Label- and slide-free tissue histology using 3D epi-mode quantitative phase imaging and virtual H&E staining.

Histological staining of tissue biopsies, especially hematoxylin and eosin (H&E) staining, serves as the benchmark for disease diagnosis and comprehensive clinical assessment of tissue. However, the process is laborious and time-consuming, often limiting its usage in crucial applications such as surgical margin assessment. To address these challenges, we combine an emerging 3D quantitative phase imaging technology, termed quantitative oblique back illumination microscopy (qOBM), with an unsupervised generative adversarial network pipeline to map qOBM phase images of unaltered thick tissues (i.e., label- and slide-free) to virtually stained H&E-like (vH&E) images. We demonstrate that the approach achieves high-fidelity conversions to H&E with subcellular detail using fresh tissue specimens from mouse liver, rat gliosarcoma, and human gliomas. We also show that the framework directly enables additional capabilities such as H&E-like contrast for volumetric imaging. The quality and fidelity of the vH&E images are validated using both a neural network classifier trained on real H&E images and tested on virtual H&E images, and a user study with neuropathologists. Given its simple and low-cost embodiment and ability to provide real-time feedback in vivo, this deep learning-enabled qOBM approach could enable new workflows for histopathology with the potential to significantly save time, labor, and costs in cancer screening, detection, treatment guidance, and more.

Optimization of a flexible fiber-optic probe for epi-mode quantitative phase imaging.

Quantitative oblique back-illumination microscopy (qOBM) is an emerging label-free optical imaging technology that enables 3D, tomographic quantitative phase imaging (QPI) with epi-illumination in thick scattering samples. In this work, we present a robust optimization of a flexible, fiber-optic-based qOBM system. Our approach enables in silico optimization of the phase signal-to-noise-ratio over a wide parameter space and obviates the need for tedious experimental optimization which could easily miss optimal conditions. Experimental validations of the simulations are also presented and sensitivity limits for the probe are assessed. The optimized probe is light-weight (∼40g) and compact (8mm in diameter) and achieves a 2µm lateral resolution, 6µm axial resolution, and a 300µm field of view, with near video-rate operation (10Hz, limited by the camera). The phase sensitivity is <20nm for a single qOBM acquisition (at 10Hz) and a lower limit of ∼3 nm via multi-frame averaging. Finally, to demonstrate the utility of the optimized probe, we image (1) thick, fixed rat brain samples from a 9L gliosarcoma tumor model and (2) freshly excised human brain tissues from neurosurgery. Acquired qOBM images using the flexible fiber-optic probe are in excellent agreement with those from a free-space qOBM system (both in-situ), as well as with gold-standard histopathology slices (after tissue processing).

Towards in-vivo label-free detection of brain tumor margins with epi-illumination tomographic quantitative phase imaging.

Brain tumor surgery involves a delicate balance between maximizing the extent of tumor resection while minimizing damage to healthy brain tissue that is vital for neurological function. However, differentiating between tumor, particularly infiltrative disease, and healthy brain in-vivo remains a significant clinical challenge. Here we demonstrate that quantitative oblique back illumination microscopy (qOBM)-a novel label-free optical imaging technique that achieves tomographic quantitative phase imaging in thick scattering samples-clearly differentiates between healthy brain tissue and tumor, including infiltrative disease. Data from a bulk and infiltrative brain tumor animal model show that qOBM enables quantitative phase imaging of thick fresh brain tissues with remarkable cellular and subcellular detail that closely resembles histopathology using hematoxylin and eosin (H&E) stained fixed tissue sections, the gold standard for cancer detection. Quantitative biophysical features are also extracted from qOBM which yield robust surrogate biomarkers of disease that enable (1) automated tumor and margin detection with high sensitivity and specificity and (2) facile visualization of tumor regions. Finally, we develop a low-cost, flexible, fiber-based handheld qOBM device which brings this technology one step closer to in-vivo clinical use. This work has significant implications for guiding neurosurgery by paving the way for a tool that delivers real-time, label-free, in-vivo brain tumor margin detection.