Research highlights
Research on amyloid-beta has failed to deliver a treatment that slows cognitive decline in Alzheimer's disease patients. Instead, some believe that the cerebral vasculature may have a role in the onset and progression of the disease. However, it is not known how the vasculature changes in both its structure (angioarchitecture) and its function (blood flow) during the progression of Alzheimer's disease. To this goal, we imaged wild-type and Alzheimer's disease mice over 7 months and quantified 25 vascular properties from this data, consisting of 1D, 2D, 3D and 4D data types.
We used data that was simultaneously acquired using OCT and two-photon microscopy to uncover the distribution of parameters governing the height, width, and inter-peak time of peaks in OCT intensity associated with the passage of RBCs. This allowed us to simulate thousands of time-series examples for different flux values and signal-to-noise ratios, which we then used to train a 1D convolutional neural network (CNN). The trained CNN enabled robust measurement of RBC flux across the entire network of hundreds of capillaries.
This paper describes an image processing pipeline to mitigate artifacts in OCT angiograms which ordinarily prevent the quantification of vascular properties. This allows for the extraction of information on the morphology and connectivity of vessels from three-dimensional images of the cortical murine vasculature.
The position of the focal plane within an imaging volume is an important factor in the accurate determination of the optical attenuation coefficient. However, it is difficult to estimate especially as it is often tilted, or curved. This method estimates the 3D position of the focal plane, even for high-NA lenses. The attenuation coefficient has a number of applications, but shown here is the possibility to delineate cancerous from healthy tissue.
Ordinarily in clinical practice, EEG data is assessed in a subjective manner. To showcase how EEG may be treated more objectively, we extracted a number of different features from data from comatose patients. We applied machine-learning techniques to build a classifier to index consciousness and predict patient outcome.