Novel hand-held eye (retina) scanning device measures biomarkers and accurately identifies brain injury in pig eyes

By Susan Klein. This article was initially published in the 1/11/24 edition of our Concussion Update newsletter; please consider subscribing.

Once the soft spot on a baby’s head closes over with the skull, the only easily accessible opening of the brain to our viewing is through our eyes. Carl Banbury and colleagues at the University of Birmingham (UK) worked with fellow UK co-investigators from the (Birmingham) and the Cavendish Physics Lab (Cambridge) to develop a handheld concussion screener (EyeD) that looks into the “window into the mind.” In their recent Science Advances article, they described a handheld device, tested on a pig model, that can scan the retina of the eye to detect biochemical markers produced by the brain immediately after mild traumatic brain injury (mTBI). 

The EyeD device promises to improve the accuracy of field diagnosis of concussion by non-medical evaluators. On-the-spot, accurate diagnosis of concussion is a key step toward identifying who will need intervention as soon as possible after an mTBI.

The authors observe that the retinal ganglion cells, visible by looking through the eye’s pupil with special equipment, are unmyelinated––they are not covered with the protective insulation of myelin, unlike the optic nerve that transmits information back to the brain. Therefore, changes that are assessed in the retinal ganglion cells “relate directly to neuronal biology.” The researchers’ device detects molecular biomarkers of brain injury that can be measured in the retinal ganglion cells, using a combination of an eye-safe parallel laser light beam with a physics method called multiplex resonance Raman spectroscopy. This specialized spectroscopy has been used by these authors and others to evaluate biomarkers of interest in mouse tissue sections, blood, and cerebrospinal fluid.

They go on to describe how this spectroscopy technique was built into a handheld device to scan the eye (and the retina). The device uses algorithmic data reduction and clustering of signal information to yield a map-type profile that identifies circulating biomarkers of interest in acute brain injury that are detected on the retinal surface. (The data reduction and clustering is accomplished with SKINET, a Kohonen network method developed by this same team.) They detail trials with:

1) a phantom eye device; a phantom is an object that is “designed to mimic the properties of human tissue.”

2) mouse eyes––brain-injured and control groups

3) ultimately, pig eyes are the closest in size and optics to human eyes. The pig eyes were obtained from pork processing plants (head-injured) and control pigs (who died on farms of natural causes). To corroborate their findings from Raman spectroscopy signals obtained with the EyeD device, the researchers did biochemical analyses of the pig eyes after death. They found chemical changes to the “relative lipid and protein composition” in the in pigs identified as brain-injured that could be linked to Raman spectroscopy signals obtained with the EyeD device.  

The authors note limitations of the study, including the inability to quantitate precisely the level of brain injury in the pigs beyond severe and less severe, with less robust results in less severely brain-injured pigs. They point out, however, the consistency of data obtained across the animal models. Aside from its obvious utility in brain injury assessment, the EyeD device offers promise for other forms of retinal disease. We look forward to trials of this interesting device in humans going forward.