Joshua Balsam

Low-cost technologies for medical diagnostics in low-resource settings

INTRODUCTION Medical diagnostics is a critical element of effective medical treatment. However, many modern and emerging diagnostic technologies are not affordable or compatible with the needs and conditions found in low- and middle-income countries. Resource-poor countries require low-cost, robust, easy-to-use, and portable diagnostic devices compatible with telemedicine that can be adapted to meet diverse medical needs. AREAS COVERED The most suitable devices are likely those that will be based on optical technologies, which are used for many types of biological analyses. This manuscript describes several prototypes of low-cost optical technologies and their application developed at the FDAs Office of Science and Engineering laboratories including a webcam-based multiwavelength fluorescence plate reader, a webcam-based fluorescence microscope demonstrated for colonic mucosa tissue pathology analysis, a lens-free optical detector used for the detection of Botulinum A neurotoxin activity, and a lab-on-a-chip which enables the performance of enzyme-linked immunosorbent assay and other immunological or enzymatic assays without the need of dedicated laboratories and complex equipment demonstrated for the detection of the toxin staphylococcal enterotoxin B. EXPERT OPINION Sensitive and effective optical detection devices can be developed using readily available consumer electronics components such as webcams, charge-coupled device cameras, and LEDs. There are challenges in developing devices with sufficient sensitivity and specificity. Several optical and computational approaches were developed to overcome these challenges to create optical detectors that can serve as low-cost medical diagnostics in resource-poor settings.

Thousand-fold fluorescent signal amplification for mHealth diagnostics

The low sensitivity of Mobile Health (mHealth) optical detectors, such as those found on mobile phones, is a limiting factor for many mHealth clinical applications. To improve sensitivity, we have combined two approaches for optical signal amplification: (1) a computational approach based on an image stacking algorithm to decrease the image noise and enhance weak signals, and (2) an optical signal amplifier utilizing a capillary tube array. These approaches were used in a detection system which includes multi-wavelength LEDs capable of exciting many fluorophores in multiple wavelengths, a mobile phone or a webcam as a detector, and capillary tube array configured with 36 capillary tubes for signal enhancement. The capillary array enables a ~100× increase in signal sensitivity for fluorescein, reducing the limit of detection (LOD) for mobile phones and webcams from 1000 nM to 10nM. Computational image stacking enables another ~10× increase in signal sensitivity, further reducing the LOD for webcam from 10nM to 1 nM. To demonstrate the feasibility of the device for the detection of disease-related biomarkers, adenovirus DNA labeled with SYBR green or fluorescein was analyzed by both our capillary array and a commercial plate reader. The LOD for the capillary array was 5 ug/mL, and that of the plate reader was 1 ug/mL. Similar results were obtained using DNA stained with fluorescein. The combination of the two signal amplification approaches enables a ~1000× increase in LOD for the webcam platform. This brings it into the range of a conventional plate reader while using a smaller sample volume (10 ul) than the plate reader requires (100 ul). This suggests that such a device could be suitable for biosensing applications where up to 10 fold smaller sample sizes are needed. The simple optical configuration for mHealth described in this paper employing the combined capillary and image processing signal amplification is capable of measuring weak fluorescent signals without the need of dedicated laboratories. It has the potential to be used to increase sensitivity of other optically based mHealth technologies, and may increase mHealths clinical utility, especially for telemedicine and for resource-poor settings and global health applications.

Improving the Sensitivity and Functionality of Mobile Webcam-Based Fluorescence Detectors for Point-of-Care Diagnostics in Global Health

Resource-poor countries and regions require effective, low-cost diagnostic devices for accurate identification and diagnosis of health conditions. Optical detection technologies used for many types of biological and clinical analysis can play a significant role in addressing this need, but must be sufficiently affordable and portable for use in global health settings. Most current clinical optical imaging technologies are accurate and sensitive, but also expensive and difficult to adapt for use in these settings. These challenges can be mitigated by taking advantage of affordable consumer electronics mobile devices such as webcams, mobile phones, charge-coupled device (CCD) cameras, lasers, and LEDs. Low-cost, portable multi-wavelength fluorescence plate readers have been developed for many applications including detection of microbial toxins such as C. Botulinum A neurotoxin, Shiga toxin, and S. aureus enterotoxin B (SEB), and flow cytometry has been used to detect very low cell concentrations. However, the relatively low sensitivities of these devices limit their clinical utility. We have developed several approaches to improve their sensitivity presented here for webcam based fluorescence detectors, including (1) image stacking to improve signal-to-noise ratios; (2) lasers to enable fluorescence excitation for flow cytometry; and (3) streak imaging to capture the trajectory of a single cell, enabling imaging sensors with high noise levels to detect rare cell events. These approaches can also help to overcome some of the limitations of other low-cost optical detection technologies such as CCD or phone-based detectors (like high noise levels or low sensitivities), and provide for their use in low-cost medical diagnostics in resource-poor settings.

Orthographic Projection Capillary Array Fluorescent Sensor for mHealth

To overcome the limited sensitivity of phone cameras for mobile health (mHealth) fluorescent detection, we have previously developed a capillary array which enables a ∼100 × increase in detection sensitivity. However, for an effective detection platform, the optical configuration must allow for uniform measurement sensitivity between channels when using such a capillary array sensor. This is a challenge due to the parallax inherent in imaging long parallel capillary tubes with typical lens configurations. To enable effective detection, we have developed an orthographic projection system in this work which forms parallel light projection images from the capillaries using an object-space telecentric lens configuration. This optical configuration results in a significantly higher degree of uniformity in measurement between channels, as well as a significantly reduced focal distance, which enables a more compact sensor. A plano-convex lens (f=150 mm) was shown to produce a uniform orthographic projection when properly combined with the phone cameras built in lens (f=4mm), enabling measurements of long capillaries (125 mm) to be made from a distance of 160 mm. The number of parallel measurements which can be made is determined by the size of the secondary lens. Based on these results, a more compact configuration with shorter 32 mm capillaries and a plano-convex lens with a shorter focal length (f=10mm) was constructed. This optical system was used to measure serial dilutions of fluorescein with a limit of detection (LOD) of 10nM, similar to the LOD of a commercial plate reader. However, many plate readers based on standard 96 well plate requires sample volumes of 100 μl for measurement, while the capillary array requires a sample volume of less than 10 μl. This optical configuration allows for a device to make use of the ∼100 × increase in fluorescent detection sensitivity produced by capillary amplification while maintaining a compact size and capability to analyze extremely small sample volumes. Such a device based on a phone or other optical mHealth technology will have the sensitivity of a conventional plate reader but have greater mHealth clinical utility, especially for telemedicine and for resource-poor settings and global health applications.

Smartphone-Based Fluorescence Detector for mHealth

We describe here a compact smartphone-based fluorescence detector for mHealth. A key element to achieving high sensitivity using low sensitivity phone cameras is a capillary array, which increases sensitivity by 100×. The capillary array was combined with a white LED illumination system to enable wide spectra fluorescent excitation in the range of 450-740 nm. The detector utilizes an orthographic projection system to form parallel light projection images from the capillaries at a close distance via an object-space telecentric lens configuration that reduces the total lens-to-object distance while maintaining uniformity in measurement between capillaries. To further increase the limit of detection (LOD), a computational image processing approach was employed to decrease the level of noise. This enables an additional 5-10× decrease in LOD. This smartphone-based detector was used to measure serial dilutions of fluorescein with a LOD of 1 nM with image stacking and 10 nM without image stacking, similar to the LOD obtained with a commercial plate reader. Moreover, the capillary array required a sample volume of less than 10 μl, which is an order of magnitude less than the 100 μl required for the plate reader.As fluorescence detection is widely used in sensitive biomedical assays, the approach described here has the potential to increase mHealth clinical utility, especially for telemedicine and for resource-poor settings in global health applications.

Effect of oxygen environment on formation of modulated Ag nanostructures along the interface of a Ag-Si heterostructure

Ag nanostructures were formed along the interface between Ag foil and Si wafer. These structures form when heated in the presence of air below the melting temperature of Ag, while heating above the melting temperature results in Ag foil melting with no nanostructure growth. Nanostructure morphology is determined by Si wafer orientation, while the size and periodicity of the structures depend on the duration of thermal processing. Electron backscatter diffraction (EBSD) analysis shows an epitaxial relation between Ag nanostructures and Si. Annealing specimens in vacuum above the Ag-Si eutectic but below pure Ag melting temperature results in solidification of melted foil without any interfacial nanostructure growth.

Evaluation of a Methodology for Automated Cell Counting for Streak ModeImaging Flow Cytometry

Identification of Circulating Tumor Cells (CTCs) has shown promising clinical applications, but since CTCs are found in very low concentration in blood large sample volumes are needed for meaningful enumeration. This issue impedes the analysis of CTCs using standard flow cytometry due to its low throughput. To address this issue, a high throughput microfluidic cytometer was recently developed using a wide field flow- flow cell instead of the conventional narrow hydrodynamic focusing cells (used in traditional flow cytometry) enabling analysis of large volumes at lower flow rate. This wide-field flow cytometer adopts a technique known as “streak photography” where exposure times and flow velocities are set such that the particles are imaged as short “streaks”. Since streaks are imaged with large number of pixels, they are easily distinguished from the noise which appears as “speckles” increasing the detection capabilities of the device, making it more suitable for analysis using current low sensitivity, high noise webcams or mobile phone cameras. The non-stationary nature of the high noisy background found in streak cytometry introduces additional challenges for automated cell counting methods using traditional cell detection techniques such TLC, CellProfiler, CellTracker and other tools based in traditional edge detection (e.g., Canny based filters) or manual thresholding. In order to address this issue, we developed a new automated enumeration approach that does not rely on edge detection or manual thresholding of individual cells, rather is based in image quantizing, morphological operations, 2D order-statistic filtering and decisions rules that take into account knowledge of the structure and expected location of the streaks in consecutive frames. We evaluated our approach comparing it with two current methods representing the major computational modalities for cell detection: CellTrack (based in edge detection) and MTrack2 (based in manual thresholding). Samples of 1 cell/mL nominal concentration were analyzed in batch size of 30 mL at flow rate of 10 mL/min and imaged at 4 frames per second (fps), the files were saved in uncompressed AVI format files. The cells were annotated and the signal to noise ratio (SNR) was calculated. For samples with average SNR greater than 4.4 dB, our method achieved a sensitivity of 91% compared to CellTrack (60%) and MTrack2 (71%). The True Positive Rate (TPR) of cells detected was 0.93 for our method compared with 0.80 for Mtrack2 and 0.29 for CellTrack. This demonstrated the ability of the algorithm to count rare cells with high accuracy for concentrations of 1 cell/mL with SNR greater than 4.4 dB. This cell counting capability will enable to automate low cost imaging flow cytometry based on CCD detector and the expansion of cell-based clinical diagnostics in resource-poor settings.