Deep learning extended depth-of-field microscope for fast and slide-free histology.

Microscopic evaluation of resected tissue plays a central role in the surgical management of cancer. Because optical microscopes have a limited depth-of-field (DOF), resected tissue is either frozen or preserved with chemical fixatives, sliced into thin sections placed on microscope slides, stained, and imaged to determine whether surgical margins are free of tumor cells-a costly and time- and labor-intensive procedure. Here, we introduce a deep-learning extended DOF (DeepDOF) microscope to quickly image large areas of freshly resected tissue to provide histologic-quality images of surgical margins without physical sectioning. The DeepDOF microscope consists of a conventional fluorescence microscope with the simple addition of an inexpensive (less than $10) phase mask inserted in the pupil plane to encode the light field and enhance the depth-invariance of the point-spread function. When used with a jointly optimized image-reconstruction algorithm, diffraction-limited optical performance to resolve subcellular features can be maintained while significantly extending the DOF (200 µm). Data from resected oral surgical specimens show that the DeepDOF microscope can consistently visualize nuclear morphology and other important diagnostic features across highly irregular resected tissue surfaces without serial refocusing. With the capability to quickly scan intact samples with subcellular detail, the DeepDOF microscope can improve tissue sampling during intraoperative tumor-margin assessment, while offering an affordable tool to provide histological information from resected tissue specimens in resource-limited settings.

First-Arrival Differential Counting for SPAD Array Design.

We present a novel architecture for the design of single-photon detecting arrays that captures relative intensity or timing information from a scene, rather than absolute. The proposed method for capturing relative information between pixels or groups of pixels requires very little circuitry, and thus allows for a significantly higher pixel packing factor than is possible with per-pixel TDC approaches. The inherently compressive nature of the differential measurements also reduces data throughput and lends itself to physical implementations of compressed sensing, such as Haar wavelets. We demonstrate this technique for HDR imaging and LiDAR, and describe possible future applications.

Snapshot polarimetric diffuse-specular separation.

We present a polarization-based approach to perform diffuse-specular separation from a single polarimetric image, acquired using a flexible, practical capture setup. Our key technical insight is that, unlike previous polarization-based separation methods that assume completely unpolarized diffuse reflectance, we use a more general polarimetric model that accounts for partially polarized diffuse reflections. We capture the scene with a polarimetric sensor and produce an initial analytical diffuse-specular separation that we further pass into a deep network trained to refine the separation. We demonstrate that our combination of analytical separation and deep network refinement produces state-of-the-art diffuse-specular separation, which enables image-based appearance editing of dynamic scenes and enhanced appearance estimation.

FlatNet3D: intensity and absolute depth from single-shot lensless capture.

Lensless cameras are ultra-thin imaging systems that replace the lens with a thin passive optical mask and computation. Passive mask-based lensless cameras encode depth information in their measurements for a certain depth range. Early works have shown that this encoded depth can be used to perform 3D reconstruction of close-range scenes. However, these approaches for 3D reconstructions are typically optimization based and require strong hand-crafted priors and hundreds of iterations to reconstruct. Moreover, the reconstructions suffer from low resolution, noise, and artifacts. In this work, we propose FlatNet3D-a feed-forward deep network that can estimate both depth and intensity from a single lensless capture. FlatNet3D is an end-to-end trainable deep network that directly reconstructs depth and intensity from a lensless measurement using an efficient physics-based 3D mapping stage and a fully convolutional network. Our algorithm is fast and produces high-quality results, which we validate using both simulated and real scenes captured using PhlatCam.

EDoF-ToF: extended depth of field time-of-flight imaging.

Conventional continuous-wave amplitude-modulated time-of-flight (CWAM ToF) cameras suffer from a fundamental trade-off between light throughput and depth of field (DoF): a larger lens aperture allows more light collection but suffers from significantly lower DoF. However, both high light throughput, which increases signal-to-noise ratio, and a wide DoF, which enlarges the system's applicable depth range, are valuable for CWAM ToF applications. In this work, we propose EDoF-ToF, an algorithmic method to extend the DoF of large-aperture CWAM ToF cameras by using a neural network to deblur objects outside of the lens's narrow focal region and thus produce an all-in-focus measurement. A key component of our work is the proposed large-aperture ToF training data simulator, which models the depth-dependent blurs and partial occlusions caused by such apertures. Contrary to conventional image deblurring where the blur model is typically linear, ToF depth maps are nonlinear functions of scene intensities, resulting in a nonlinear blur model that we also derive for our simulator. Unlike extended DoF for conventional photography where depth information needs to be encoded (or made depth-invariant) using additional hardware (phase masks, focal sweeping, etc.), ToF sensor measurements naturally encode depth information, allowing a completely software solution to extended DoF. We experimentally demonstrate EDoF-ToF increasing the DoF of a conventional ToF system by 3.6 ×, effectively achieving the DoF of a smaller lens aperture that allows 22.1 × less light. Ultimately, EDoF-ToF enables CWAM ToF cameras to enjoy the benefits of both high light throughput and a wide DoF.

Ultrafast and Ultrahigh-Resolution Diffuse Optical Tomography for Brain Imaging with Sensitivity Equation based Noniterative Sparse Optical Reconstruction (SENSOR).

We introduce a novel image reconstruction method for time-resolved diffuse optical tomography (DOT) that yields submillimeter resolution in less than a second. This opens the door to high-resolution real-time DOT in imaging of the brain activity. We call this approach the sensitivity equation based noniterative sparse optical reconstruction (SENSOR) method. The high spatial resolution is achieved by implementing an asymptotic l 0-norm operator that guarantees to obtain sparsest representation of reconstructed targets. The high computational speed is achieved by employing the nontruncated sensitivity equation based noniterative inverse formulation combined with reduced sensing matrix and parallel computing. We tested the new method with numerical and experimental data. The results demonstrate that the SENSOR algorithm can achieve 1 mm3 spatial-resolution optical tomographic imaging at depth of ∼60 mean free paths (MFPs) in 20∼30 milliseconds on an Intel Core i9 processor.

Vision and Deep Learning-Based Algorithms to Detect and Quantify Cracks on Concrete Surfaces from UAV Videos.

Immediate assessment of structural integrity of important civil infrastructures, like bridges, hospitals, or dams, is of utmost importance after natural disasters. Currently, inspection is performed manually by engineers who look for local damages and their extent on significant locations of the structure to understand its implication on its global stability. However, the whole process is time-consuming and prone to human errors. Due to their size and extent, some regions of civil structures are hard to gain access for manual inspection. In such situations, a vision-based system of Unmanned Aerial Vehicles (UAVs) programmed with Artificial Intelligence algorithms may be an effective alternative to carry out a health assessment of civil infrastructures in a timely manner. This paper proposes a framework of achieving the above-mentioned goal using computer vision and deep learning algorithms for detection of cracks on the concrete surface from its image by carrying out image segmentation of pixels, i.e., classification of pixels in an image of the concrete surface and whether it belongs to cracks or not. The image segmentation or dense pixel level classification is carried out using a deep neural network architecture named U-Net. Further, morphological operations on the segmented images result in dense measurements of crack geometry, like length, width, area, and crack orientation for individual cracks present in the image. The efficacy and robustness of the proposed method as a viable real-life application was validated by carrying out a laboratory experiment of a four-point bending test on an 8-foot-long concrete beam of which the video is recorded using a camera mounted on a UAV-based, as well as a still ground-based, video camera. Detection, quantification, and localization of damage on a civil infrastructure using the proposed framework can directly be used in the prognosis of the structure's ability to withstand service loads.

Imaging through plasmonic nanoparticles.

The optical properties of metallic nanoparticles with plasmon resonances have been studied extensively, typically by measuring the transmission of light, as a function of wavelength, through a nanoparticle suspension. One question that has not yet been addressed, however, is how an image is transmitted through such a suspension of absorber-scatterers, in other words, how the various spatial frequencies are attenuated as they pass through the nanoparticle host medium. Here, we examine how the optical properties of a suspension of plasmonic nanoparticles affect the transmitted image. We use two distinct ways to assess transmitted image quality: the structural similarity index (SSIM), a perceptual distortion metric based on the human visual system, and the modulation transfer function (MTF), which assesses the resolvable spatial frequencies. We show that perceived image quality, as well as spatial resolution, are both dependent on the scattering and absorption cross-sections of the constituent nanoparticles. Surprisingly, we observe a nonlinear dependence of image quality on optical density by varying optical path length and nanoparticle concentration. This work is a first step toward understanding the requirements for visualizing and resolving objects through media consisting of subwavelength absorber-scatterer structures, an approach that should also prove useful in the assessment of metamaterial or metasurface-based optical imaging systems.

Integrated light-sheet illumination using metallic slit microlenses.

Light sheet microscopy (LSM) - also known as selective plane illumination microscopy (SPIM) - enables high-speed, volumetric imaging by illuminating a two-dimensional cross-section of a specimen. Typically, this light sheet is created by table-top optics, which limits the ability to miniaturize the overall SPIM system. Replacing this table-top illumination system with miniature, integrated devices would reduce the cost and footprint of SPIM systems. One important element for a miniature SPIM system is a flat, easily manufactured lens that can form a light sheet. Here we investigate planar metallic lenses as the beam shaping element of an integrated SPIM illuminator. Based on finite difference time domain (FDTD) simulations, we find that diffraction from a single slit can create planar illumination with a higher light throughput than zone plate or plasmonic lenses. Metallic slit microlenses also show broadband operation across the entire visible range and are nearly polarization insensitive. Furthermore, compared to meta-lenses based on sub-wavelength-scale diffractive elements, metallic slit lenses have micron-scale features compatible with low-cost photolithographic manufacturing. These features allow us to create inexpensive integrated devices that generate light-sheet illumination comparable to tabletop microscopy systems. Further miniaturization of this type of integrated SPIM illuminators will open new avenues for flat, implantable photonic devices for in vivo biological imaging.

CS-ToF: High-resolution compressive time-of-flight imaging.

Three-dimensional imaging using Time-of-flight (ToF) sensors is rapidly gaining widespread adoption in many applications due to their cost effectiveness, simplicity, and compact size. However, the current generation of ToF cameras suffers from low spatial resolution due to physical fabrication limitations. In this paper, we propose CS-ToF, an imaging architecture to achieve high spatial resolution ToF imaging via optical multiplexing and compressive sensing. Our approach is based on the observation that, while depth is non-linearly related to ToF pixel measurements, a phasor representation of captured images results in a linear image formation model. We utilize this property to develop a CS-based technique that is used to recover high resolution 3D images. Based on the proposed architecture, we developed a prototype 1-megapixel compressive ToF camera that achieves as much as 4× improvement in spatial resolution and 3× improvement for natural scenes. We believe that our proposed CS-ToF architecture provides a simple and low-cost solution to improve the spatial resolution of ToF and related sensors.

Mesoscopic calcium imaging in a head-unrestrained male non-human primate using a lensless microscope.

Mesoscopic calcium imaging enables studies of cell-type specific neural activity over large areas. A growing body of literature suggests that neural activity can be different when animals are free to move compared to when they are restrained. Unfortunately, existing systems for imaging calcium dynamics over large areas in non-human primates (NHPs) are table-top devices that require restraint of the animal's head. Here, we demonstrate an imaging device capable of imaging mesoscale calcium activity in a head-unrestrained male non-human primate. We successfully miniaturize our system by replacing lenses with an optical mask and computational algorithms. The resulting lensless microscope can fit comfortably on an NHP, allowing its head to move freely while imaging. We are able to measure orientation columns maps over a 20 mm2 field-of-view in a head-unrestrained macaque. Our work establishes mesoscopic imaging using a lensless microscope as a powerful approach for studying neural activity under more naturalistic conditions.

Metric-guided Image Reconstruction Bounds via Conformal Prediction.

Recent advancements in machine learning have led to novel imaging systems and algorithms that address ill-posed problems. Assessing their trustworthiness and understanding how to deploy them safely at test time remains an important and open problem. We propose a method that leverages conformal prediction to retrieve upper/lower bounds and statistical inliers/outliers of reconstructions based on the prediction intervals of downstream metrics. We apply our method to sparse-view CT for downstream radiotherapy planning and show 1) that metric-guided bounds have valid coverage for downstream metrics while conventional pixel-wise bounds do not and 2) anatomical differences of upper/lower bounds between metric-guided and pixel-wise methods. Our work paves the way for more meaningful reconstruction bounds. Code available at https://github.com/matthewyccheung/conformal-metric.

DeepDOF-SE: affordable deep-learning microscopy platform for slide-free histology.

Histopathology plays a critical role in the diagnosis and surgical management of cancer. However, access to histopathology services, especially frozen section pathology during surgery, is limited in resource-constrained settings because preparing slides from resected tissue is time-consuming, labor-intensive, and requires expensive infrastructure. Here, we report a deep-learning-enabled microscope, named DeepDOF-SE, to rapidly scan intact tissue at cellular resolution without the need for physical sectioning. Three key features jointly make DeepDOF-SE practical. First, tissue specimens are stained directly with inexpensive vital fluorescent dyes and optically sectioned with ultra-violet excitation that localizes fluorescent emission to a thin surface layer. Second, a deep-learning algorithm extends the depth-of-field, allowing rapid acquisition of in-focus images from large areas of tissue even when the tissue surface is highly irregular. Finally, a semi-supervised generative adversarial network virtually stains DeepDOF-SE fluorescence images with hematoxylin-and-eosin appearance, facilitating image interpretation by pathologists without significant additional training. We developed the DeepDOF-SE platform using a data-driven approach and validated its performance by imaging surgical resections of suspected oral tumors. Our results show that DeepDOF-SE provides histological information of diagnostic importance, offering a rapid and affordable slide-free histology platform for intraoperative tumor margin assessment and in low-resource settings.

Broadband thermal imaging using meta-optics.

Subwavelength diffractive optics known as meta-optics have demonstrated the potential to significantly miniaturize imaging systems. However, despite impressive demonstrations, most meta-optical imaging systems suffer from strong chromatic aberrations, limiting their utilities. Here, we employ inverse-design to create broadband meta-optics operating in the long-wave infrared (LWIR) regime (8-12 µm). Via a deep-learning assisted multi-scale differentiable framework that links meta-atoms to the phase, we maximize the wavelength-averaged volume under the modulation transfer function (MTF) surface of the meta-optics. Our design framework merges local phase-engineering via meta-atoms and global engineering of the scatterer within a single pipeline. We corroborate our design by fabricating and experimentally characterizing all-silicon LWIR meta-optics. Our engineered meta-optic is complemented by a simple computational backend that dramatically improves the quality of the captured image. We experimentally demonstrate a six-fold improvement of the wavelength-averaged Strehl ratio over the traditional hyperboloid metalens for broadband imaging.

Roadmap for Optical Metasurfaces.

Metasurfaces have recently risen to prominence in optical research, providing unique functionalities that can be used for imaging, beam forming, holography, polarimetry, and many more, while keeping device dimensions small. Despite the fact that a vast range of basic metasurface designs has already been thoroughly studied in the literature, the number of metasurface-related papers is still growing at a rapid pace, as metasurface research is now spreading to adjacent fields, including computational imaging, augmented and virtual reality, automotive, display, biosensing, nonlinear, quantum and topological optics, optical computing, and more. At the same time, the ability of metasurfaces to perform optical functions in much more compact optical systems has triggered strong and constantly growing interest from various industries that greatly benefit from the availability of miniaturized, highly functional, and efficient optical components that can be integrated in optoelectronic systems at low cost. This creates a truly unique opportunity for the field of metasurfaces to make both a scientific and an industrial impact. The goal of this Roadmap is to mark this "golden age" of metasurface research and define future directions to encourage scientists and engineers to drive research and development in the field of metasurfaces toward both scientific excellence and broad industrial adoption.

Real-time, deep-learning aided lensless microscope.

Traditional miniaturized fluorescence microscopes are critical tools for modern biology. Invariably, they struggle to simultaneously image with a high spatial resolution and a large field of view (FOV). Lensless microscopes offer a solution to this limitation. However, real-time visualization of samples is not possible with lensless imaging, as image reconstruction can take minutes to complete. This poses a challenge for usability, as real-time visualization is a crucial feature that assists users in identifying and locating the imaging target. The issue is particularly pronounced in lensless microscopes that operate at close imaging distances. Imaging at close distances requires shift-varying deconvolution to account for the variation of the point spread function (PSF) across the FOV. Here, we present a lensless microscope that achieves real-time image reconstruction by eliminating the use of an iterative reconstruction algorithm. The neural network-based reconstruction method we show here, achieves more than 10000 times increase in reconstruction speed compared to iterative reconstruction. The increased reconstruction speed allows us to visualize the results of our lensless microscope at more than 25 frames per second (fps), while achieving better than 7 µm resolution over a FOV of 10 mm2. This ability to reconstruct and visualize samples in real-time empowers a more user-friendly interaction with lensless microscopes. The users are able to use these microscopes much like they currently do with conventional microscopes.

CoIR: Compressive Implicit Radar.

Using millimeter wave (mmWave) signals for imaging has an important advantage in that they can penetrate through poor environmental conditions such as fog, dust, and smoke that severely degrade optical-based imaging systems. However, mmWave radars, contrary to cameras and LiDARs, suffer from low angular resolution because of small physical apertures and conventional signal processing techniques. Sparse radar imaging, on the other hand, can increase the aperture size while minimizing the power consumption and read out bandwidth. This paper presents CoIR, an analysis by synthesis method that leverages the implicit neural network bias in convolutional decoders and compressed sensing to perform high accuracy sparse radar imaging. The proposed system is data set-agnostic and does not require any auxiliary sensors for training or testing. We introduce a sparse array design that allows for a 5.5× reduction in the number of antenna elements needed compared to conventional MIMO array designs. We demonstrate our system's improved imaging performance over standard mmWave radars and other competitive untrained methods on both simulated and experimental mmWave radar data.

SpeckleCam: high-resolution computational speckle contrast tomography for deep blood flow imaging.

Laser speckle contrast imaging is widely used in clinical studies to monitor blood flow distribution. Speckle contrast tomography, similar to diffuse optical tomography, extends speckle contrast imaging to provide deep tissue blood flow information. However, the current speckle contrast tomography techniques suffer from poor spatial resolution and involve both computation and memory intensive reconstruction algorithms. In this work, we present SpeckleCam, a camera-based system to reconstruct high resolution 3D blood flow distribution deep inside the skin. Our approach replaces the traditional forward model using diffuse approximations with Monte-Carlo simulations-based convolutional forward model, which enables us to develop an improved deep tissue blood flow reconstruction algorithm. We show that our proposed approach can recover complex structures up to 6 mm deep inside a tissue-like scattering medium in the reflection geometry. We also conduct human experiments to demonstrate that our approach can detect reduced flow in major blood vessels during vascular occlusion.

High-Speed Time-Domain Diffuse Optical Tomography with a Sensitivity Equation-based Neural Network.

Steady progress in time-domain diffuse optical tomography (TD-DOT) technology is allowing for the first time the design of low-cost, compact, and high-performance systems, thus promising more widespread clinical TD-DOT use, such as for recording brain tissue hemodynamics. TD-DOT is known to provide more accurate values of optical properties and physiological parameters compared to its frequency-domain or steady-state counterparts. However, achieving high temporal resolution is still difficult, as solving the inverse problem is computationally demanding, leading to relatively long reconstruction times. The runtime is further compromised by processes that involve 'nontrivial' empirical tuning of reconstruction parameters, which increases complexity and inefficiency. To address these challenges, we present a new reconstruction algorithm that combines a deep-learning approach with our previously introduced sensitivity-equation-based, non-iterative sparse optical reconstruction (SENSOR) code. The new algorithm (called SENSOR-NET) unfolds the iterations of SENSOR into a deep neural network. In this way, we achieve high-resolution sparse reconstruction using only learned parameters, thus eliminating the need to tune parameters prior to reconstruction empirically. Furthermore, once trained, the reconstruction time is not dependent on the number of sources or wavelengths used. We validate our method with numerical and experimental data and show that accurate reconstructions with 1 mm spatial resolution can be obtained in under 20 milliseconds regardless of the number of sources used in the setup. This opens the door for real-time brain monitoring and other high-speed DOT applications.

Unrolled-DOT: an interpretable deep network for diffuse optical tomography.

Significance: Imaging through scattering media is critical in many biomedical imaging applications, such as breast tumor detection and functional neuroimaging. Time-of-flight diffuse optical tomography (ToF-DOT) is one of the most promising methods for high-resolution imaging through scattering media. ToF-DOT and many traditional DOT methods require an image reconstruction algorithm. Unfortunately, this algorithm often requires long computational runtimes and may produce lower quality reconstructions in the presence of model mismatch or improper hyperparameter tuning. Aim: We used a data-driven unrolled network as our ToF-DOT inverse solver. The unrolled network is faster than traditional inverse solvers and achieves higher reconstruction quality by accounting for model mismatch. Approach: Our model "Unrolled-DOT" uses the learned iterative shrinkage thresholding algorithm. In addition, we incorporate a refinement U-Net and Visual Geometry Group (VGG) perceptual loss to further increase the reconstruction quality. We trained and tested our model on simulated and real-world data and benchmarked against physics-based and learning-based inverse solvers. Results: In experiments on real-world data, Unrolled-DOT outperformed learning-based algorithms and achieved over 10× reduction in runtime and mean-squared error, compared to traditional physics-based solvers. Conclusion: We demonstrated a learning-based ToF-DOT inverse solver that achieves state-of-the-art performance in speed and reconstruction quality, which can aid in future applications for noninvasive biomedical imaging.