Panoramic hyperspectral imaging system links heart scar tissue to arrhythmias

A panoramic imaging system that unites hyperspectral and optical mapping techniques has linked the structure of scarred heart tissue directly to the abnormal electrical signals that can trigger dangerous arrhythmias

When a heart attack damages the heart muscle, the injured tissue is gradually replaced by scarring which can interfere with the electrical signals that coordinate each heartbeat. This has long been associated with an increased risk of arrhythmias – dangerous heartbeat rhythm disorders that can follow from a cardiac injury. While scientists have been able to map the heart's electrical activity for decades, it has proved difficult to relate those signals directly to scar tissues.

A team at the George Washington University, Washington DC, USA, led by Professor Matthew Kay of the university's Department of Biomedical Engineering, has now developed a novel imaging system designed to close that gap. The technique combines hyperspectral imaging which identifies different tissue types according to how they interact with light, with optical mapping, an established method used to track electrical activity in heart muscle. Together, the two produce a panoramic view of the heart that shows where scar tissue is located and how electrical signals move around it.

Most existing optical mapping systems can show how electrical activity spreads across the heart but they offer little insight into the condition of the tissue beneath. This makes it hard to establish why abnormal rhythms tend to begin in particular areas. The team designed the system to measure tissue composition and electrical behaviour at the same time – across the whole heart surface – rather than in isolated sections.

To achieve this, the team built an imaging platform that incorporates six cameras, each positioned to capture the heart from a different angle. Four high-speed complementary metal-oxide-semiconductor cameras recorded electrical signals using a fluorescent dye that responds to changes in voltage. A further camera gathered detailed spectral information from the tissue itself, while an additional camera was used to reconstruct the heart's three-dimensional shape. Specialised software then merged all of this data to produce a single map of the entire heart surface.

To test the system, the researchers studied the hearts of rat for a month after a heart attack had been induced. Each heart was removed, kept alive by means of a perfusion system, and scanned while the team recorded tissue properties and electrical activity simultaneously.

The hyperspectral measurements proved particularly useful for locating scar tissue. Scarred regions were found to emit a stronger fluorescence signal associated with collagen, a protein that accumulates in connective tissue after cardiac injury, within a specific band of the visible-light spectrum, approximately 400–520 nanometres. To classify tissue type from this spectral data, the team applied an automated k-means clustering algorithm, which produced high-resolution maps distinguishing healthy heart muscle, damaged tissue and the border zone that separates the two. Laboratory analysis of the heart tissue subsequently confirmed the accuracy of the imaging results.

Once the tissue and electrical maps had been combined, the team could see directly how electrical signals behaved close to the scar. In three of the four hearts studied, abnormal beats originated close to the border between healthy and damaged tissue. Electrical waves generally travelled quickly around the scar itself but slowed markedly, or were blocked altogether, when they attempted to pass through scarred regions – disruptions of this kind are thought to be the conditions in which arrhythmias develop.

The system also revealed clear differences in the electrical properties of the various tissue types. Signals recorded from scarred regions remained active for longer than those from healthy muscle, while the border zone showed an intermediate pattern. This behaviour was consistent across every heart examined and reflects changes already known to occur in the aftermath of a heart attack.

Beyond the study of heart attacks specifically, the researchers believe the technique could help investigate other conditions that alter the structure of heart tissue, including ageing, fibrosis, heart failure and the effects of treatments such as catheter ablation. According to the team, this is the first panoramic system to combine detailed tissue characterisation with electrical mapping across the whole surface of a living heart.

For further reading please visit: 10.1117/1.JBO.31.7.076004