Vivian Pera

A Computer Vision Approach to Rare Cell In Vivo Fluorescence Flow Cytometry

Noninvasive enumeration of rare circulating cell populations in small animals is of great importance in many areas of biomedical research. In this work, we describe a macroscopic fluorescence imaging system and automated computer vision algorithm that allows in vivo detection, enumeration and tracking of circulating fluorescently‐labeled cells from multiple large blood vessels in the ear of a mouse. This imaging system uses a 660 nm laser and a high sensitivity electron‐multiplied charge coupled device camera (EMCCD) to acquire fluorescence image sequences from relatively large (∼5 × 5 mm2) imaging areas. The primary technical challenge was developing an automated method for identifying and tracking rare cell events in image sequences with substantial autofluorescence and noise content. To achieve this, we developed a two‐step image analysis algorithm that first identified cell candidates in individual frames, and then merged cell candidates into tracks by dynamic analysis of image sequences. The second step was critical since it allowed rejection of >97% of false positive cell counts. Overall, our computer vision IVFC (CV‐IVFC) approach allows single‐cell detection sensitivity at estimated concentrations of 20 cells/mL of peripheral blood. In addition to simple enumeration, the technique recovers the cells trajectory, which in the future could be used to automatically identify, for example, in vivo homing and docking events.

Wearable near-infrared optical probe for continuous monitoring during breast cancer neoadjuvant chemotherapy infusions

Abstract. We present a new continuous-wave wearable diffuse optical probe aimed at investigating the hemodynamic response of locally advanced breast cancer patients during neoadjuvant chemotherapy infusions. The system consists of a flexible printed circuit board that supports an array of six dual wavelength surface-mount LED and photodiode pairs. The probe is encased in a soft silicone housing that conforms to natural breast shape. Probe performance was evaluated using tissue-simulating phantoms and in vivo normal volunteer measurements. High SNR (71 dB), low source-detector crosstalk (−60  dB), high measurement precision (0.17%), and good thermal stability (0.22% Vrms/°C) were achieved in phantom studies. A cuff occlusion experiment was performed on the forearm of a healthy volunteer to demonstrate the ability to track rapid hemodynamic changes. Proof-of-principle normal volunteer measurements were taken to demonstrate the ability to collect continuous in vivo breast measurements. This wearable probe is a first of its kind tool to explore prognostic hemodynamic changes during chemotherapy in breast cancer patients.

On the use of the Cramér–Rao lower bound for diffuse optical imaging system design

Abstract. We evaluated the potential of the Cramér–Rao lower bound (CRLB) to serve as a design metric for diffuse optical imaging systems. The CRLB defines the best achievable precision of any estimator for a given data model; it is often used in the statistical signal processing community for feasibility studies and system design. Computing the CRLB requires inverting the Fisher information matrix (FIM), however, which is usually ill-conditioned (and often underdetermined) in the case of diffuse optical tomography (DOT). We regularized the FIM by assuming that the inhomogeneity to be imaged was a point target and assessed the ability of point-target CRLBs to predict system performance in a typical DOT setting in silico. Our reconstructions, obtained with a common iterative algebraic technique, revealed that these bounds are not good predictors of imaging performance across different system configurations, even in a relative sense. This study demonstrates that agreement between the trends predicted by the CRLBs and imaging performance obtained with reconstruction algorithms that rely on a different regularization approach cannot be assumed a priori. Moreover, it underscores the importance of taking into account the intended regularization method when attempting to optimize source–detector configurations.

Diffuse fluorescence fiber probe for in vivo detection of circulating cells

Abstract. There has been significant recent interest in the development of technologies for enumeration of rare circulating cells directly in the bloodstream in many areas of research, for example, in small animal models of circulating tumor cell dissemination during cancer metastasis. We describe a fiber-based optical probe that allows fluorescence detection of labeled circulating cells in vivo in a diffuse reflectance configuration. We validated this probe in a tissue-mimicking flow phantom model in vitro and in nude mice injected with fluorescently labeled multiple myeloma cells in vivo. Compared to our previous work, this design yields an improvement in detection signal-to-noise ratio of 10 dB, virtually eliminates problematic motion artifacts due to mouse breathing, and potentially allows operation in larger animals and limbs.

Long-term longitudinal monitoring of chemotherapy response using Spatial Frequency Domain Imaging using an improved two-layer Monte Carlo based inverse model (Conference Presentation)

Spatial Frequency Domain Imaging (SFDI) is a Diffuse Optical Imaging (DOI) technique that is well suited for preclinical functional imaging. Recently, we have shown that SFDI can successfully be used for longitudinal monitoring of a prostate subcutaneous tumor xenograft, where we have applied a look-up-table (LUT) based approach to extract tissue absorption (μa) and scattering properties (μs’). This LUT assumes a semi-infinite homogeneous medium and simulates reflectance (Rd) in spatial domain, and scales Rd for all μa and μs’ of interest from a single Monte Carlo simulation. However, converting Rd to spatial frequency domain (SFD) and scaling for μs’ may introduces unacceptable errors. Most importantly, the homogeneous model fails to mimic the actual physiology of a subcutaneous tumor, which can be described as a two-layer medium with a thin skin layer above the tumor layer. To overcome these limitations, we have developed a Monte Carlo based two-layer LUT with a wide range of tumor (bottom) layer optical properties, and fixed skin (top) properties. The two-layer LUT will be validated by two-layer silicone phantoms and tested for sensitivity to inaccurate layer assumptions. Additionally, the homogeneous and two-layer LUTs will be used on a large mouse tumor database (n=54 mice monitored over 3 months) to identify how the two-layer LUT can improve accuracy of SFDI by more accurately reflecting in vivo physiology, and reducing discretization and scaling errors. Improved SFDI findings in small animals, in the long run, will help establish clinical DOI tools for early detection of chemotherapy efficacy during treatment.

Near Infrared Spectroscopy for Measuring Changes in Bone Hemoglobin Content after Exercise in Individuals with Spinal Cord Injury

Bone blood perfusion has an essential role in maintaining a healthy bone. However, current methods for measuring bone blood perfusion are expensive and highly invasive. This study presents a custom built near‐infrared spectroscopy (NIRS) instrument to measure changes in bone blood perfusion. We demonstrated the efficacy of this device by monitoring oxygenated and deoxygenated hemoglobin changes in the human tibia during and after exercise in able‐bodied and in individuals with spinal cord injury (SCI), a population with known impaired peripheral blood perfusion. Nine able‐bodied individuals and six volunteers with SCI performed a 10 min rowing exercise (functional electrical stimulation rowing for those with SCI). With exercise, during rowing, able‐bodied showed an increase in deoxygenated hemoglobin in the tibia. Post rowing, able‐bodied showed an increase in total blood content, characterized by an increase in total hemoglobin content due primarily to an increase in deoxygenated hemoglobin. During rowing and post‐rowing, those with SCI showed no change in total blood content in the tibia. The current study demonstrates that NIRS can non‐invasively detect changes in hemoglobin concentration in the tibia.

Multiplexed fluorescence mediated tomography with temporal and spectral data

Abstract. We recently developed an algorithm for multiplexed fluorescence tomographic imaging of at least four fluorophores concurrently in the red and near-infrared wavelength region by jointly using spectral and temporal data. We report the design of a fluorescence tomography instrument that acquires spectral and temporal data, and validate its use in tissue-mimicking phantoms with four embedded fluorescent targets with highly overlapped spectral signatures. Critically, this requires measurement or computation of extended fluorophore signature libraries, which capture the variability in the measured signal due to the unknown position of the targets in the media. We demonstrate that we can demix and tomographically image all four fluorophores with zero image cross-talk, and 1 mm or better spatial resolution.

A sparse nonnegative demixing algorithm with intrinsic regularization for multiplexed fluorescence tomography

Fluorescence molecular tomography is becoming an important tool in preclinical biomedical imaging of small animals. However, the inability to perform high-throughput imaging of multiple fluorescent targets in bulk tissue remains a limitation. Recent work in our group suggests that joint measurement of spectral and temporal fluorophore data can enable robust identification (“demixing”) and localization of at least four concurrent fluorophores. Here we present a novel demixing strategy for this data, which incorporates ideas from sparse subspace clustering and compressed sensing. It uses a suitable “library” of fluorophore signatures to compute a nonnegative least-squares estimate of each fluorophore signal in the sample. The algorithm does not require a regularization parameter, even when the library is rank-deficient. In simulations, we simultaneously demixed four fluorophores with closely overlapping spectral and temporal profiles in a 25 mm diameter cross-sectional area with an RMS error of less than 3% per fluorophore.

Multi-Wavelength Fluorescence Diffuse Fluorescence Flow Cytometry for Detection and Enumeration of Very Rare Circulating Cells In Vivo

We describe a new instrument for fluorescence detection and enumeration of very rare circulating cells with diffuse light in mice and validate its operation at cell concentrations less than 5e3 / mL in vivo.