Search for: All records

DeepTensor is a computationally efficient framework for low-rank decomposition of matrices and tensors using deep generative networks. We decompose a tensor as the product of low-rank tensor factors where each low-rank tensor is generated by a deep network (DN) that is trained in a self-supervised manner to minimize the mean-square approximation error. Our key observation is that the implicit regularization inherent in DNs enables them to capture nonlinear signal structures that are out of the reach of classical linear methods like the singular value decomposition (SVD) and principal components analysis (PCA). We demonstrate that the performance of DeepTensor is robust to a wide range of distributions and a computationally efficient drop-in replacement for the SVD, PCA, nonnegative matrix factorization (NMF), and similar decompositions by exploring a range of real-world applications, including hyperspectral image denoising, 3D MRI tomography, and image classification. 

Significance: Accurate identification between pathologic (e.g., tumors) and healthy brain tissue is a critical need in neurosurgery. However, conventional surgical adjuncts have significant limitations toward achieving this goal (e.g., image guidance based on pre-operative imaging becomes inaccurate up to 3 cm as surgery proceeds). Hyperspectral imaging (HSI) has emerged as a potential powerful surgical adjunct to enable surgeons to accurately distinguish pathologic from normal tissues. Aim: We review HSI techniques in neurosurgery; categorize, explain, and summarize their technical and clinical details; and present some promising directions for future work. Approach: We performed a literature search on HSI methods in neurosurgery focusing on their hardware and implementation details; classification, estimation, and band selection methods; publicly available labeled and unlabeled data; image processing and augmented reality visualization systems; and clinical study conclusions. Results: We present a detailed review of HSI results in neurosurgery with a discussion of over 25 imaging systems, 45 clinical studies, and 60 computational methods. We first provide a short overview of HSI and the main branches of neurosurgery. Then, we describe in detail the imaging systems, computational methods, and clinical results for HSI using reflectance or fluorescence. Clinical implementations of HSI yield promising results in estimating perfusion and mapping brain function, classifying tumors and healthy tissues (e.g., in fluorescence-guided tumor surgery, detecting infiltrating margins not visible with conventional systems), and detecting epileptogenic regions. Finally, we discuss the advantages and disadvantages of HSI approaches and interesting research directions as a means to encourage future development. Conclusions: We describe a number of HSI applications across every major branch of neurosurgery. We believe these results demonstrate the potential of HSI as a powerful neurosurgical adjunct as more work continues to enable rapid acquisition with smaller footprints, greater spectral and spatial resolutions, and improved detection. 

Abstract 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 (8-12μm). 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. 

Foveated imaging provides a better tradeoff between situational awareness (field of view) and resolution, and is critical in long wavelength infrared regimes because of the size, weight, power, and cost of thermal sensors. We demonstrate computational foveated imaging by exploiting the ability of a meta-optical frontend to discriminate between different polarization states and a computational backend to reconstruct the captured image/video. The frontend is a three-element optic: the first element, which we call the “foveal” element, is a metalens that focuses s-polarized light at a distance off₁without affecting the p-polarized light; the second element, which we call the “perifovea” element, is another metalens that focuses p-polarized light at a distance off₂without affecting thes-polarized light. The third element is a freely rotating polarizer that dynamically changes the mixing ratios between the two polarization states. Both the foveal element (focal length=150mm; diameter=75mm) and the perifoveal element (focal length=25mm; diameter=25mm) were fabricated as polarization-sensitive, all-silicon, meta surfaces resulting in a large-aperture, 1:6 foveal expansion, thermal imaging capability. A computational backend then utilizes a deep image prior to separate the resultant multiplexed image or video into a foveated image consisting of a high resolution center and a lower-resolution large field of view context. We build a prototype system and demonstrate 12 frames per second real-time, thermal, foveated image and video capture.. 

We introduce a new neural signal model designed for efficient high-resolution representation of large-scale signals. The key innovation in our multiscale implicit neural representation (MINER) is an internal representation via a Laplacian pyramid, which provides a sparse multiscale decomposition of the signal that captures orthogonal parts of the signal across scales. We leverage the advantages of the Laplacian pyramid by representing small disjoint patches of the pyramid at each scale with a small MLP. This enables the capacity of the network to adaptively increase from coarse to fine scales, and only represent parts of the signal with strong signal energy. The parameters of each MLP are optimized from coarse-to-fine scale which results in faster approximations at coarser scales, thereby ultimately an extremely fast training process. We apply MINER to a range of large-scale signal representation tasks, including gigapixel images and very large point clouds, and demonstrate that it requires fewer than 25% of the parameters, 33% of the memory footprint, and 10% of the computation time of competing techniques such as ACORN to reach the same representation accuracy. 

We introduce DeepIR, a new thermal image processing framework that combines physically accurate sensor modeling with deep network-based image representation. Our key enabling observations are that the images captured by thermal sensors can be factored into slowly changing, scene-independent sensor non-uniformities (that can be accurately modeled using physics) and a scene-specific radiance flux (that is well-represented using a deep network-based regularizer). DeepIR requires neither training data nor periodic ground-truth calibration with a known black body target--making it well suited for practical computer vision tasks. We demonstrate the power of going DeepIR by developing new denoising and super-resolution algorithms that exploit multiple images of the scene captured with camera jitter. Simulated and real data experiments demonstrate that DeepIR can perform high-quality non-uniformity correction with as few as three images, achieving a 10dB PSNR improvement over competing approaches.