Abstract
Purpose :
The purpose of this study was to evaluate the potential of multispectral high-resolution imaging to identify amyloid β-protein (Aβ) deposits in the retina as a biomarker for neurodegenerative diseases such as Alzheimer’s (AD). We combined simultaneous multi-channels offset and confocal AO-SLO imaging, which provides isotropic images of retinal microstructures free of directionality artifacts, with spectral signature analysis of chemical compounds to identify such biomarkers at cellular-level.
Methods :
We developed a multispectral imaging technique for identifying chemical compounds in vivo in the retina with cellular-level resolution and without the use of contrast agents. Since various retinal structures and cells have different reflectance and absorbance properties, information about cell metabolism and structure can be obtained using spectral imaging. Our method is based on using multiple wavelengths for which Aβ has different absorbances (400-600 nm range). In this way, Aβ can be spectrally differentiated from the surrounding tissue. Our instrument is based on PSI high-resolution retinal imaging platform (MAORI for human and MAOSI for rodent imaging) with multi-offset capabilities. However, similar to all current AO-SLO systems, our instruments use a single light source (20-50 nm bandwidth) and detect the reflected light with detectors that integrate over the entire spectrum, without spectral differentiation capabilities. To add multispectral capabilities, we replaced the SLO light source with a time-wavelength multiplexed source to which we synchronized the detection unit.
Results :
One confocal and four offset images are acquired for each individual illumination wavelength. Multiple split, phase, phase gradient, sum (ring detection), and standard deviation images are derived for each color and are combined in a manner that highlights spectrally the local variations of absorbance. A similar algorithm to the one used in blood oximetry based on multiple wavelengths has been developed and implemented for Aβ identification. The new concept has been demonstrated in a model eye using commercial Aβ. An animal study on a rat AD model is ongoing to demonstrate the technique in vivo.
Conclusions :
PSI has developed a novel multispectral adaptive optics-based technique for detection of Aβ at cellular-level in the retina. This technique may pave a path forward for better understanding of the onset of various neurodegenerative diseases.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.