Search

Changes in 3D pathophysiological images for healthy finger joints during cuff occlusion are presented, recovered using multispectral diffuse optical tomography, to evaluate its potential use for early detection of rheumatoid arthritis.

In the context of traumatic brain injury, two NIRS systems (NIRO and ISS) were evaluated during induced hypoxia in healthy participants to demonstrate their limits in classifying normal and hypoxic brains for absolute oxygenation measurements

Within 72 hours of stroke onset, High-Density Diffuse Optical Tomography can detect disruptions in functional connectivity patterns that significantly differ from healthy subjects (p

High density near infrared probes and registered subject specific head models are used to show the potential improvements to the quantitative accuracy of recovered parameters relevant to monitoring an injured brain.

A multi-modal optical imaging system is presented which uses diffuse optical tomography to determine subject-specific, heterogeneous tissue attenuation to increase the quantitative accuracy of bioluminescence tomography.

Bioluminescence imaging (BLI) is a widely used pre-clinical imaging technique, but there are a number of limitations to its quantitative accuracy. This work uses an animal model to demonstrate some significant limitations of BLI and presents processing methods and algorithms which overcome these limitations, increasing the quantitative accuracy of the technique. The position of the imaging subject and source depth are both shown to affect the measured luminescence intensity. Free Space Modelling is used to eliminate the systematic error due to the camera/subject geometry, removing the dependence of luminescence intensity on animal position. Bioluminescence tomography (BLT) is then used to provide additional information about the depth and intensity of the source. A substantial limitation in the number of sources identified using BLI is also presented. It is shown that when a given source is at a significant depth, it can appear as multiple sources when imaged using BLI, while the use of BLT recovers the true number of sources present.

Spectral bioluminescence imaging and prior knowledge of bioluminescence shape were used to accurately localise, and reconstruct two luminescent sources in a phantom mouse via Markov Chain Monte Carlo sampling, and estimate reconstruction reliability.

Atlas-based DOT consists of three steps: registration,sensitivity matrix calculation and activation recovery. Registration methods provide an accuracy of ~7 mm, leading to ~150% inaccuracy in sensitivity, with little detriment in activation recovery.

Realistic head models are used to assess the information content in regions of the brain relevant to monitoring traumatic brain injury patients with commercial NIRS probes, highlighting system variability and the need for better standardization.

Computational and image recovery tools for Near Infrared spectroscopic imaging have undergone major development in the last 10 years. Some of the key significant advancements will be presented together with clinical and preclinical examples.

We present integrative advances in high-density diffuse optical tomography imaging arrays, large field-of-view instrumentation, and anatomical head modeling. The system was tested by imaging four language tasks and multiple resting-state networks.

Free-space modelling is important for quantitative bioluminescence imaging. Data using a new imaging system and a commercial system are compared to highlight this important phenomena showing inaccuracies of ~50% when free-space models are ignored.

We present a novel near-infrared non-contact multispectral optical imaging system by utilizing a CCD camera and an accurate 3D shape recovery method to study tomographic information of the rheumatoid arthritic and normal hand joints

Bioluminescence Tomography characteristics suggest its suitability to Compressive Sensing-based reconstruction. Numerical simulations examining reconstruction quality in the presence of noise demonstrate improved robustness and compactness.

An automated, multi-view, spectral bioluminescence tomography system that utilises structured-light-based surface capture techniques along with a model-based approach to image reconstruction is presented.

Image-quality of high-density diffuse optical tomography is evaluated against fMRI using functional maps of visual cortex as a benchmark. Co-registered subject-specific light models have an average localization error of 4.4 +/- 1 mm.

Employing multiple time-bins in the image reconstruction of a time-domain fluorescence tomography dataset provides improved spatial resolution and contrast. Furthermore, these gains can be used to minimize tissue sampling and reduce imaging duration

Using fluorophore specific to zinc, a well-established prostate cancer marker, to detect prostate cancer will be challenged by the “reverse-uptake” of the fluorophore. A sensitivity-adapted reconstruction method may improve the target recovery in axial-imaging geometry.

A surface-shrinking algorithm is demonstrated for automated source position localization one scattering depth into discretized simulation domains, in near-infrared imaging. The algorithm allows users to accurately place source fiber locations with minimal guidance.

This work demonstrates the benefits of using subject specific models for image reconstruction in neuro-imaging of humans. It also investigates depth related information available from the increased number of tomographic measurements.

Optimized bands of wavelengths in spectral optical imaging is presented showing improvement in cross talk between parameters. A normalization technique is presented which creates a more uniform update within a spectral image reconstruction model.

For diffuse optical tomography reconstruction, DC-based method outperforms frequency-domain method in background artifacts, at the known cost of increased coupling between absorption and scattering. The differences of these methods diminish when spatial priors are available.

The use of higher order approximations to the Radiative transport equation, through simplified spherical harmonics expansion (SPN) in optical tomography are presented. At higher orders, the image reconstruction becomes highly under-determined due to the large increase in the number of unknowns which cannot be adequately recovered. However, reconstruction of diffuse parameters, namely optical absorption and reduced scatter have shown to be more accurate where only the sensitivity matrix used in the inverse problem is based on SPN method and image reconstruction is limited to these two diffuse parameters.

Model based image reconstruction in Optical and Molecular imaging is rapidly becoming an optimal method due to the flexibility and robustness of this method to problem specific cases. An online imaging and Finite Element modeling tool for optical imaging has been developed which allows the user to run problem specific cases, as well as providing an online tutorial for light propagation in tissue. This online package, based on NIRFAST is presented and is being made available to users worldwide to allow further development of this novel imaging technique.

3D MRg-NIRS imaging of breast cancer has been studied in simulation and in vivo using parallel plate and circular geometries. Optimal methods to minimize errors in measurement and modeling are presented.

Protoporphyrin IX fluorescence activity in mouse model gliomas is imaged in nude mice using an MRI-coupled spectroscopy scanner. Segmented MR images acquired simultaneously with fluorescence spectra are used to guide fluorescence yield reconstruction.

Spectrally-resolved BioLuminescence Tomography is used to image Luciferase activity within a 3-D model. The Adjoint theorem for 3-D image reconstruction is presented and validated using experimental data with a complete computation time of 5 minutes.

Using a Jacobian reconstruction method using experimental data, we show a dramatic increase in image reconstruction time in 3-D DOT, without detriment to image quality.

The SPnapproximation is implemented into NIRFAST. Results are presented on a small geometry and are shown for SP1, SP3and SP5. Further work is underway to test the accuracy using Monte Carlo data.

A method is developed to show that instead of using data from the entire spectrum, only selected spectral bands are required to improve image reconstruction accuracy in spectral diffuse optical tomography.

The reconstruction of sagittal trans-rectal optical tomography for prostate imaging is presented with assumption of structural guidance from trans-rectal ultrasound. The spatial prior combined with Jacobian weighing improves the recovery of lesion depth.

We have developed a high-density neuroimaging DOT system with improved depth discrimination and lateral resolution. The advantages are demonstrated through cortical maps of phase-encoded traveling waves in the full visual field of adult humans.

Diffuse optical tomography of tissue can be used to characterize molecular features of tissue during detection, diagnosis, or management of cancer therapy. This tutorial will review the process and current applications in the field.

Developing diffuse optical tomography methods for neuroimaging of humans is challenging due geometry and light level constraints. Analysis of multi-distant high-density imaging arrays show feasibility of imaging up to 20-mm depth within the adult brain.

Near-Infrared (NIR) Diffuse Optical Tomography (DOT) is a non-invasive imaging technique which is used to obtain functional and physiological images of soft tissue, such as the female breast, specifically for the detection and characterization of breast cancer. The vast majority of the work to date has been limited to two dimensional (2D) models which have provided valuable insight into tissue function and physiology enabling a better understanding of tumor development and treatment. Although the 2D image reconstruction approach is fast and computationally efficient, it has limitations as it does not correctly represent the volume under investigation and therefore do not provide the most accurate model for image reconstruction. Three dimensional (3D) modeling and image reconstruction is becoming more accessible through the development of sophisticated numerical models and computationally fast algorithms. A robust and general method is presented which reconstructs 3D functional images using a more accurate and realistic spectral model of 3D light propagation in tissue. Results from a single patient example are presented to demonstrate the clinical importance of 3D image reconstruction in optical tomography for the detection and characterization of breast cancer.

Spectrally-constrained diffuse tomography is used to image endogenous and exogenous tissue chromophore concentrations with higher accuracy than when using individual wavelengths. Reconstructed images from simulated data demonstrate that concentrations of Lutetium Texaphyrin are readily recovered.

The algorithms for near-infrared (NIR) diffuse optical tomography within magnetic resonance imaging (MRI) of the female breast are analyzed in both circular and curved slab geometries.

Spectrally-resolved BioLuminescence Tomography is used to image Luciferase activity within a 3D mouse model using multi-wavelength data from multiple bioluminescence sources. Images show the sources reconstructed accurately using a fast, direct and linear 3D algorithm.

A fluorescence diffuse optical tomography system capable of producing B-scan-type images of localized fluorescence regions is presented. The B-Scan mode is analogous to ultrasound where the excitation and remission signals are delivered from the tissue surface.

A novel scheme for implementing spectral constraints into a three-dimensional image reconstruction model is presented. This approach, based on selecting measurements relating to spatial areas of maximum heterogeneity, makes spectral imaging computationally efficient in 3-D.

Accuracy of chromophore concentrations in near-infrared tomography is affected by negative bias in the recovery of absorption coefficients and availability of limited wavelengths. The effect of these factors has been studied with suitable error models.

Near-infrared spectral tomographic imaging was used to show that a statistically significant difference exists in the spectral scattering power and amplitude between a group of benign (n=7) and malignant (n=6) tumor tissues imaged in vivo.

8-Channel broadband spectroscopy and tomography CCD system was designed and frequency domain reconstruction was applied to get both absorption and scattering images at two wavelengths based on the second-differential spectroscopy analysis of CW data.

In NIR tomography of the breast, good contact is needed by the fibers, resulting in breast deformation. We present a deformation model to account for the change of shape and discuss implications in image reconstruction.

A multi-spectral, frequency-domain near infrared tomography system has been constructed and evaluated. Measurements of light transmission through female breast can be acquired simultaneously with MRI scans.

Near infrared (NIR) image reconstruction can be improved using a-priori structural knowledge provided by MRI. In this work we show that combining these modalities and optimizing the NIR reconstruction algorithm improves recovery of internal optical property distributions.

Oxygen saturation imaging has been calibrated relative to the measured pO2 values in tissue phantoms of Intralipid and blood. These calibration results are used to interpret the oxygen saturation of patient tumors and normal tissues.

Simulations of near-infrared light propagation based upon MRI images of a rat cranium are used to determine the optimum arrangement for maximum sensitivity in the brain. Singular value decomposition analysis is used to provide a quantitative measure for amount of information obtained in different fiber arrangements.