Multiscale Entropy

Interest in entropy-based tools for biological signal analysis has surged since the introduction of the multiscale entropy (MSE) method. This approach quantifies the disorder, or entropy, of physiological outputs across multiple time scales, resulting in an MSE curve that estimates the system’s complexity. According to the complexity-loss theory, aging and disease degrade this complexity, a notion supported by recent findings in conditions such as depression, Parkinson’s disease, schizophrenia, and Alzheimer’s disease. In this talk, we will provide a concise overview of the fundamentals of complexity analysis in electrophysiological signals and explore their relevance to neurodegenerative disorder diagnosis. We will demonstrate the utility of MSE tools using micro-electroretinogram recordings from healthy and transgenic mice (5XFAD, Tg XBP1s, and Tg XBP1s/5XFAD) obtained via multi-electrode arrays in response to various visual stimuli. We will highlight the potential of non-traditional retinal electrophysiology assessment in developing novel diagnostic methods for Alzheimer’s disease through the eye, including the use of novel stimulation protocols. The MSE analyses show great promise and could appeal to a broad neuroscience and BME audience. For researchers examining biological system complexity, MSE tools offer valuable applications across a range of biomedical challenges, potentially paving the way for innovative diagnostic techniques and further research opportunities.

The retina has long been considered a window into the brain, generating significant interest in identifying potential ocular biomarkers for brain disorders. Functional changes in the retina can be detected using electroretinography (ERG). In fact, recent research has revealed alterations in the characteristics of the ERG in patients with Alzheimer’s disease (AD). However, the effects of aging and neurodegeneration on retinal electrophysiology remain unclear, as do the most effective tools to detect these changes. Furthermore, recently proposed protective mechanisms against brain cell senescence, such as the overexpression of the transcription factor XBP1s, may be reflected in retinal electrophysiology. In this study we used entropy tools to explore the complexity of retinal dynamics in two age stages: young (2-5 months) and adult (5-8 months) in four mouse models: wild-type (WT) (n= 8 young, 9 adult), AD 5xFAD transgenic model (n = 9, 10), Tg XBP1s mice (n = 7, 6), and a crossbreed Tg XBP1s / 5xFAD (n = 6, 7). We applied a chirp light stimulus—a flashlight followed by sinusoidal light of increasing frequency and then increasing amplitude light—to retinas ex vivo while recording μERG responses from a multielectrode recording array (MEA). We calculated the multiscale entropy (MSE) of the μERG signals and estimated a complexity index as the slope of the linear regression of the MSE curves. The median complexity index (upper: lower quartile) of all young groups was positive: WT young 4.6 (6:1.4), 5xFAD young 2.4 (3.5: -1.7), Tg XBP1s young 1.7 (2.4: 0.18) and Tg XBP1s / 5xFAD young 0.94 (1.7: -1.4), which was higher than those of the respective older groups: WT adult -2.2 (0.98:-11.4), 5xFAD adult -5.3 (-3:-8.9), Tg XBP1s adult 0.09 (3.4:-2.1) and Tg XBP1s / 5xFAD adult 0.89 (1.7:-1.4). This reduction in complexity among older groups supports the theory of loss of complexity with age. Furthermore, the 5xFAD groups exhibited lower complexity compared to their respective WT groups, consistent with a theory of loss of complexity in the disease. Tg XBP1s overexpression appeared to confer protection against aging, a benefit that extended to Tg XBP1s / 5xFAD animals, suggesting that Tg XBP1s may counteract aging and neurodegenerative mechanisms in the retina. These findings not only advance our understanding of retinal electrophysiology with aging but also have significant implications for the identification of new AD biomarkers and the development of novel therapeutic strategies.