful oversight. AI is designed to aid humans, not replace them. This principle is especially important within the HTM environment, where the reliability of medical equipment directly impacts the safety of the patients.
