LLNL’s Distributed Implicit Neural Representation (DINR) is a novel approach to 4D time-space reconstruction of dynamic objects.  DINR is the first technology to enable 4D imaging of dynamic objects at sufficiently high spatial and temporal resolutions that are necessary for real world medical and industrial applications. 

The essence of this invention is a method that couples network architecture using neural implicit representations coupled with a novel parametric motion field to perform limited angle 4D-CT reconstruction of deforming scenes.

LLNL researchers have invented an ultrafast PCSS to drive a high-power laser diode with arbitrary pulse widths.  These devices operate by supplying a high voltage (>10 kV) to one side of the switch.  A short pulse of light illuminates the semiconductor, instantly turning it from highly resistive to highly conductive. Ultrawide bandgap (UWBG) semiconductors are used to achieve sub-…

LLNL’s novel technology automates the inspection process by using a scanning system that captures data within the walnut shell without having to open the shell. The system output gives a visual image inside the walnut shell sufficient to evaluate and rate the quality of the walnut.  The system uses a camara and radar that can capture data at a rapid rate. This improves speed and…

The approach is to develop a solid-state X-ray imager based on the architecture of the Silicon Drift Detector (SDD) which uses a series of cathode strips on both sides of a silicon wafer to achieve bulk depletion and electron drift.  The invention leverages this SDD functionality to achieve signal stretching of liberated charge carriers from X-Ray photons that converts the time domain…

LLNL’s novel approach is to use diamond substrates with the desired donor (nitrogen) and acceptor (boron) impurities.   In order to optically activate these deep impurities, the invention requires at least one externally or internally integrated light source.  The initial exposure to light can set up the desired conduction current, after which the light source could be turned…

Instead of producing individual DSRDs and bonding them, Tunnel DSRD's entire stack structure is grown epitaxially on a n- or p-type silicon wafer, resulting in a novel, “monolithic” stacked DSRD.  A tunnel diode is essentially a diode with very highly doped p and n regions such that the reverse breakdown voltage is 200 meV or lower. 

LLNL’s novel approach is to use a continuous moving camera with a scan speed of >1 mm/sec and a frame rate of 100 frames per second.  The key is to have a light source that flashes with a duration of one nanosecond, thus essentially freezing the image with no blur.  Clear images of high resolution can then be captured through a high-magnification objective lens (reflection mode)…

For cooling a high power device, the novel approach is to use a thermoelectric cooler (TEC)-based embedded substrate with proper selection of the TEC material as an active cooler.  The packaging configuration of TEC allows cooling the entire die without the use of a fluid.  The process is compatible with the thin film TEC material.  Standard semiconductor processes can be used…

LLNL’s approach to the development of a wide-field, three-dimensional quantum (3DQ) microscope is to harness quantum entangled photons to form simultaneous 3D optical images, which could be a new paradigm for 3D volumetric imaging of biological specimens.  The 3DQ microscope is comprised of a novel optical system with highly sensitive detectors and an on-demand light source of entangled…

For this method, a Silicon on Insulator (SOI) wafer is used to tailor etch rates and thickness in initial steps of the process.  The simple three step process approach is comprised of grayscale lithography, deep reactive-ion etch (DRIE) and liftoff of the SOI wafer.  The liftoff process is used to dissolve the insulating layer, thus separating sections of the wafer as individual…

The approach is to use Charge Balance Layers (CBLs) to create a superjunction device in wide bandgap materials.  These CBLs enable the device to effectively spread the electric field over 2- or 3-dimensions within a semiconductor voltage sustaining layer instead of 1-dimension, thereby increasing the maximum voltage a device is capable of withstanding.  The challenge of using CBLs is…

Lawrence Livermore National Laboratory (LLNL) is offering the opportunity to collaborate in accelerating artificial intelligence (AI) for applied science, including research in key areas such as advanced material design, 3D printing, predictive biology, energy systems, “self-driving” lasers and fusion energy research.

To solve these challenges using new and existing CT system designs, LLNL has developed an innovative software package for CT data processing and reconstruction. Livermore Tomography Tools (LTT) is a modern integrated software package that includes all aspects of CT modeling, simulation, reconstruction, and analysis algorithms based on the latest research in the field. LTT contains the most…

Clinical images have a wealth of data that are currently untapped by physicians and machine learning (ML) methods alike. Most ML methods require more data than is available to sufficiently train them. In order to obtain all data contained in a clinical image, it is imperative to be able to utilize multimodal, or various types of, data such as tags or identifications, especially where spatial…

Some COVID-19 diagnoses are utilizing computed tomography (CT)-scans for triage. CT-scans produce immediate results with high sensitivity. The digital images produced by a CT-scan require physicians to identify objects within the image to determine the presence of disease. Object identification can be done using machine learning (ML) techniques such as deep learning (DL) to improve speed and…

MimicGAN represents a new generation of methods that can “self-correct” for unseen corruptions in the data out in the field. This is particularly useful for systems that need to be deployed autonomously without needing constant intervention such as Automated Driver Assistance Systems. MimicGAN achieves this by treating every test sample as “corrupt” by default. The goal is to determine (a) the…

LLNL's 3D X-ray imager combines two different hardware pieces. The first is an x-ray optic with a depth-of-field that is small compared to the object under investigation. Reflective Wolter type x-ray optics are one such design. These hollow optics have a relatively large collection efficiency and can be designed with a large field of view. The depth of focus, which is the distance over which a…

LLNL has developed a new system, called the Segmentation Ensembles System, that provides a simple and general way to fuse high-level and low-level information and leads to a substantial increase in overall performance of digital image analysis. LLNL researchers have demonstrated the effectiveness of the approach on applications ranging from automatic threat detection for airport security, to…

LLNL's high fidelity hydrocode is capable of predicting blast loads and directly coupling those loads to structures to predict a mechanical response. By combining this code and our expertise in modeling blast-structure interaction and damage, along with our access to experimental data and testing facilities, we can contribute to the design of protective equipment that can better mitigate the…