Rapidly adaptable automated interpretation of point-of-care COVID-19 diagnostics.

BACKGROUND: Point-of-care diagnostic devices, such as lateral-flow assays, are becoming widely used by the public. However, efforts to ensure correct assay operation and result interpretation rely on hardware that cannot be easily scaled or image processing approaches requiring large training datasets, necessitating large numbers of tests and expert labeling with validated specimens for every new test kit format. METHODS: We developed a software architecture called AutoAdapt POC that integrates automated membrane extraction, self-supervised learning, and few-shot learning to automate the interpretation of POC diagnostic tests using smartphone cameras in a scalable manner. A base model pre-trained on a single LFA kit is adapted to five different COVID-19 tests (three antigen, two antibody) using just 20 labeled images. RESULTS: Here we show AutoAdapt POC to yield 99% to 100% accuracy over 726 tests (350 positive, 376 negative). In a COVID-19 drive-through study with 74 untrained users self-testing, 98% found image collection easy, and the rapidly adapted models achieved classification accuracies of 100% on both COVID-19 antigen and antibody test kits. Compared with traditional visual interpretation on 105 test kit results, the algorithm correctly identified 100% of images; without a false negative as interpreted by experts. Finally, compared to a traditional convolutional neural network trained on an HIV test kit, the algorithm showed high accuracy while requiring only 1/50th of the training images. CONCLUSIONS: The study demonstrates how rapid domain adaptation in machine learning can provide quality assurance, linkage to care, and public health tracking for untrained users across diverse POC diagnostic tests.

It can be difficult to correctly interpret the results of rapid diagnostic tests that give a visual readout, such as COVID rapid tests. We developed a computational algorithm to interpret rapid test results using an image taken by a smartphone camera. This algorithm can easily be adapted for use on results from different test kits. The algorithm was accurate at interpreting results obtained by members of the public using various COVID rapid tests and diagnostic tests with similar outputs used for other infections. The use of this algorithm should enable accurate interpretation of rapid diagnostic tests by members of the public and hence enable improved medical care.

Snapshots of nascent RNA reveal cell- and stimulus-specific responses to acute kidney injury.

The current strategy to detect acute injury of kidney tubular cells relies on changes in serum levels of creatinine. Yet serum creatinine (sCr) is a marker of both functional and pathological processes and does not adequately assay tubular injury. In addition, sCr may require days to reach diagnostic thresholds, yet tubular cells respond with programs of damage and repair within minutes or hours. To detect acute responses to clinically relevant stimuli, we created mice expressing Rosa26-floxed-stop uracil phosphoribosyltransferase (Uprt) and inoculated 4-thiouracil (4-TU) to tag nascent RNA at selected time points. Cre-driven 4-TU-tagged RNA was isolated from intact kidneys and demonstrated that volume depletion and ischemia induced different genetic programs in collecting ducts and intercalated cells. Even lineage-related cell types expressed different genes in response to the 2 stressors. TU tagging also demonstrated the transient nature of the responses. Because we placed Uprt in the ubiquitously active Rosa26 locus, nascent RNAs from many cell types can be tagged in vivo and their roles interrogated under various conditions. In short, 4-TU labeling identifies stimulus-specific, cell-specific, and time-dependent acute responses that are otherwise difficult to detect with other technologies and are entirely obscured when sCr is the sole metric of kidney damage.

Acceptability and Use of a Dual HIV/Syphilis Rapid Test and Accompanying Smartphone App to Facilitate Self- and Partner-Testing Among Cisgender Men and Transgender Women Who Have Sex with Men.

At home self- and partner-testing may reduce HIV and syphilis transmission by detecting undiagnosed infections. Forty-eight cisgender men and transgender women who men who have sex with men were given ten INSTI Multiplex kits and downloaded the SMARTtest app to facilitate self- and partner testing over the next three months. Thirty-seven (77%) participants self-tested using the INSTI (mean = 3.7 times, SD = 3.9); 26 (54%) tested partners (mean = 1.6 times, SD = 2.2). Participants liked the test for its ease of use, quick results, and dual HIV/syphilis testing but its blood-based nature hindered use with partners. Participants with reactive syphilis results always attributed them to a past infection and these results presented a challenge to testing with partners and the ability to accurately assess risk of infection. Most participants stated they would use the INSTI for self-testing (100%) and for partner-testing (89%). Acceptability of the SMARTtest app was high for functionality (M = 4.16 of max 5, SD = 0.85) and helpfulness (M = 6.12 of max 7, SD = 1.09). Participants often used the app as needed, eschewing its use if they felt comfortable conducting the test and interpreting its results. Seventy-eight percent would recommend the app to a friend. Availability of the INSTI Multiplex as a self-test with the accompanying SMARTtest app might increase frequency of HIV and syphilis testing, allowing for earlier detection of infection and reduced transmission.

RESUMEN: El uso de pruebas rápidas caseras con parejas y como auto-pruebas puede reducir la transmisión del VIH y la sifilis al detectar infecciones no diagnosticadas. Cuarenta y ocho hombres cisgénero y mujeres transgénero que tienen sexo con hombres recibieron diez kits del INSTI Multiplex y descargaron la aplicación SMARTtest para facilitar su uso con parejas y para auto-pruebas durante los próximos tres meses. Treinta y siete (77%) participantes se auto-testearon utilizando el INSTI (media = 3.7 veces, DE = 3.9); 26 (54%) testearon a sus parejas (media = 1.6 veces, DE = 2.2). A los participantes les gustó la prueba por su facilidad de uso, rapidez de los resultados y por ser una prueba dual de VIH/sífilis, pero al ser una prueba basada en sangre dificultó su uso con parejas. Los participantes con resultados de sífilis reactivos siempre atribuyeron éstos a una infección pasada y sus resultados presentaron un desafío para el uso de pruebas con parejas. La mayoría de los participantes afirmaron que utilizarían el INSTI como auto-pruebas (100%) y para testear a sus parejas (89%). La aceptabilidad de la aplicación SMARTtest fue alta para la funcionalidad (M = 4.16 de un máximo de 5, SD = 0.85) y utilidad (M = 6.13 de un máximo de 7, SD = 1.09). Los participantes solían utilizar la aplicación según fuera necesario, evitando su uso si se sentían cómodos realizando la prueba e interpretando sus resultados. El 78% recomendaría la aplicación a un amigo. La disponibilidad del INSTI Multiplex como auto-prueba con la aplicación SMARTtest podría aumentar la frecuencia de las pruebas de VIH y sífilis, lo que permite una detección más temprana de la infección y reduce la transmisión.

Point-of-care diagnostics: recent developments in a pandemic age.

In this review, we provide an overview of developments in point-of-care (POC) diagnostics during the COVID-19 pandemic. We review these advances within the framework of a holistic POC ecosystem, focusing on points of interest - both technological and non-technological - to POC researchers and test developers. Technologically, we review design choices in assay chemistry, microfluidics, and instrumentation towards nucleic acid and protein detection for severe acute respiratory coronavirus 2 (SARS-CoV-2), and away from the lab bench, developments that supported the unprecedented rapid development, scale up, and deployment of POC devices. We describe common features in the POC technologies that obtained Emergency Use Authorization (EUA) for nucleic acid, antigen, and antibody tests, and how these tests fit into four distinct POC use cases. We conclude with implications for future pandemics, infectious disease monitoring, and digital health.

Rule Out Acute Kidney Injury in the Emergency Department With a Urinary Dipstick.

INTRODUCTION: The identification of acute injury of the kidney relies on serum creatinine (SCr), a functional marker with poor temporal resolution as well as limited sensitivity and specificity for cellular injury. In contrast, urinary biomarkers of kidney injury have the potential to detect cellular stress and damage in real time. METHODS: To detect the response of the kidney to injury, we have tested a lateral flow dipstick that measures a urinary protein called neutrophil gelatinase-associated lipocalin (NGAL). Analysis of urine was performed in a prospective cohort of 479 patients (final cohort N = 426) entering an emergency department in New York City and subsequently admitted for inpatient care. RESULTS: Colorimetric development had high interrater reliability (88% concordance rate) and correlated with traditional enzyme-linked immunosorbent assay (ELISA) measurements (ρ = 0.732, P < .0001). Of the 14% of the cohort who met Acute Kidney Injury Network (AKIN) SCr criteria for acute kidney injury (AKI), 67% demonstrated transient (<2 days) and 33% demonstrated sustained (>2 days) elevation of SCr. Comparing the outcomes of patients with sustained versus transient or undetectable changes in SCr revealed that the urinary NGAL (uNGAL) dipstick had high specificity and negative predictive value (NPV) (high- vs. low-intermediate readings, sensitivity = 0.55, specificity = 0.91, positive predictive value = 0.24, NPV = 0.97, χ2 = 20.39, P < 0.001). CONCLUSION: We show that the introduction of a bedside uNGAL dipstick permits accurate triage by identifying individuals who do not have tubular injury. In an era of shortening length of stay and rapid decisions based on isolated SCr measurements, real-time exclusion of kidney injury by a dipstick will be particularly useful to overcome the retrospective, insensitive, and nonspecific attributes of SCr.

Biosensors for Personal Mobile Health: A System Architecture Perspective.

Advances in mobile biosensors, integrating developments in materials science and instrumentation, are fueling an expansion in health data being collected and analyzed in decentralized settings. For example, semiconductor-based sensors are enabling measurement of vital signs, and microfluidic-based sensors are enabling measurement of biochemical markers. As biosensors for mobile health are becoming increasingly paired with smart devices, it will become critical for researchers to design biosensors - with appropriate functionalities and specifications - to work seamlessly with accompanying connected hardware and software. This article describes recent research in biosensors, as well as current mobile health devices in use, as classified into four distinct system architectures that take into account the biosensing and data processing functions required in personal mobile health devices. We also discuss the path forward for integrating biosensors into smartphone-based mobile health devices.

SMARTtest: A Smartphone App to Facilitate HIV and Syphilis Self- and Partner-Testing, Interpretation of Results, and Linkage to Care.

Acceptability of rapid HIV self-testing is high but potential users remain concerned about correct use, interpretation of test results, and linkage to care. This article describes user preferences for a smartphone app to mitigate these challenges and how these were integrated into the SMARTtest app to support self- and partner-testing using the INSTI Multiplex®. Sixty men and transgender women who have sex with men self-tested for HIV and syphilis while guided by a prototype app that provided a video, pictorial step-by-step instructions, and sample test results presented textually ("positive," "negative"). Subsequently, participants provided feedback on revisions and additional app content. Participants recommended offering different user modes (self, partner, both), and retaining the video, step-by-step instructions, and textual test results. They strongly favored the ability to save and send test results to sexual partners or providers. These features were integrated into the SMARTtest app to facilitate HIV/syphilis self- and partner-testing, HIV/syphilis status awareness and disclosure, and linkage to care.

A Multiplexed Serologic Test for Diagnosis of Lyme Disease for Point-of-Care Use.

Single multiplexed assays could replace the standard 2-tiered (STT) algorithm recommended for the laboratory diagnosis of Lyme disease if they perform with a specificity and a sensitivity superior or equal to those of the STT algorithm. We used human serum rigorously characterized to be sera from patients with acute- and convalescent-phase early Lyme disease, Lyme arthritis, and posttreatment Lyme disease syndrome, as well as the necessary controls (n = 241 samples), to select the best of 12 Borrelia burgdorferi proteins to improve our microfluidic assay (mChip-Ld). We then evaluated its serodiagnostic performance in comparison to that of a first-tier enzyme immunoassay and the STT algorithm. We observed that more antigens became positive as Lyme disease progressed from early to late stages. We selected three antigens (3Ag) to include in the mChip-Ld: VlsE and a proprietary synthetic 33-mer peptide (PepVF) to capture sensitivity in all disease stages and OspC for early Lyme disease. With the specificity set at 95%, the sensitivity of the mChip-Ld with 3Ag ranged from 80% (95% confidence interval [CI], 56% to 94%) and 85% (95% CI, 74% to 96%) for two panels of serum from patients with early Lyme disease and was 100% (95% CI, 83% to 100%) for serum from patients with Lyme arthritis; the STT algorithm detected early Lyme disease in the same two panels of serum from patients with early Lyme disease with a sensitivity of 48.5% and 75% and Lyme arthritis in serum from patients with Lyme arthritis with a sensitivity of 100%, and the specificity was 97.5% to 100%. The mChip-Ld platform outperformed the STT algorithm according to sensitivity. These results open the door for the development of a single, rapid, multiplexed diagnostic test for point-of-care use that can be designed to identify the Lyme disease stage.

Integrating user behavior with engineering design of point-of-care diagnostic devices: theoretical framework and empirical findings.

With point-of-care (POC) diagnostic devices becoming increasingly available to untrained users, it will be critical to understand how real-world user behavior can best inform and guide the engineering design process. Social sciences present frameworks for analyzing user behavior, but they have not yet been applied to POC diagnostics in a methodical manner. Here, we develop a framework that synthesizes two models that can collectively account for user behavior and experience with POC diagnostic devices: a social psychological information-motivation-behavior (IMB) model (first described by Fisher and Fisher) for identifying determinants for health-related behavior, and user experience (UX) elements for studying interactions between users and products. Based on studies of 40 naïve users of our smartphone-enabled microfluidics device that can be used for HIV home-testing, we found that untrained participants could perform 90% of steps correctly, with engineering design elements that provided feedback that was either direct (e.g., a light or click) or binary (e.g., a switch) enhancing usability. Interestingly, of the steps performed incorrectly, over 70% were due not to errors in the device or user operation, but user-to-user variability (e.g. time in collecting fingerstick and force applied to initiate vacuum), which could be addressed by further modifications to the device. Overall, this study suggests that microfluidic POC HIV home-testing is likely to benefit from smartphone integration, and that engineering design of POC diagnostic devices can benefit from a structured evaluation of user behavior and experience, as guided by a social-psychological framework, which emphasizes user credibility, accessibility, acceptability, usability, and value.

Comparative review of interferometric detection of plasmonic nanoparticles.

Noble metal nanoparticles exhibit enhanced scattering and absorption at specific wavelengths due to a localized surface plamson resonance. This unique property can be exploited to enable the use of plasmonic nanoparticles as contrast agents in optical imaging. A range of optical techniques have been developed to detect nanoparticles in order to implement imaging schemes. Here we review several different approaches for using optical interferometry to detect the presence and concentration of nanoparticles. The strengths and weaknesses of the various approaches are discussed and quantitative comparisons of the achievable signal to noise ratios are presented. The benefits of each approach are outlined as they relate to specific application goals.