AI and Perfusion: A Partnership, Not a Replacement

Artificial intelligence (AI) has woven itself into the fabric of nearly every industry—from finance and education to creative arts and clinical medicine. As perfusionists, we’ve watched this evolution with both curiosity and caution. There’s no denying the remarkable capabilities of AI in pattern recognition, predictive analytics, and automation. But when it comes to what we do in the operating room—responding to the unexpected, making real-time decisions under pressure, and managing the highly tactile and nuanced world of extracorporeal circulation—AI has a long way to go.

Let’s be clear: AI cannot do our job.

Perfusion requires a level of situational awareness and anticipatory thinking that can’t be coded into an algorithm. We manage circuits with our hands and eyes, our instincts honed by experience. We interpret subtle changes in hemodynamics, communicate with surgical and anesthesia teams, and pivot instantly during high-stakes moments. Whether it's rapidly priming a backup circuit or recognizing air entrainment before a machine alarm even sounds, our role is human and physical.

But that doesn’t mean AI has no place in our field. In fact, it’s quite the opposite—AI can serve as a powerful ally, especially when it comes to data evaluation and research.

Where AI Shines in Perfusion

1. Data Analysis and Clinical Decision Support

Perfusion generates a mountain of data—arterial and venous blood gases, ACTs, flow rates, pressures, temperature trends, and more. Manually sorting through these variables to draw meaningful conclusions is both time-consuming and limited by human bias. AI, particularly machine learning algorithms, can identify patterns and correlations across large datasets we might miss.

For example, predictive analytics can help identify early warning signs of post-operative complications like AKI or low cardiac output syndrome. AI can also assist in benchmarking outcomes across cases and institutions, helping us refine protocols based on real-world data.

2. Optimizing Protocols and Guidelines

AI-driven tools can analyze historical perfusion records to evaluate the efficacy of different cardioplegia strategies, circuit configurations, or anticoagulation protocols. By comparing large volumes of cases, AI can help us identify what works best for certain patient populations—offering a layer of insight to support evidence-based practice.

3. Enhancing Research

Perfusion research often involves retrospective chart reviews, data mining, and statistical modeling. AI can automate much of this process, reducing the time it takes to go from question to publication. Natural language processing (NLP), a branch of AI, can even be used to scan and extract data from operative notes or electronic medical records—tasks that would take a human hours to complete.

4. Education and Simulation

AI can also be used in the training of future perfusionists. Machine learning models integrated into simulation software can replicate patient responses more accurately, creating realistic and challenging scenarios. As educators, we can use these tools to better prepare students for the unpredictability of the operating room.

The Human Element Will Always Matter

The adoption of AI in healthcare raises important ethical and professional questions: Who is responsible when AI makes a mistake? How do we protect patient data? How do we ensure transparency in how decisions are made?

As perfusionists, we must be involved in shaping how AI is implemented in our field. We should be asking hard questions, contributing to research, and advocating for tools that support—not replace—clinical judgment.

Just as we use centrifugal pumps, gas blenders, and automated charting to make our work more efficient and precise, AI can be another tool in our toolbox. But it’s our experience, empathy, and hands-on skill that will always remain at the heart of patient care.

Final Thoughts

AI isn’t a threat to perfusion—it’s an opportunity. An opportunity to become more efficient, informed, and connected to the evolving world of medicine. Let’s embrace the technology with the same thoughtful approach we apply to every bypass run: informed, precise, and always with the patient in mind.