By Ciprian (Chip) Ionita, PhD

Artificial intelligence (AI) is transforming just about every industry, and especially medical care. One pioneering application of AI is QAS.AI, a technology designed to assist neurosurgeons by providing them with real-time predictions of surgical outcomes. QAS.AI stands for Quantitative Angiographic Systems-Artificial Intelligence, a groundbreaking software developed at University at Buffalo that merges advanced imaging techniques with machine learning to enhance the precision and safety of brain surgeries.

For example, an aneurysm, which is a bulge or ballooning in a blood vessel caused by a weakness in the blood vessel wall, can lead to life-threatening bleeding if it ruptures. To detect and assess the severity of an aneurysm, doctors often use angiography, an imaging technique in which a special dye is injected into the patient’s bloodstream that is visible under X-rays. Angiography enables a neurosurgeon to observe blood flow and identify abnormalities like aneurysms.

Treating an aneurysm could require highly invasive open brain surgery which carries significant risks and recovery time. Fortunately, new technologies and minimally invasive treatments performed during an angiography allow a neurosurgeon to navigate through the vascular system using catheters (thin tubes) introduced from a small incision in the groin or wrist. The neurosurgeon can then deploy devices like coils or flow diverters directly to the site of the aneurysm to stabilize the vessel and prevent it from rupturing.

QAS.AI enhances these minimally invasive procedures by providing neurosurgeons with sophisticated tools to predict surgical outcomes in real time. By integrating AI with the detailed visuals from angiography, surgeons can view the current state of the aneurysm and how it might respond to different treatment approaches. This predictive power is vital for making informed decisions during surgery, significantly improving the chances of a successful outcome and reducing the risks associated with aneurysm treatment.

The effectiveness of QAS.AI depends significantly on its algorithm, initially trained using data from over 500 patient cases. To accomplish this, comprehensive angiographic imaging data were collected and paired with detailed records of surgical outcomes in patients. The AI was trained to recognize and interpret patients’ contrast flow patterns within the angiograms to understand how these patterns correlated with their outcomes post-surgery. Through analyzing how specific contrast flow patterns aligned with surgical successes or complications, the AI was trained to predict intra-operatively, six-months potential outcomes for new cases. In other words, as a surgery progresses, QAS.AI provides real-time feedback based on the analysis of contrast flow patterns observed during the procedure. This feedback allows neurosurgeons to make informed decisions in real-time, adjusting surgical techniques such as the placement of coils or stents to optimize patient outcomes.

The ability to adapt surgical strategy based on live AI-generated insights significantly enhances the potential for successful interventions. Integrating AI into surgical practice improves diagnostic accuracy and allows for dynamic surgical planning and adjustments in strategy based on AI-generated predictions of surgical success or complications. Additionally, the more data QAS.AI processes, the more accurate it becomes. Thus, each new case contributes to the ‘learning’ of the system to help continuously improve its predictive algorithms.

As QAS.AI continues to evolve, its developers aim to expand its application to other neurovascular conditions and integrate it into routine surgical workflows across the US and the globe. However, challenges such as data privacy, integration with existing medical systems, and the need for regulatory approvals remain. Addressing these challenges is crucial for the wider adoption and ethical use of AI in medical settings.

QAS.AI exemplifies the incredible potential of artificial intelligence to revolutionize medical treatments. By providing neurosurgeons with a powerful tool that offers real-time, data-driven insights, QAS.AI is not just enhancing surgical precision, but is also paving the way for a new era in medical technology where AI and human expertise converge to save lives and improve patient outcomes.

It is important to note that the development of QAS.AI is a result of UB Neurosurgery’s dedication to excellence and L. Nelson Hopkins, MD, FACS’ vison for the Gates Vascular Institute environment. Without Dr. Hopkins’ vision of bringing engineers and doctors in the same place this would have probably not happened.

Ciprian N. Ionita PhD is Associate Professor of the Departments of Biomedical Engineering and Neurosurgery, Canon Stroke and Vascular Research Center, and CEO and cofounder of QAS.AI, 8052 CTRC, 875 Ellicot Street, Buffalo, NY 14203. Learn more at https://www.qas.ai. You can reach Dr. Ionita at 716-829-5413 or cnionita@buffalo.edu.