Illinois bioengineering professor Rohit Bhargava, research scientist Sudipta Mukherjee, and a multidisciplinary team have developed a groundbreaking diagnostic method for cardiac amyloidosis using infrared spectroscopic imaging and AI. Current diagnostic techniques require invasive biopsies, specialized facilities, and extensive wait times. The new method uses a specialized infrared microscope to detect molecular “fingerprints” of proteins, analyzed by an AI neural network trained to identify those causing amyloidosis. This fully automated approach delivers results in just 10 minutes, is non-destructive, and doesn’t require costly mass spectrometry facilities. With clinical trial discussions underway with the Mayo Clinic, the method promises faster, more accessible, and cost-effective diagnoses, advancing healthcare equity and improving patient outcomes.
Written by Ben Libman
Research scientist Sudipta Mukherjee
The importance of heart health cannot be overstated. As the centerpiece of our circulatory system, the heart's well-being is essential to overall health. Yet heart disease is pervasive, and cardiac amyloidosis—a condition where misfolded proteins accumulate in the heart—can be lethal. Different types of proteins can cause these buildups, and each requires a unique treatment. This makes early and accurate diagnosis critical, but current diagnostic methods are often inaccessible. The traditional method involves a biopsy and examination by a certified cardiac pathologist, who will then dissect the sample and send it to a specialized facility to identify the offending protein. This is time-intensive, costly, and destroys the tissue sample, often leaving no room for further analysis.
A multidisciplinary team from Illinois, including research scientist Sudipta Mukherjee and bioengineering professor Rohit Bhargava, collaborated with clinicians from multiple medical centers to overcome these drawbacks. Their solution was to use a technique called infrared spectroscopic imaging, combined with artificial intelligence - an approach pioneered by the Bhargava laboratory. Infrared spectroscopy is a technique that identifies molecules by analyzing how they absorb light. Each molecule in a sample responds to infrared light in a unique way, giving off an identifiable “fingerprint.” The team imaged tissue samples using a specialized infrared microscope, which measured how light interacted with the sample. This data was analyzed by an AI neural network trained to recognize the specific spectral patterns of the proteins and thereby identify the one that led to cardiac amyloidosis with accuracy comparable to current diagnostic protocol. Author Sudipta Mukherjee explains the significance of this finding: “While infrared spectroscopy is known for its sensitivity to protein structure and identity, human tissue chemistry is exceptionally complex. In this complex biochemical milieu, pinpointing a single protein among thousands of others and myriads of non-protein components is challenging. Achieving this was the most surprising and groundbreaking aspect of our research.”
Bioengineering professor Rohit Bhargava, Grainger Distinguished Chair in the Department of Bioengineering
This new method has a lot of potential. It’s fully automated, meaning no human intervention is needed during the detection process. It’s faster- while the current standard method can take two weeks or longer, this AI network provides results in just ten minutes. And finally, it does not need a Mass Spectrometry-based proteomics facility, which have large overhead expenses and are inaccessible to most hospitals in the US and worldwide. This savings in time and human labor results in a significant reduction in cost. The new method also has the advantage of being non-destructive, meaning the tissue sample survives the test and can be reused, reducing the necessity for repeat biopsies.
While the potential of this new method is enormous, there are still more steps to be taken before it can be implemented in a clinical setting. A large-scale clinical trial would be necessary before this technology could get to patients. However, this paper represents the incredible potential such a trial could have and establishes the statistical parameters that are necessary for appropriate study design. Discussions to design a comprehensive clinical trial are underway with the Mayo Clinic, and this technology could be implemented after their successful completion. “Illinois’ department of bioengineering, the Carle Illinois College of Medicine, and the Cancer Center at Illinois are leading the development of technologies that can greatly improve human health,” says professor Bhargava. “Translating this study to use with our clinical colleagues can lead to faster and accurate diagnoses that improve patients’ lives.”
The importance of accessibility in medicine is paramount. Yet too often, patients suffer with extended wait times and inadequate resources. The most impactful advancements in medicine today are the ones that make the best healthcare available to all, regardless of the patient’s circumstances. For those suffering from cardiac amyloidosis, this new method holds the promise to do just that, with earlier intervention, improved outcomes, and better equity in healthcare.