KEYWORDS: Monte Carlo methods, Structured light, Photon transport, Computer simulations, Diffuse optical tomography, Signal to noise ratio, Optical tomography, Tomography, Data acquisition, Whole body imaging
With the rapid development of spatial light modulators, structured light strategies have been readily implemented for efficient diffuse optical tomography (DOT) and fluorescent molecular tomography (FMT) applications. Compared to traditional pencil-beam sources, wide-field illumination enables larger imaging field-of-view (FOV), higher signal-to-noise ratios (SNR), and faster data acquisition speed, making it attractive for small-animal whole-body imaging.
As the gold-standard for simulating photon propagations inside complex biological tissues, the Monte Carlo (MC) method is one of the most accurate approaches for imaging general media. Recently, thanks to parallel hardware such as graphics processing units (GPUs), MC simulations can be computed with high efficiency even with personal computers. While we have added support for wide-field illumination patterns in our widely distributed MC platform (http://mcx.space), the computation for multiple patterns is currently performed sequentially.
To further accelerate forward modeling of large number of wide-field patterns, we propose a new method, referred to as “photon sharing”, to simultaneously simulate multiple structure-light sources. We demonstrate a 5- to 10-fold reduction of the MC simulation time. This technique is particularly valuable in DOT or FMT applications using structured light illumination and/or single-pixel-camera based systems. The proposed algorithm has been implemented in our open-source MC simulation platforms, supporting both CPUs and GPUs.
Over the past three decades, non-invasive molecular imaging via optical tomography has garnered attention in the field of preclinical imaging thanks to its high sensitivity and ability to image multiple biomarkers simultaneously. However, it is still very challenging to image intact tissues with high resolution while retaining the two aforementioned characteristics.
Over the last few years, our group has pioneered Mesoscopic Fluorescence Molecular Tomography (MFMT), a novel imaging modality that recapitulates the 3D distribution of fluorescent markers within thick and diffuse samples (< 3 mm) with spatial resolution ~100 µm. Still, as a diffuse optical inverse problem, the image formation can be challenging due to its ill-conditioned nature. Herein, we report on the fusion of MFMT with Optical Coherence Tomography (OCT) to provide both structural and molecular imaging capabilities. Moreover, we leverage the OCT information to impart structural priors that facilitate the optical inverse problem in MFMT. We demonstrate the capability and utility of this novel platform on bioprinted tissues, fluorescent polymer letters in agar phantoms, and on microfabricated beads at different imaging depths.
Fluorescence lifetime imaging (FLI) has become an invaluable tool in the biomedical field by providing unique, quantitative information about biochemical events and interactions taking place within specimens of interest. Applications of FLI range from superresolution microscopy to whole body imaging using visible and near-infrared fluorophores. However, quantifying lifetime can still be a challenging task especially in the case of bi-exponential applications. In such cases, model based iterative fitting is typically employed but necessitate setting up multiple parameters ad hoc and can be computationally expensive. These facts have limited the universal appeal of the technique and methodologies can be specific to certain applications/technology or laboratory bound. Herein, we propose a novel approach based on Deep Learning (DL) to quantify bi-lifetime Forster Resonance Energy Transfer (FRET). Our deep neural network outputs three images consisting of both lifetimes and fractional amplitude. The network is trained using synthetic data and then validated on experimental FLI microscopic (FLIM) and macroscopic data sets (MFLI). Our results demonstrate that DL is well suited to quantify wide-field bi-exponential fluorescence lifetime accurately and in real time, even when it is difficult to obtain large scale experimental training data.
Single-pixel imaging based on compressive sensing theory has been a highlighted technique in the biomedical imaging field for many years. This interest has been driven by the possibility of performing microscopic or macroscopic imaging based on low-cost detector arrays, increased SNR (signal-to-noise ratio) in the acquired data sets and the ability to perform high quality image reconstruction with compressed data sets by exploiting signal sparsity. In this work, we present our recent work in implementing this technique to perform time domain fluorescence-labeled investigations in preclinical settings. More precisely, we report on our time-resolved hyperspectral single-pixel camera for fast, wide-field mapping of molecular labels and lifetime-based quantification. The hyperspectral single-pixel camera implements a DMD (Digital micro-mirror device) to generate optical masks for modulating the illumination field before it is delivered onto the sample and focuses the emission light signals into a multi-anode hyperspectral time-resolved PMT (Photomultiplier tube) to acquire spatial, temporal and spectral information enriched 4-D data sets. Fluorescence dyes with lifetime and spectral contrast are embedded in well plates and thin tissues. L-1 norm based regularization or the least square method, is applied to solve the underdetermined inverse problem during image reconstruction. These experimental results prove the possibility of fast, wide-field mapping of fluorescent labels with lifetime and spectral contrast in thin media.
KEYWORDS: Neon, Kidney, Luminescence, Tomography, Fluorescence tomography, Sensors, Optical tomography, Monte Carlo methods, Signal to noise ratio, Multiplexing
Fluorescence Molecular Tomography (FMT) is a powerful optical imaging tool for preclinical research. Especially, its implementation with time-domain (TD) techniques allows lifetime multiplexing for simultaneously imaging multiple biomarkers and provides enhanced data sets for improved resolution and quantification compared to continuous wave (CW) and frequency domain (FD) methodologies. When performing time-domain reconstructions, one important aspect is the selection of a temporal sub-data set. Typically, such selection is performed a posteriori after dense temporal sampling during the acquisition. In this work, we investigate the potential to collect a priori sparse data sets for fast experimental acquisition without compromising FMT performances.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.