This PDF file contains the front matter associated with SPIE Proceedings Volume 12365, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Functional Near-InfraRed Spectroscopy (NIRS) (fNIRS) is a powerful method for non-invasively measuring cerebral hemodynamics on human subjects. Measurement contamination from superficial tissue which do not represent the brain continues to be an issue. We have proposed the Dual-Slope (DS) approach which is less sensitive to superficial tissue compared to typical Single-Distance (SD) methods. This DS method has been applied to Diffuse Optical Imaging (DOI), designing and constructing a large source-detector array. Previous results suggested that DS phase (Φ) has intrinsically higher sensitivity to the brain compared to SD Intensity (I). To further investigate this finding, on a large population of subjects, a modular DS array is designed. Allowing for collection from different cortical locations during various protocols. These source-detector modules are hexagonal and contain 4 intra-module DS sets. Tessellation greatly expands the number of measurement sets through the creation of inter-module DS sets. In one example, we found a tessellation of 7 modules which generated 94 DS sets. The modules will be used to enable large population DS DOI studies. Here we present one example trace during a 3-back protocol. Examination of the DS traces suggest the expected higher DS Φ sensitivity to cerebral hemodynamics. Further, close observation of the results demonstrate the importance of considering both the Oxy-hemoglobin concentration change (ΔO) and Deoxy-hemoglobin concentration change (ΔD) during such protocols. The results indicated that if one observed only ΔO they would have mis-identified brain activation in the short SD I measurement. Other data-types and ΔD dynamics suggested that the short SD I was dominated by superficial blood-volume instead of the blood-flow dynamics associated with brain activation.

3D reconstruction of the human brain at high resolution is one of the most important challenges of neuroscience. We present a new clearing method named SHORT that in combination with an advanced double-view light-sheet fluorescence microscope and an automated machine-learning based images analysis allow to perform volumetric study of the human brain. We applied our methodology to a Broca’s area block of 4 x 4 x 2 cm³, demonstrating the possibility of obtaining a fast 3D reconstruction of the human brain at high-resolution, paving the way to the possibility of finally mapping a comprehensive atlas of the human brain.

Optical imaging technology serves as a powerful tool in neuroscience study for recording large populations of neurons in vivo. Here we present a compact lensless microscope that breaks the fundamental tradeoff in lens-based imaging systems to simultaneously achieve large field-of-view (FOV) and high-resolution imaging. Our prototype lensless microscope incorporates a “contour” phase mask with an integrated illumination system giving us improved performance for allowing us to demonstrate the first functional imaging with a lensless microscope in behaving non-human primates (NHPs). Specifically, we successfully imaged over a 16 mm² FOV on primary visual cortex of NHPs, and measured how cortical activity changes as a function of the stimulus position. The extracted position tuning information from our lensless microscope has good correspondence to the ground truth captured by a tabletop widefield microscope system.

High-speed low-light two-photon voltage imaging is an emerging tool to simultaneously monitor neuronal activity from a large number of neurons. However, shot noise dominates pixel-wise measurements and the neuronal signals are difficult to be identified in the single-frame raw measurement. We developed a self-supervised deep learning framework for voltage imaging denoising, DeepVID, without the need for any high-SNR measurements. DeepVID infers the underlying fluorescence signal based on independent temporal and spatial statistics of the measurement that is attributable to shot noise. DeepVID achieved a 15-fold improvement in SNR when comparing denoised and raw image data.

We present a new functional near infrared spectroscopy (fNIRs) technique based on dual-comb optical interrogation applied to dispersive media (DC-fNIRS) that can retrieve the frequency response of a living tissue (such as the brain) by parallel sampling of its frequency response in amplitude and phase at specific frequencies. With this information, we can retrieve the impulse response (diffuse-time-of-flight measurements, DTOF) of the medium and extract the absolute optical properties of the tissue and the spatial localization of perturbations for functional analysis with millisecond temporal resolution and noiseless optical gain, increasing the penetration. We have tested these predictions studying a biomimetic phantom with the same optical characteristics as brain tissue confirming the capacities of the DC-fNIRs technique for diffuse media. The system is patent pending PCT/ES2022/070176.

Functional magnetic resonance imaging (fMRI) is a common medical device to diagnose Alzheimer’s disease (AD), but it is not for all subjects due to its cost and other issues. We investigated the potential of functional near-infrared spectroscopy (fNIRS) to separate AD patients from controls as a pre-screening prior to more thorough examination using fMRI. For this purpose, two-channel fNIRS device with 690 nm and 830 nm, sampled at 10 Hz, was placed on the forehead with 3 cm distance between light source and detector to provide resting state fNIRS signals from both sides of pre-frontal cortex. We applied fractional amplitude of physiological fluctuation (fAPF), modified from fractional amplitude of low frequency fluctuation (fALFF), to oxy-, deoxy-, and total-hemoglobin in very low frequency (0.008-0.1 Hz), respiratory (0.1-0.6 Hz), and cardiac (0.6-5 Hz) bands. A t-test at 0.05 significance level was used to evaluate if the fAPF score from AD patients and healthy controls is significantly different. We found that fAPF score of total hemoglobin from both side at cardiac band showed its potential to distinguish AD patients from healthy controls. This finding was in-line with the recent finding that heart failure may co-occur in AD patients with the prevalence of one third of cases.

By chronically implanting optical fibers and measuring each modality over time, the relationship between changes in animal behavior and changes in each modality can be clarified. While there are some challenges to the practical application of optical fiber. First, the video of neural activity obtained by optical fiber is a mixture of changes in fluorescent signals accompanying neural activity and the fiber’s autofluorescence; hence the obtained neural image is first smoothed by a Gaussian function. A minimum value image is created for the smoothed input image, which is used as the background image to compensate for autofluorescence. In addition, the optical fiber inserted into the brain of a mouse shifts its position as the mouse moves, because the measurement is performed under free-running conditions, hence the differential image using the input image is generated in the marker domain as a pre-processing step. Template matching and angle/position correction of the marker are applied to the signal domain in the actual processing section to obtain corrected neural activity data. As a result, the amount of cell migration significantly improved due to the cell migration correction. This algorithm increases the signal-to-noise ratio of the data, enables observation of the same neurons over a long period, such as two hours, and makes detecting neural activity more accurate with the correction. In the future, we plan to correspond with the obtained activity to video images of mouse behavior to be taken separately.

Genetically encoded fluorescent calcium indicators (GECIs) allow for imaging of neuronal calcium activity with high spatiotemporal resolution. However, to date only very few GECIs were developed in the near-infrared (NIR) range of 700 to 900 nm, where optical scattering and attenuation are minimal in tissue. NIR GECIs generally suffer from low brightness and weak fluorescence responses and are thus not deemed as suitable for in vivo imaging.

NIR-GECO2G is a recently developed GECI with excitation and emission maxima at 678 nm and 704 nm, respectively. Using widefield fluorescence imaging in the live the mouse brain, we demonstrate several-fold improved response magnitude compared to the original NIR-GECO1 variant. We further show several-fold increased NIR-GECO2G brightness levels in Blvra-/- mice, where high concentrations of biliverdin (BV) result from deleting the gene for the enzyme that aids in the breakdown of BV.

Our results show that NIR-GECO2G demonstrates strong in vivo responses to stimuli in mice and can be used over multiple experiments on minutes- to hours-long timescales.

The Allen Mouse Brain Connectivity Atlas (AMBCA) offers a high-resolution map of neural connections detailing axonal projections labeled by viral tracers. It is a unique tool for studying structural connectivity and better understanding the white matter pathways of the gene mouse brain. But, the analysis and comparison of these data are limited to a simple visualization on the Allen website and have no direct relationship with specific User data. Here, we propose a series of python-based tools to operate with AMBCA data in the User’s data space. Our method is based on ”back and forth” actions between Allen and User data using the Allen Software Development Kit (AllenSDK) to import data from the Allen Institute and the Python package ANTsPyX for registration. A transformation matrix is calculated with ANTsPyX to overlay, for instance, Allen’s projection density maps with a diffusion MRI-based tractography in the User space. Conversely, applying the inverse transformation to a specific location along a white matter bundle within the User space allows us to recover which experiments were done at this particular location in the Allen Mouse brain Common Coordinate Framework (CCFv3). Thus, both data can be used in a natural interaction, e.g., by inspecting them in a visualization tool such as the MI-Brain software. This series of tools will offer an attractive solution for researchers with neural tracing and/or tractography data to be combined with the AMBCA. The code is available at: https: //github.com/linum-uqam/m2m.

Wide-field calcium imaging (WFCI) with genetically encoded calcium indicators allows spatiotemporal recordings of neuronal activity in preclinical models. When applied to the study of sleep, WFCI data are manually scored into sleep states of wakefulness, non-rapid eye movement (NREM) and REM by use of adjunct electroencephalogram (EEG) and electromyogram (EMG) recordings. However, this process is time-consuming, invasive and suffers from low inter- and intra-rater reliability. To overcome these limitations, an automated sleep state classification method that operates on spatiotemporal WFCI recordings is desired. Previous work that classifies sleep states from WFCI data by use of multiplex visibility graphs and deep learning only leverages shared information derived from average time series across parcellated brain regions, and thus fails to fully explore the spatiotemporal calcium dynamics recorded. In this work, a hybrid network architecture consisting of a convolutional neural network (CNN) to extract spatial features of image frames and a bidirectional long short-term memory network (BiLSTM) with attention mechanism to identify temporal dependencies among different time points was proposed to jointly learn spatial and temporal information from the WFCI sleep data. Nineteen transgenic mice expressing GCaMP6f in excitatory neurons were used for network training and testing. The CNN-BiLSTM achieved a weighted F1-score of 0.84 and Cohen’s κ of 0.64, indicating substantial agreement with EEG/EMG-based human scoring. The gradient-weighted class activation maps were computed to provide deeper insights into the brain regions most relevant to the inference of individual sleep state. This work will enable further investigation of sleep neural activity using WFCI.

The human cerebral cortex is composed of gyri of approximately 10–15 mm in width that have independent functions. To explore their activities with functional near-infrared spectroscopy (fNIRS), hemodynamic responses at single gyri should be measured. Most fNIRS devices only offer a channel arrangement with a larger interval size than the gyrus width, which can cause false negative errors in detecting cortical activation localized within 10–15 mm, and this has been an obstacle using fNIRS to explore cortical activities. Previously, we demonstrated doubling of the channel density using a triangular arrangement of dual-purpose optodes with a minimum number of optodes that was almost equivalent to that used in conventional arrangements. To implement this method as a wearable device for human measurement, we developed a dual-purpose optode to function both as the source and detector with the base unit triangularly mounted by three optodes, and the connectors joining plural base units with three-way joints. Optodes in this triangular arrangement illuminated and detected in sequence between adjacent optodes and performed high-density 15-mm measurements in channel intervals. Measurements of 30 channels on an adult human successfully detected hemodynamic responses to unilateral finger movements at the motor-related cortical regions according to their functions.

Dementia afflicts more than 55 million patients worldwide, and Alzheimer’s disease (AD) accounts for around 60% to 80% of total cases¹. During the onset of AD, the hippocampus (HC) is among the first-affected brain regions to experience pathological changes. Therefore, identifying changes to HC in AD subjects will be extremely helpful in providing early-stage diagnosis and interventions. We are developing in vivo techniques to investigate microscopic alterations to hippocampal structure and function in animal models of AD. Imaging will be performed using removable GRIN (gradient-index) lenses to chronically access subcortical brain structures with two-photon microscopy. Here, we report the development and characterization of a customized cannula (1 mm diameter, 6 – 7 mm length) to repeatedly insert a GRIN lens for two-photon imaging. The cannula allows for easy removal of the lens after imaging sessions and enables detailed investigations of hippocampal changes during AD progression in mouse models. The cannula is made of polyimide tubing and tipped with transparent acrylic coverslip. We compare acrylic coverslips and existing glass coverslips in terms of physical and optical properties. Acrylic coverslips display comparable imaging quality and therefore serves as a reliable alternative to glass coverslips which is more economical, reproducible, and mechanically stable. We also present preliminary hippocampal images collected in vivo with our custom cannula. These results will guide more extensive efforts to measure hippocampal metabolic and hemodynamic alterations in awake animal models of AD.