We developed a new imaging platform using hyperspectral SRS imaging to detect metabolic dynamics in living organisms. Within the broad vibrational spectra, we visualized more than 20 different molecules including lipids subtypes-, protein-, and DNA-specific Raman profiles and develop hyperspectral detection methods to obtain various macromolecular multiplex imaging. We further developed deconvolution algorithm to enhance the spatial resolution to generate super resolution SRS hyperspectral images for visualizing subcellular distribution of various molecules. We applied this method to study the diet regulated metabolic dynamics in animals during aging processes, the quantitative lipid and protein turnover rate, the intra-cellular metabolic heterogeneity.

Label-free nonlinear optical microscopy has become a powerful tool for biomedical research. However, the low imaging speed and the accompanying photodamage risk hinder further clinical applications. To reduce these adverse effects, in this study, we constructed a new generation of simultaneous label-free autofluorescence multiharmonic (SLAM) microscopy, featuring high-speed, multimodal imaging, monitorable-photodamage, and tunable ultrashort pulses. By using birefringent photonic crystal fiber and a pulse shaper, this system has the ability to allow users to independently adjust repetition rate, pulse width, and average power without overlapping interference, and can realize multiphoton generation in each channel from a single excitation pulse. These outstanding advantages represent a powerful and user-friendly imaging platform.

We report the utility of fiberoptic Raman spectroscopy for realizing post-treatment NPC patients surveying and accurate detection of tumor recurrence. Distinct Raman spectral differences are observed among the tissue Raman spectra of normal, NPC, and non-recurring post-treatment patients. The classification models using the in vivo fingerprint and high-wavenumber (FP/HW) tissue Raman spectra together with the partial-least-squares linear-discriminant-analysis (PLS-LDA) provide the high diagnostic accuracy for detecting recurrent NPC from both inflammation and long-term post-treatment fibrosis. We further quantitatively analyze the major biochemicals related to the NPC malignancy (e.g., triolein, elastin, keratin, fibrillar collagen, and type IV collagen, etc.). This study suggests that fiberoptic Raman spectroscopy can enable real-time in-vivo post-treatment patients surveying and tumor recurrence detection with high biomolecular sensitiveness.

We present a supervised learning approach to train a deep neural network which can transform images of H&E stained tissue sections into special stains (e.g., PAS, Jones silver stain and Masson’s Trichrome). We performed a diagnostic study using tissue sections from 58 subjects covering a variety of non-neoplastic kidney diseases to show that when the pathologists performed their diagnoses using the three virtually-created special stains in addition to H&E, a statistically significant diagnostic improvement was made over the use of H&E only. This virtual staining technique can be used to improve preliminary diagnoses while saving time and reducing costs.

In this study, we trained a convolutional neural network (CNN) utilizing a mix of recent CNN architectural design strategies. Our goals are to leverage these modern techniques to improve the binary classification of kidney tumor images obtained using Multi-Photon Microscopy (MPM). We demonstrate that incorporating these newer model design elements, coupled with transfer learning, image standardization, and data augmentation, leads to significantly increased classification performance over previous results. Our best model averages over 90% sensitivity, specificity, accuracy, area under the receiver operating characteristic curve (AUROC) in image-level classification across cross-validation folds, superior to the previous best in all four metrics.

Differential interference contrast (DIC) microscopy is widely used in biomedical studies for imaging transparent samples due to its excellent lateral resolution and axial discrimination. Aided by computation, quantitative DIC, which inherits the advantages of the DIC approach, has emerged as one attractive quantitative phase imaging modality. For the routine Hematoxylin-Eosin (H and E) stained histopathology specimens, the concentration maps of Hematoxylin and Eosin carry additional diagnostic information because of the negatively charged affinity of acidic eosin for cytoplasmic proteins and the positively charged affinity of basic hematoxylin for nuclear structures. It will be desirable to quantify the stain concentration maps in addition to the quantitative phase for H and E stained specimens. Here we present Quantitative phase and stain concentration imaging using Single-shot color differential interference contrast (C-DIC) microscopy for photonic histopathology for hematoxylin and eosin-stained tissue sections. The recorded image is first normalized by the corresponding image for blank slides under identical imaging conditions. The absorbance contributions from the H and E stains and the contributions from the phase gradient of the H and E stained specimen are unmixed from the recorded color-DIC image. The concentrations of the H and E stains are computed based on their respective extinction coefficients at the R: G: B: channels of the color camera. The quantitative phase map of the specimen is obtained from the results of the phase gradient. We apply C-DIC to image a cohort of routine prostate cancer specimens. The accuracy of diagnosis of prostate cancer improves with the addition of stain concentration maps compared to the quantitative phase alone.

The availability and the cost of 3D imaging systems are still a problem nowadays, which brings us to the need and urgency of a new way to democratize optical biopsy. Inspired by structural illumination and diffuse optical imaging, we propose a 3D-multiplexed diffused optical imaging (3D-mDOI) solution, a technique to reconstruct the 3D optical properties of the tissue from 2D diffuse images and estimate the depth of tissue lesions. 3D-mDOI uses a low-cost and contact-free design of the imaging acquisition platform, integrating a digital micromirror device (DMD) and an infrared-enhanced CCD camera. The imaging setup that creates custom sampling patterns for tissue photon migration enables spatial multiplexing to overcome low photon signals. We design a hybrid reconstruction pipeline for harvesting the benefits from existing mathematical solutions. The analytical solution of the steady-state radiation transfer equation is utilized to compute each pixel's optical properties in 2D. Monte Carlo simulation provides the stochastic solution for 3D photon diffusion patterns on the discretized tissue volume. We then map the 2D optical properties to the corresponding 3D photon diffusion patterns between a light source and a detector. To better correct the instrumental noises, we design multiple calibrations. 3D-mDOI is versatile, non-invasive, and cost-effective, containing 3D insights to subsurface molecular composition. The technique reconstructs lesions up to 5mm below the surface with 0.2mm axial spatial resolution. We could apply the solution to broad applications in the scientific and medical fields, including the rapid estimation of melanoma staging.

The aim of this study was to investigate the feasibility of multimodal optical imaging for reliable detection of thyroid cancer. Samples were collected from discarded malignant, benign, and normal thyroid tissues following surgery. The results indicate that the proposed method may provide a rapid and accurate tool to assist cytopathological differentiation of thyroid lesions.