Session A1: Focus Session: Novel Instrumentation and measurements for Biomedical Research

Instrumentation of Molecular Imaging on Site-Specific Targeting Fluorescent Peptide for Early Detection of Breast Cancer

In this work we developed two biomedical imaging techniques for early detection of breast cancer. Both image modalities provide molecular imaging capability to probe site-specific targeting dyes. The first technique, heterodyne CCD fluorescence mediated tomography, is a non-invasive biomedical imaging that uses fluorescent photons from the targeted dye on the tumor cells inside human breast tissue. The technique detects a large volume of tissue (20 cm) with a moderate resolution (1 mm) and provides the high sensitivity. The second technique, dual-band spectral-domain optical coherence tomography, is a high-resolution tissue imaging modality. It uses a low coherence interferometer to detect coherent photons hidden in the incoherent background. Due to the coherence detection, a high resolution (20 microns) is possible. We have finished prototype imaging systems for the development of both image modalities and performed imaging experiments on tumor tissues. The spectroscopic/tomographic images show contrasts of dense tumor tissues and tumor necrotic regions. In order to correlate the findings from our results, a diffusion-weighted magnetic resonance imaging (MRI) of the tumors was performed using a small animal 7-Telsa MRI and demonstrated excellent agreement.   

Digital Image Speckle Correlation for the Quantification of the Cosmetic Treatment with Botulinum Toxin Type A (BTX-A)

The skin on the face is directly attached to the underlying muscles. Here, we successfully introduce a non-invasive, non-contact technique, Digital Image Speckle Correlation (DISC), to measure the precise magnitude and duration of facial muscle paralysis inflicted by BTX-A. Subjective evaluation by clinicians and patients fail to objectively quantify the direct effect and duration of BTX-A on the facial musculature. By using DISC, we can (a) Directly measure deformation field of the facial skin and determine the locus of facial muscular tension(b)Quantify and monitor muscular paralysis and subsequent re-innervation following injection; (c)~Continuously correlate the appearance of wrinkles and muscular tension. Two sequential photographs of slight facial motion (frowning, raising eyebrows) are taken. DISC processes the images to produce a vector map of muscular displacement from which spatially resolved information is obtained regarding facial tension. DISC can track the ability of different muscle groups to contract and can be used to predict the site of injection, quantify muscle paralysis and the rate of recovery following BOTOX injection.   

Ultra-low field SQUID electron paramagnetic resonance for biomedical applications

We have constructed a SQUID-magnetometer system operating at 4K, for electron paramagnetic resonance (EPR) detection from room temperature samples in magnetic fields of the order of Earth$^\prime$s field. The magnetometer consists of a home-built gradiometer pick-up coil inductively coupled to the input of a commercially available 2-stage dc SQUID amplifier. Operation at low EPR excitation frequency of a few MHz has advantages of negligible sample heating and finite penetration depth effects in biological systems. The same system can be adapted to detect NMR signals at several kHz range. We describe our detection scheme and discuss the prospects for \textit{in vivo} biomedical EPR imaging.   

Single cell microfluidics for systems oncology

The singular term ``cancer'' is never one kind of disease, but deceivingly encompasses a large number of heterogeneous disease states, which makes it impossible to completely treat cancer using a generic approach. Rather systems approaches are urgently required to assess cancer heterogeneity, stratify patients and enable the most effective, individualized treatment. The heterogeneity of tumors at the single cell level is reflected by the hierarchical complexity of the tumor microenvironment. To identify all the cellular components, including both tumor and infiltrating immune cells, and to delineate the associated cell-to-cell signaling network that dictates tumor initiation, progression and metastasis, we developed a single cell microfluidics chip that can analyze a panel of proteins that are potentially associated inter-cellular signaling network in tumor microenvironment from hundreds of single cells in parallel. This platform integrates two advanced technologies -- microfluidic single cell handling and ultra-high density protein array. This device was first tested for highly multiplexed profiling of secreted proteins including tumor-immune signaling molecules from monocytic leukemia cells. We observed profound cellular heterogeneity with all functional phenotypes quantitatively identified. Correlation analysis further indicated the existence of an intercellular cytokine network in which TNF$\alpha $-induced secondary signaling cascades further increased functional cellular diversity. It was also exploited to evaluate polyfunctionality of tumor antigen-specific T cells from melanoma patients being treated with adoptive T cell transfer immunotherapy. This platform could be further extended to analyze both solid tumor cells (e.g. human lung carcinoma cells) and infiltrating immune cells (e.g. macrophages) so as to enable systems analysis of the complex tumor microenvironment from small amounts of clinical specimens, e.g. skinny needle biopsies. Thus, it could potentially become a clinical tool for patient stratification based upon the inter-cellular signaling network and designing new anti-cancer therapy by targeting microenvironmental components.   

High resolution phase-sensitive optical coherence microscopy tracking of magnetic microbeads for cellular mechanics

We present a real-time multi-modal near-infrared imaging technology that tracks externally induced axial motion of magnetic microbeads in mouse macrophages, human breast epithelial cells, and human breast primary ductal carcinoma cells. The imaging technique includes nanometer scale phase-sensitive magnetomotive integrated optical coherence (MM-OCM) and multiphoton (MPM) microscopy. MPM is used to visualize multifunctional microbeads based on their fluorescence, while MM-OCM detects sinusoidal displacements of the magnetic microbeads induced by a magnetic field. This methodology demonstrates imaging of cell dynamics and has the potential for measuring cell biomechanical properties, which are important in assessing cell health.   

Reagent-free ultrasensitive spectroscopic probes for long term diabetes monitoring

Long-term glycemic control is essential in developing therapeutics for diabetics. Glycated hemoglobin (HbA1c) and glycated albumin have been increasingly accepted as a functional metric of glycemic control over the past two to three months and three weeks, respectively. In this talk, we present the first demonstration of non-enhanced Raman spectroscopy as a novel analytical method for quantitative detection of HbA1c and glycated albumin. Using the drop coating deposition Raman technique, we observe that the non-enzymatic glycosylation of these proteins results in subtle, but consistent, changes in vibrational features, which with the help of multivariate classification techniques can be used to discriminate the glycated proteins from their unglycated variants with 100\%. Additionally, the developed multivariate calibration models show a high degree of prediction accuracy even at substantially lower concentrations than those typically encountered in clinical practice. The excellent accuracy and reproducibility achieved in this proof-of-concept study opens substantive avenues for basic investigations of glycated proteins as well as in high-throughput glycemic marker sensing in multi-component mixtures and potentially even in serum and whole blood samples.   

The measurand problem in infrared breath alcohol testing

Measurements are made to determine the value of a quantity known as a measurand. The measurand is not always the quantity subject to measurement, however. Often, a distinct quantity will be measured and related to the measurand through a measurement function. When the identities of the measurand and the quantity actually measured are not well defined or distinguished, it can lead to the misinterpretation of results. This is referred to as the measurand problem. The measurand problem can present significant difficulties when the law and not science determines the measurand. This arises when the law requires that a particular quantity be measured. Legal definitions are seldom as rigorous or complete as those utilized in science. Thus, legally defined measurands often fall prey to the measurand problem. An example is the measurement of breath alcohol concentration by infrared spectroscopy. All 50 states authorize such tests but the measurand differs by jurisdiction. This leads to misinterpretation of results in both the forensic and legal communities due to the measurand problem with the consequence that the innocent are convicted and guilty set free. Correct interpretation of breath test results requires that the measurand be properly understood and accounted for. I set forth the varying measurands defined by law, the impact these differing measurands have on the interpretation of breath test results and how the measurand problem can be avoided in the measurement of breath alcohol concentration.   

Nanoscale neuroelectronic interface based on open-ended nanocoax arrays

We describe the development of a nanoscale neuroelectronic array with submicron pixelation for recording and stimulation with high spatial resolution. The device is composed of an array of nanoscale coaxial electrodes, either network- or individually-configured. As a neuroelectronic interface, it will employ noninvasive real-time capacitive coupling to the plasma membrane with potential for extracellular recording of intra- and interneural synaptic activity, with one target being precision measurement of electrical signals associated with induced and spontaneous synapse firing in pre- and post-synaptic somata. Subarrays or even individual pixels can also be actuated for precisely-localized stimulation. We report initial results from measurements using the rat adrenal pheochromocytoma PC12 cell line, which terminally differentiates in response to nerve growth factor, as well as SH-SY5Y neuroblastoma cells in response to retinoic acid, characterizing the basic performance of the fabricated device.   

Phosphorus-31 MRI of bones using quadratic echo line-narrowing

There is a great need to probe the internal composition of bone on the sub-0.1 mm length scale, both to study normal features and to look for signs of disease. Despite the obvious importance of the mineral fraction to the biomechanical properties of skeletal tissue, few non-destructive techniques are available to evaluate changes in its chemical structure and functional microarchitecture on the interior of bones. MRI would be an excellent candidate, but bone is a particularly challenging tissue to study given the relatively low water density and wider linewidths of its solid components. Recent fundamental research in quantum computing gave rise to a new NMR pulse sequence - the quadratic echo - that can be used to narrow the broad NMR spectrum of solids. This offers a new route to do high spatial resolution, 3D $^{31}$P MRI of bone that complements conventional MRI and x-ray based techniques to study bone physiology and structure. We have used our pulse sequence to do 3D $^{31}$P MRI of \textit{ex vivo} bones with a spatial resolution of (sub-450 $\mu$m)$^3$, limited only by the specifications of a conventional 4 Tesla liquid-state MRI system. We will describe our plans to push this technique towards the factor of 1000 increase in spatial resolution imposed by fundamental limits.   

Phosphorus-31 MRI of cell membranes using quadratic echo line-narrowing

Soft biological tissues have phosphorus concentrated in the membranes, metabolites, RNA and DNA of cells. This leads to a complicated, multi-peak $^{31}$P nuclear magnetic resonance spectrum (including a broad membrane peak and narrow metabolite peaks), which precludes high-resolution $^{31}$P MRI of soft tissues. This long-standing barrier has been overcome by a novel pulse sequence - the quadratic echo - recently discovered in fundamental quantum computation research. Applying time-dependent gradients in synch with a repeating pulse block enables a new route to high spatial resolution, three-dimensional $^{31}$P MRI of the soft solid components of cells and tissues. This is a functionally different kind of MR image, since conventional $^1$H MRI probes the intracellular and extracellular free water, whereas our $^{31}$P MRI signal is dominated by the cell membrane contribution, which in turn depends on the density of mitochondria. The unique aspects of the signal should provide new insights into cellular and tissue function that compliment the information revealed by $^1$H MRI. So far, various \textit{ex vivo} soft tissue samples have been imaged with (sub-mm)$^{3}$ voxels. We will describe plans to enhance the spatial resolution in future work, to open a new window into cells.   

Accelerating the magnetic resonance imaging of hard and soft solids

Recent fundamental research in quantum computing gave rise to a new NMR pulse sequence - the quadratic echo - that can be used to narrow the broad MR spectrum of solids by orders of magnitude. This advance enables high spatial resolution, 3D MRI of hard and soft solids (e.g., the $^{31}$P MRI of bone and soft tissues, which has recently been demonstrated in our group). While this is great progress, the next challenge to address is the slow imaging time in solids. The $T_1$ time of solids can be orders of magnitude larger than that of a liquid (e.g., for some of our samples the $T_1$ time can be $>$100 s). Since imaging sequences wait a repetition time (of order $T_1$) between scans, a three-dimensional image (requiring many scans) can take $>$24 hours. Here we discuss approaches under study that aim to decrease the total imaging time of our MRI of solids technique. One method is to implement a variant on the driven equilibrium approach. Another method is to take less data, namely, sparse MRI (i.e., under-sampling of \textbf{k}-space). This leads to artifacts in standard FFT image construction; more advanced alternatives, such as a difference map algorithm, may be used to produce an image closer to the ideal, which is a promising approach to reduce total imaging time.   

A Comparison of Raman Spectral Features of Frozen and Deparaffinized Tissues in Neuroblastoma and Ganglioneuroma

We have investigated the cellular regions in neuroblastoma and ganglioneuroma using Raman spectroscopy and compared their spectral characteristics with those of normal adrenal gland. Thin sections from both frozen and deparaffinized tissues, obtained from the same tissue specimen, were studied in conjunction with the pathological examination of the tissues. We found a significant difference in the spectral features of frozen sections of normal adrenal gland, neuroblastoma, and ganglioneuroma when compared to deparaffinized tissues. The quantitative analysis of the Raman data using chemometric methods of principal component analysis and discriminant function analysis obtained from the frozen tissues show a sensitivity and specificity of 100{\%} each. The biochemical identification based on the spectral differences shows that the normal adrenal gland tissues have higher levels of carotenoids, lipids, and cholesterol compared to the neuroblastoma and ganglioneuroma frozen tissues. However, deparaffinized tissues show complete removal of these biochemicals in adrenal tissues. This study demonstrates that Raman spectroscopy combined with chemometric methods can successfully distinguish neuroblastoma and ganglioneuroma at cellular level.   

Expanding Cancer Detection Using Molecular Imprinting for a Novel Point-of-Care Diagnostic Device

We propose the use of a potentiometric biosensor that incorporates the efficient and specific molecular imprinting (MI) method with a self-assembled monolayer (SAM). We first tested the biosensor using carcinoembryonic antigen, CEA, a biomarker associated with pancreatic cancer. No change in detection efficiency was observed when detection was performed in the presence of 100{\%} serum albumin, indicating that the sensor is able to discriminate for the template analyte even in concentrated solution of similar substances. Computer simulations of the protein structure were performed in order to estimate the changes in morphology and determine the sensitivity of the biosensor to conformational changes in the proteins. We found that even small changes in PH can generate rotation of the surface functional groups, without significant change in the morphology. Yet, the results show that only when the detection and imprinting conditions are similar, robust signals occurs. Hence we concluded that both morphology and surface chemistry play a role in the recognition.