Using neural networks to calibrate agent based models enables improved regional evidence for vaccine strategy and policy.

Distribution and administration strategy are critical to successful population immunization efforts. Agent-based modeling (ABM) can reflect the complexity of real-world populations and can experimentally evaluate vaccine strategy and policy. However, ABMs historically have been limited in their time-to-development, long runtime, and difficulty calibrating. Our team had several technical advances in the development of our GradABMs: a novel class of scalable, fast and differentiable simulations. GradABMs can simulate million-size populations in a few seconds on commodity hardware, integrate with deep neural networks and ingest heterogeneous sources. This allows for rapid and real-world sensitivity analyses. Our first epidemiological GradABM (EpiABMv1) enabled simulation interventions over real million-scale populations and was used in vaccine strategy and policy during the COVID-19 pandemic. Literature suggests decisions aided by evidence from these models saved thousands of lives. Our most recent model (EpiABMv2) extends EpiABMv1 to allow improved regional calibration using deep neural networks to incorporate local population data, and in some cases different policy recommendations versus our prior models. This is an important advance for our model to be more effective at vaccine strategy and policy decisions at the local public health level.

Detection and mapping of specular surfaces using multibounce LiDAR returns.

We propose methods that use specular, multibounce LiDAR returns to detect and map specular surfaces that might be invisible to conventional LiDAR systems that rely on direct, single-scatter returns. We derive expressions that relate the time- and angle-of-arrival of these multibounce returns to scattering points on the specular surface, and then use these expressions to formulate techniques for retrieving specular surface geometry when the scene is scanned by a single beam or illuminated with a multi-beam flash. We also consider the special case of transparent specular surfaces, for which surface reflections can be mixed together with light that scatters off of objects lying behind the surface.

Public health impact of delaying second dose of BNT162b2 or mRNA-1273 covid-19 vaccine: simulation agent based modeling study.

OBJECTIVE: To estimate population health outcomes with delayed second dose versus standard schedule of SARS-CoV-2 mRNA vaccination. DESIGN: Simulation agent based modeling study. SETTING: Simulated population based on real world US county. PARTICIPANTS: The simulation included 100 000 agents, with a representative distribution of demographics and occupations. Networks of contacts were established to simulate potentially infectious interactions though occupation, household, and random interactions. INTERVENTIONS: Simulation of standard covid-19 vaccination versus delayed second dose vaccination prioritizing the first dose. The simulation runs were replicated 10 times. Sensitivity analyses included first dose vaccine efficacy of 50%, 60%, 70%, 80%, and 90% after day 12 post-vaccination; vaccination rate of 0.1%, 0.3%, and 1% of population per day; assuming the vaccine prevents only symptoms but not asymptomatic spread (that is, non-sterilizing vaccine); and an alternative vaccination strategy that implements delayed second dose for people under 65 years of age, but not until all those above this age have been vaccinated. MAIN OUTCOME MEASURES: Cumulative covid-19 mortality, cumulative SARS-CoV-2 infections, and cumulative hospital admissions due to covid-19 over 180 days. RESULTS: Over all simulation replications, the median cumulative mortality per 100 000 for standard dosing versus delayed second dose was 226 v 179, 233 v 207, and 235 v 236 for 90%, 80%, and 70% first dose efficacy, respectively. The delayed second dose strategy was optimal for vaccine efficacies at or above 80% and vaccination rates at or below 0.3% of the population per day, under both sterilizing and non-sterilizing vaccine assumptions, resulting in absolute cumulative mortality reductions between 26 and 47 per 100 000. The delayed second dose strategy for people under 65 performed consistently well under all vaccination rates tested. CONCLUSIONS: A delayed second dose vaccination strategy, at least for people aged under 65, could result in reduced cumulative mortality under certain conditions.

Universal Shelter-in-Place Versus Advanced Automated Contact Tracing and Targeted Isolation: A Case for 21st-Century Technologies for SARS-CoV-2 and Future Pandemics.

OBJECTIVE: To model and compare effect of digital contact tracing versus shelter-in-place on severe acute respiratory syndrome - coronavirus 2 (SARS-CoV-2) spread. METHODS: Using a classical epidemiologic framework and parameters estimated from literature published between February 1, 2020, and May 25, 2020, we modeled two non-pharmacologic interventions - shelter-in-place and digital contact tracing - to curb spread of SARS-CoV-2. For contact tracing, we assumed an advanced automated contact tracing (AACT) application that sends alerts to individuals advising self-isolation based on individual exposure profile. Model parameters included percentage population ordered to shelter-in-place, adoption rate of AACT, and percentage individuals who appropriately follow recommendations. Under influence of these variables, the number of individuals infected, exposed, and isolated were estimated. RESULTS: Without any intervention, a high rate of infection (>10 million) with early peak is predicted. Shelter-in-place results in rapid decline in infection rate at the expense of impacting a large population segment. The AACT model achieves reduction in infected and exposed individuals similar to shelter-in-place without impacting a large number of individuals. For example, a 50% AACT adoption rate mimics a shelter-in-place order for 40% of the population and results in a greater than 90% decrease in peak number of infections. However, as compared to shelter-in-place, with AACT significantly fewer individuals would be isolated. CONCLUSION: Wide adoption of digital contact tracing can mitigate infection spread similar to universal shelter-in-place, but with considerably fewer individuals isolated.

Author Correction: Active Printed Materials for Complex Self-Evolving Deformations.

A correction has been published and is appended to both the HTML and PDF versions of this paper. The error has not been fixed in the paper.

Machine learning approaches for large scale classification of produce.

The analysis and identification of different attributes of produce such as taxonomy, vendor, and organic nature is vital to verifying product authenticity in a distribution network. Though a variety of analysis techniques have been studied in the past, we present a novel data-centric approach to classifying produce attributes. We employed visible and near infrared (NIR) spectroscopy on over 75,000 samples across several fruit and vegetable varieties. This yielded 0.90-0.98 and 0.98-0.99 classification accuracies for taxonomy and farmer classes, respectively. The most significant factors in the visible spectrum were variations in the produce color due to chlorophyll and anthocyanins. In the infrared spectrum, we observed that the varying water and sugar content levels were critical to obtaining high classification accuracies. High quality spectral data along with an optimal tuning of hyperparameters in the support vector machine (SVM) was also key to achieving high classification accuracies. In addition to demonstrating exceptional accuracies on test data, we explored insights behind the classifications, and identified the highest performing approaches using cross validation. We presented data collection guidelines, experimental design parameters, and machine learning optimization parameters for the replication of studies involving large sample sizes.

Terahertz scattering and water absorption for porosimetry.

We use terahertz transmission through limestone sedimentary rock samples to assess the macro and micro porosity. We exploit the notable water absorption in the terahertz spectrum to interact with the pores that are two orders of magnitude smaller (<1µm) than the terahertz wavelength. Terahertz water sensitivity provides us with the dehydration profile of the rock samples. The results show that there is a linear correlation between such a profile and the ratio of micro to macro porosity of the rock. Furthermore, this study estimates the absolute value of total porosity based on optical diffusion theory. We compare our results with that of mercury injection capillary pressure as a benchmark to confirm our analytic framework. The porosimetry method presented here sets a foundation for a new generation of less invasive porosimetry methods with higher penetration depth based on lower frequency (f<10THz) scattering and absorption. The technique has applications in geological studies and in other industries without the need for hazardous mercury or ionizing radiation.

Object classification through scattering media with deep learning on time resolved measurement.

We demonstrate an imaging technique that allows identification and classification of objects hidden behind scattering media and is invariant to changes in calibration parameters within a training range. Traditional techniques to image through scattering solve an inverse problem and are limited by the need to tune a forward model with multiple calibration parameters (like camera field of view, illumination position etc.). Instead of tuning a forward model and directly inverting the optical scattering, we use a data driven approach and leverage convolutional neural networks (CNN) to learn a model that is invariant to calibration parameters variations within the training range and nearly invariant beyond that. This effectively allows robust imaging through scattering conditions that is not sensitive to calibration. The CNN is trained with a large synthetic dataset generated with a Monte Carlo (MC) model that contains random realizations of major calibration parameters. The method is evaluated with a time-resolved camera and multiple experimental results are provided including pose estimation of a mannequin hidden behind a paper sheet with 23 correct classifications out of 30 tests in three poses (76.6% accuracy on real-world measurements). This approach paves the way towards real-time practical non line of sight (NLOS) imaging applications.

A method for rapid 3D scanning and replication of large paleontological specimens.

We demonstrate a fast and cost-effective technique to perform three dimensional (3D) scanning and replication of large paleontological specimens, in this case the entire skull of a Tyrannosaurus rex (T.rex) with a volume in the range of 2 m3. The technique involves time-of-flight (TOF) depth sensing using the Kinect scanning module commonly used in gesture recognition in gaming. Raw data from the Kinect sensor was captured using open source software and the reconstruction was done rapidly making this a viable method that can be adopted by museums and researchers in paleontology. The current method has the advantage of being low-cost as compared to industrial scanners and photogrammetric methods but also of accurately scanning a substantial volume range which is well suited for large specimens. The depth resolution from the Kinect sensor was measured to be around 0.6 mm which is ideal for scanning large specimens with reasonable structural detail. We demonstrate the efficacy of this method on the skull of FMNH PR 2081, also known as SUE, a near complete T.rex at the Field Museum of Natural History.

Computer vision uncovers predictors of physical urban change.

Which neighborhoods experience physical improvements? In this paper, we introduce a computer vision method to measure changes in the physical appearances of neighborhoods from time-series street-level imagery. We connect changes in the physical appearance of five US cities with economic and demographic data and find three factors that predict neighborhood improvement. First, neighborhoods that are densely populated by college-educated adults are more likely to experience physical improvements-an observation that is compatible with the economic literature linking human capital and local success. Second, neighborhoods with better initial appearances experience, on average, larger positive improvements-an observation that is consistent with "tipping" theories of urban change. Third, neighborhood improvement correlates positively with physical proximity to the central business district and to other physically attractive neighborhoods-an observation that is consistent with the "invasion" theories of urban sociology. Together, our results provide support for three classical theories of urban change and illustrate the value of using computer vision methods and street-level imagery to understand the physical dynamics of cities.

Automated retinal imaging and trend analysis - a tool for health monitoring.

Most current diagnostic devices are expensive, require trained specialists to operate and gather static images with sparse data points. This leads to preventable diseases going undetected until late stage, resulting in greatly narrowed treatment options. This is especially true for retinal imaging. Future solutions are low cost, portable, self-administered by the patient, and capable of providing multiple data points, population analysis, and trending. This enables preventative interventions through mass accessibility, constant monitoring, and predictive modeling.

All Photons Imaging Through Volumetric Scattering.

Imaging through thick highly scattering media (sample thickness ≫ mean free path) can realize broad applications in biomedical and industrial imaging as well as remote sensing. Here we propose a computational "All Photons Imaging" (API) framework that utilizes time-resolved measurement for imaging through thick volumetric scattering by using both early arrived (non-scattered) and diffused photons. As opposed to other methods which aim to lock on specific photons (coherent, ballistic, acoustically modulated, etc.), this framework aims to use all of the optical signal. Compared to conventional early photon measurements for imaging through a 15 mm tissue phantom, our method shows a two fold improvement in spatial resolution (4db increase in Peak SNR). This all optical, calibration-free framework enables widefield imaging through thick turbid media, and opens new avenues in non-invasive testing, analysis, and diagnosis.

Ultra-portable, wireless smartphone spectrometer for rapid, non-destructive testing of fruit ripeness.

We demonstrate a smartphone based spectrometer design that is standalone and supported on a wireless platform. The device is inherently low-cost and the power consumption is minimal making it portable to carry out a range of studies in the field. All essential components of the device like the light source, spectrometer, filters, microcontroller and wireless circuits have been assembled in a housing of dimensions 88 mm × 37 mm × 22 mm and the entire device weighs 48 g. The resolution of the spectrometer is 15 nm, delivering accurate and repeatable measurements. The device has a dedicated app interface on the smartphone to communicate, receive, plot and analyze spectral data. The performance of the smartphone spectrometer is comparable to existing bench-top spectrometers in terms of stability and wavelength resolution. Validations of the device were carried out by demonstrating non-destructive ripeness testing in fruit samples. Ultra-Violet (UV) fluorescence from Chlorophyll present in the skin was measured across various apple varieties during the ripening process and correlated with destructive firmness tests. A satisfactory agreement was observed between ripeness and fluorescence signals. This demonstration is a step towards possible consumer, bio-sensing and diagnostic applications that can be carried out in a rapid manner.

Terahertz time-gated spectral imaging for content extraction through layered structures.

Spatial resolution, spectral contrast and occlusion are three major bottlenecks for non-invasive inspection of complex samples with current imaging technologies. We exploit the sub-picosecond time resolution along with spectral resolution provided by terahertz time-domain spectroscopy to computationally extract occluding content from layers whose thicknesses are wavelength comparable. The method uses the statistics of the reflected terahertz electric field at subwavelength gaps to lock into each layer position and then uses a time-gated spectral kurtosis to tune to highest spectral contrast of the content on that specific layer. To demonstrate, occluding textual content was successfully extracted from a packed stack of paper pages down to nine pages without human supervision. The method provides over an order of magnitude enhancement in the signal contrast and can impact inspection of structural defects in wooden objects, plastic components, composites, drugs and especially cultural artefacts with subwavelength or wavelength comparable layers.

Volume estimation of tonsil phantoms using an oral camera with 3D imaging.

Three-dimensional (3D) visualization of oral cavity and oropharyngeal anatomy may play an important role in the evaluation for obstructive sleep apnea (OSA). Although computed tomography (CT) and magnetic resonance (MRI) imaging are capable of providing 3D anatomical descriptions, this type of technology is not readily available in a clinic setting. Current imaging of the oropharynx is performed using a light source and tongue depressors. For better assessment of the inferior pole of the tonsils and tongue base flexible laryngoscopes are required which only provide a two dimensional (2D) rendering. As a result, clinical diagnosis is generally subjective in tonsillar hypertrophy where current physical examination has limitations. In this report, we designed a hand held portable oral camera with 3D imaging capability to reconstruct the anatomy of the oropharynx in tonsillar hypertrophy where the tonsils get enlarged and can lead to increased airway resistance. We were able to precisely reconstruct the 3D shape of the tonsils and from that estimate airway obstruction percentage and volume of the tonsils in 3D printed realistic models. Our results correlate well with Brodsky's classification of tonsillar hypertrophy as well as intraoperative volume estimations.

Optical brush: Imaging through permuted probes.

The combination of computational techniques and ultrafast imaging have enabled sensing through unconventional settings such as around corners, and through diffusive media. We exploit time of flight (ToF) measurements to enable a flexible interface for imaging through permuted set of fibers. The fibers are randomly distributed in the scene and are packed on the camera end, thus making a brush-like structure. The scene is illuminated by two off-axis optical pulses. Temporal signatures of fiber tips in the scene are used to localize each fiber. Finally, by combining the position and measured intensity of each fiber, the original input is reconstructed. Unlike conventional fiber bundles with packed set of fibers that are limited by a narrow field of view (FOV), lack of flexibility, and extended coaxial precalibration, the proposed optical brush is flexible and uses off-axis calibration method based on ToF. The enabled brush form can couple to other types of ToF imaging systems. This can impact probe-based applications such as, endoscopy, tomography, and industrial imaging and sensing.

Identi-wheez - A device for in-home diagnosis of asthma.

Asthma is the most common chronic illness among children. The skills required to diagnose it make it an even greater concern. In this work, we present a child-friendly wearable device, which allows in-home diagnosis of asthma. The device acquires simultaneous measurements from multiple stethoscopes. The recordings are then sent to a specialist who uses assistive diagnosis algorithms that enable auscultation (listening to lung sounds with a stethoscope) at any location in the lungs volume by sound refocusing. The specialist is also presented with a sound "heat map" which shows the location of sound sources in the lungs. We present design considerations of our device, as well as the algorithms for assistive diagnosis and their analysis which demonstrate reduction of ambient and measurement noise by over 10dB.

Time-of-Flight Microwave Camera.

Microwaves can penetrate many obstructions that are opaque at visible wavelengths, however microwave imaging is challenging due to resolution limits associated with relatively small apertures and unrecoverable "stealth" regions due to the specularity of most objects at microwave frequencies. We demonstrate a multispectral time-of-flight microwave imaging system which overcomes these challenges with a large passive aperture to improve lateral resolution, multiple illumination points with a data fusion method to reduce stealth regions, and a frequency modulated continuous wave (FMCW) receiver to achieve depth resolution. The camera captures images with a resolution of 1.5 degrees, multispectral images across the X frequency band (8 GHz-12 GHz), and a time resolution of 200 ps (6 cm optical path in free space). Images are taken of objects in free space as well as behind drywall and plywood. This architecture allows "camera-like" behavior from a microwave imaging system and is practical for imaging everyday objects in the microwave spectrum.

Modeling "wiggling" as a multi-path interference problem in AMCW ToF imaging.

Amplitude modulated continuous wave time-of-flight range cameras suffer from an inherent depth measurement error due to aliasing of the emitted signal vs reference signal correlation function. This is due to higher harmonics present in both signals which are not accounted for in the model or measurements. This "wiggling" error is generally corrected by employing a correction function based on frequency and depth dependent calibration data. This problem is shown to be equivalent to a multi-path interference problem. Casting the problem into the multi-path interference domain and utilizing multiple modulation frequencies provides tools for dealing with the depth error without calibration in a frequency independent way.