M.H.P. gratefully acknowledges the scholarship research fund from Universitas Islam Indonesia (UII). The authors extend their gratitude to Mr. Nutchanon Ninyawee for his valuable expertise and assistance throughout the development of the mobile application and his contributions to the manuscript. Furthermore, the authors appreciate the financial support provided by Thailand Science Research and Innovation (TSRI), Basic Research Fund: Fiscal year 2023 (project no. FRB660073/0164) under Program Smart Healthcare of King Mongkut&#39;s University of Technology Thonburi.

The Ethics Committee of the Institutional Review Board, Royal Thai Army Medical Department, Phramongkutklao Hospital, Bangkok, Thailand (IRBRTA 1218/2562) granted approval. In carrying out this study, we complied with all necessary ethical regulations.
1. Device fabrication of the paper-based Immunoassay
NOTE: The paper-based immunoassay device was fabricated following previously established methods36,37, and Thai patent request no. 19010081638.
<ol>
	<li>Design and pattern drawing: Design the paper analytical device (Figure 1A,B) with 18 PAD wax patterns on a computer. 
	NOTE: The design is specific and intended for A5-sized paper. The number of PADs is related to the size of the paper, as the user requires.</li>
	<li>Print the designed pattern onto the cellulose paper using a wax printer (Table of Materials).</li>
	<li>Melt the wax-printed paper in a laboratory oven for 75 s at 150 &#176;C. Subsequently, store it in a silica box until needed for the subsequent steps.</li>
	<li>Apply 0.5 &#181;L of 0.025% poly-L-lysine (PLL) to both the test and control areas. Incubate at room temperature (RT) for 2 min in a silica box and then heat in the oven at 65 &#176;C for 5 min.</li>
	<li>Apply 0.5 &#181;L of 1 &#181;g &#181;L-1 of goat anti-mouse IgG antibody on the control area and 0.5 &#181;L of 1 &#181;g &#181;L-1 of&#160;the capture antibody to the test area. Allow the drops to dry in a silica gel box at RT for 30 min.</li>
	<li>Apply 2 &#181;L of the blocking buffer to the sample area, 3 &#181;L to the conjugate area, and 2 &#181;L to the detection area. Let the drops dry at RT in a silica gel box for 30 min.</li>
	<li>Apply 2 &#181;L of gold nanoparticle-antibody complex (AuNPs-Ab) solution to the conjugate area and allow it to dry in a silica gel box at RT for 30 min.</li>
</ol>
2. Assembly of the paper-based Immunoassay
<ol>
	<li>Carefully remove the protective film on the reverse side of the adhesive plastic backing card to expose the adhesive.</li>
	<li>Align the treated cellulose paper with the adhesive plastic backing card and firmly press the two layers together. 
	NOTE: Avoid touching the hydrophilic field to minimize the risk of contamination or damage to the device.</li>
	<li>Apply a plastic film to coat the paper and press them together.</li>
	<li>Cut the desired piece of devices using scissors from sheets of completely assembled devices.</li>
	<li>The DEN-NS1-PADs (Figure 1C) are now ready for use. For long-term stability, store them at 4 &#176;C.</li>
</ol>
3. Preparation of the AuNPs-Ab conjugate
NOTE: The AuNPs-Ab was prepared as described previously by Prabowo et al.36.
<ol>
	<li>Combine 10 &#181;L of 1 mg mL&#8722;1 anti-NS1 antibody in PBS, 1 mL of 40 nm AuNPs colloid, and 0.1 mL of 0.1 M borate buffer (pH 8.5).</li>
	<li>Rotate the mixture at 50 rpm for 60 min and incubate at RT.</li>
	<li>Apply 0.1 mL of 10 mg mL&#8722;1 BSA in BBS, rotate at 50 rpm, and incubate at RT for 15 min.</li>
	<li>Centrifuge the solution at 20,187 &#215; g and 4 &#176;C for 30 min.</li>
	<li>Carefully pipette and separate the supernatant from the precipitated AuNPs-Ab.</li>
	<li>Resuspend the AuNPs-Ab in 500 &#181;L of BBS and disperse it using sonication.</li>
	<li>Repeat centrifugation at 20,187 &#215; g and 4 &#176;C for 30 min. 
	NOTE: Repeat the dispersion and centrifugation processes 3x.</li>
	<li>Add 50 &#181;L of the conjugate buffer to the suspension, making it ready for application onto the conjugate area.</li>
</ol>
4. Mobile application development
<ol>
	<li>Image processing and machine learning development
	<ol>
		<li>Gather a dataset for a supervised image model by collecting over 900 auto-focusing images of DEN-NS1-PADs, capturing various conditions such as different concentrations, camera brands (12-13 megapixels), rotations (90&#176; and 180&#176;), and lighting settings. Aim for 30 images under each specific condition.</li>
		<li>Label the ground truth by Identifying and annotating two regions of interest as the test and control areas within the collected images for supervised learning.</li>
		<li>Design an algorithm to identify the background strip. Locate the centerline between the test and control regions, calculate its midpoint, and establish a square region proportional to the average size of the two main regions while maintaining the same rotational orientation.</li>
		<li>Create an image segmentation model using the dataset and ground truth labels from steps 4.1.1 and 4.1.2 to train an image segmentation model for identifying the regions of interest.</li>
	</ol>
	</li>
	<li>Application algorithm
	<ol>
		<li>Apply the trained image segmentation model to new images to locate the test, control, and background regions automatically.</li>
		<li>Use basic image processing techniques to obtain a single intensity value for each of the three regions of interest (test, control, and background).</li>
		<li>Transform the image into a 3D array representation (y, x channel) to access pixel values.</li>
		<li>Convert the image to grayscale by averaging RGB values and apply inversion with the formula (255-x).</li>
		<li>Normalize the values of the test and control area by subtracting the background area value.</li>
		<li>Use the preestablished calibration curve to calculate the concentration of NS1.</li>
		<li>Classify the results as positive or negative based on a cut-off value of 0.1103 derived from the normalized grayscale intensities37.</li>
	</ol>
	</li>
</ol>
5. Calibration curve and sensitivities
<ol>
	<li>Prepare NS1 sample in human serum for calibration with concentrations of 0, 0.1, 0.2, 0.4, 0.6, 0.8, and 1.0 &#181;g mL-1.</li>
	<li>Drop 50 &#181;L of each concentration onto the sample area and perform measurements in triplicate.</li>
	<li>Allow the samples to wick completely into the device, which may take 20-30 min to obtain results.</li>
	<li>Capture images of the device using a digital camera or smartphone after 5 min of incubation.</li>
	<li>Analyze the test and control areas using ImageJ and a custom mobile application.</li>
	<li>Construct the calibration curve based on data from ImageJ and the mobile application.</li>
	<li>Calculate the limit of blank (LOB), the limit of detection (LOD), and the limit of quantification (LOQ) using equations (1-3) below: 
	LOB = Mean of the blank data + 1:645* &#240; (standard deviation of blank data) (1) 
	LOD = LOB +1:645*&#240; (standard deviation of the lowest concentration data) (2) 
	LOQ = Mean of blank data + 10*&#240; (standard deviation of blank data) (3)</li>
</ol>
6. Performing a paper-based Immunoassay with clinical samples
<ol>
	<li>Collect and process 300 &#181;L of peripheral blood from 30 patients on the first day of hospitalization into purple top EDTA tubes, following good clinical practices.</li>
	<li>Centrifuge the blood at 2,884 &#215; g and 4 &#176;C for 20 min.</li>
	<li>Transfer the liquid component (plasma) into a clean polypropylene tube using a pipette.</li>
	<li>Store the plasma in the freezer immediately at &#8722;20 &#176;C for subsequent analysis.</li>
	<li>Apply 20 &#181;L of plasma to the sample area on top of the device. Then, add 30 &#181;L of wash buffer (0.05% v/v Tween 20 in 1x phosphate buffered saline).</li>
	<li>Allow the sample to wick completely into the device, which may take 20-30 min to obtain the results.</li>
	<li>Capture images of the device using a digital camera or smartphone after 5 min of incubation at room temperature.</li>
	<li>Analyze the test and control areas using ImageJ and a custom mobile application.</li>
</ol>
7. Quantification with mobile application
NOTE: The intensity of paper-based immunoassay is analyzed in the mobile application (Figure 2).
<ol>
	<li>Open the developed mobile application on the smartphone.</li>
	<li>Select Use camera or Upload from Gallery to choose or upload the data source. Do this through camera capturing or by selecting an image from the device&#39;s gallery.</li>
	<li>Navigate to the analytic section and touch the Analyze button on the screen.</li>
	<li>Wait for the application to analyze the data and display the results.</li>
</ol>

Smartphone-Integrated Dengue NS1 Paper-Based Diagnostic Device

The authors have no conflicts of interest to disclose.

One of the important design parameters for a smartphone-based reader system is the ability to provide reproducible imaging processing of samples. In this study, for simplicity and convenience, the images were captured from three different smartphone brands with 12-13 MP cameras without using an imaging box or accessories. Variable conditions of image capturing, such as the resolution of the camera, image capturing time, lighting conditions, and environment, can influence the color intensity of the test and control spots ...

Dengue virus (DENV) infection is the fastest-spreading mosquito-borne disease1, and more than 390 million people are infected with 96 million symptomatic infections, 2 million cases of severe disease, and more than 25,000 deaths per year occur in the world1,2. According to the World Health Organization (WHO), an estimated 3.9 billion people are at risk for dengue; ~70% live in Asia Pacific countries and mainly in Southeast Asia3. In 2019, the number of dengue cases reported to WHO was 4.2 million, and Thailand contributed at least 136,000 dengue cases and 144 death cases from dengue infection4. The dengue outbreak in Thailand occurs during the rainy season, from April to December, in both urban and rural areas, especially in the northeastern area.
DENV infections have different clinical manifestations ranging from subclinical symptoms, mild dengue fever (DF) to severe dengue hemorrhagic fever (DHF). The main characteristic of severe DHF condition is increased vascular permeability followed by shock and organ dysfunction1. Understanding the molecular pathway that can cause the vascular leak is very important in developing effective dengue treatments. Dengue non-structural protein 1 (NS1) is a secreted glycoprotein during early virus infection5,6, and it functions as a cofactor for viral RNA replication7. NS1 can trigger cytokine release and contribute to vascular leak by binding to toll-like receptor 4 (TLR4) and endothelial glycocalyx8,9. In vitro research has shown that NS1 interacts with endothelial cells and induces apoptosis. This condition can contribute to endothelial dysfunction and vascular leak10. NS1 antigen levels, correlated with serum Interleukin (IL)-10 levels, were increased significantly in patients with severe clinical disease11. Dengue NS1 also contributes to disease pathogenesis by inducing IL-10 and suppressing DENV-specific T-cell responses12,13. In addition, dengue NS1 protein was related to severe clinical disease, and the concentration of NS1 &#62; 600 ng mL-1 in the first 3 days of illness was associated with the development of DHF14.
The persistence of the dengue NS1 antigen in patients with DHF could be used as a marker of severe dengue6. There are several methods to detect NS1 in clinical samples such as enzyme-linked immunosorbent assay (ELISA)and the rapid test15. The gold standard for measuring the concentration of NS1 proteins in a clinical setting is the ELISA method. However, the ELISA method is expensive and requires skilled personnel, and laboratory facilities16. Therefore, the development of technology for detecting and quantifying NS1 proteins in the point-of-care test (POCT) is still ongoing. In the last decade, paper-based immunoassays such as lateral flow assays (LFAs) and microfluidic paper-based analytical devices (&#181;PADs) have become popular as diagnostic tests because of their simplicity, rapidity, inexpensiveness, and specificity17,18,19. In a paper-based immunoassay, several labels have been used to generate signals, such as gold nanoparticles (AuNPs)20, magnetic nanoparticles21,22, quantum dots23, and fluorescence materials24,25. AuNPs are the most common labels used in paper-based immunoassays due to their inexpensive cost of production, ease of manufacture, stability, and simple readout. Currently, lateral flow assays (LFAs) for dengue NS1 are famously used in the clinical setting26,27. However, conventional LFA label detection commonly uses the naked eye and only provides qualitative results.
In the last decade, more than 5 billion smartphones have been widely used globally, and there is potential for developing portable detection28,29. Smartphones have multi-functional capacities such as built-in physical sensors, multi-core processors, digital cameras, USB ports, audio ports, wireless, and application software, making them suitable for use in various biosensor platforms30. In addition, wireless technologies allow data to be sent quickly and can be used for real-time and on-site monitoring31. Mudanyali et al. combined the paper-based immunoassay and smartphones to develop a portable, equipment-free, rapid, low-cost, and user-friendly POCT platform for malaria, tuberculosis, and HIV32. Ling et al. reported a lateral flow assay combined with a smartphone camera to detect alkaline phosphatase activity in milk quantitatively33. Hou et al. also developed a smartphone-based, dual-modality imaging system for quantitative signals from color or fluorescence in the lateral flow assay34. In addition, using the smartphone as a colorimetric and quantitative reader can improve the sensitivity while the naked eye cannot confidently report the presence of the target35.
Presenting a breakthrough in dengue diagnostics, the DEN-NS1-PAD36,37,38 (referred to as the device henceforth) offers a portable and efficient solution. Using wax-printed microfluidic paper-based technology, this device quantifies NS1 with high sensitivity and specificity through image processing. To further enhance its utility, we have developed a user-friendly smartphone app for colorimetric and quantitative reading. Clinical validation using patient samples from Thai hospitals underscores its immediate impact on real-time patient assessment. Our innovation marks a pivotal advancement in streamlined, point-of-care-dengue management, promising to revolutionize diagnostics in resource-limited healthcare landscapes.

<table border="1" fo:keep-together.within-page="1" fo:keep-with-next.within-page="always">
	<tbody>
		<tr>
			<td>Materials</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>0.1 M phosphate-buffered saline (PBS, pH 7.2)&#160;</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>BBS containing 0.1% Tween 20, 10% sucrose, and 1% casein&#160;</td>
			<td></td>
			<td></td>
			<td>&#160;the conjugate area treatment and blocking buffer</td>
		</tr>
		<tr>
			<td>Borate buffered saline (BBS) (25 mM sodium borate and 150 mM sodium chloride at pH 8.2) supplemented with 1% BSA&#160;</td>
			<td></td>
			<td></td>
			<td>the washing buffer during the conjugation process AuNPs with the antibody</td>
		</tr>
		<tr>
			<td>Boric acid</td>
			<td>Merck</td>
			<td>10043-35-3</td>
			<td></td>
		</tr>
		<tr>
			<td>Bovine serum albumin fraction V (BSA)&#160;&#160;</td>
			<td>PAA Lab GmbH (Germany)</td>
			<td>K41-001&#160;</td>
			<td></td>
		</tr>
		<tr>
			<td>Casein</td>
			<td>Merck</td>
			<td>9005-46-3</td>
			<td></td>
		</tr>
		<tr>
			<td>Chromatography paper Grade 2&#160;</td>
			<td>GE Healthcare</td>
			<td>3002-911&#160;</td>
			<td></td>
		</tr>
		<tr>
			<td>Clear laminate film</td>
			<td>3M (Stationery shops)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Disodium hydrogen phosphate</td>
			<td>Merck</td>
			<td>7558-79-4</td>
			<td></td>
		</tr>
		<tr>
			<td>Double tape side</td>
			<td>Stationery shops</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Goat anti-mouse IgG antibody&#160;</td>
			<td>MyBiosource (USA)</td>
			<td>MBS435013</td>
			<td></td>
		</tr>
		<tr>
			<td>Gold nanoparticles (40 nm)&#160;&#160;</td>
			<td>Serve Science Co., Ltd. (Thailand)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Human IgG polyclonal antibody&#160;&#160;</td>
			<td>Merck</td>
			<td>AG711-M</td>
			<td></td>
		</tr>
		<tr>
			<td>Mouse dengue NS1 monoclonal antibody&#160;</td>
			<td>MyBiosource (USA)</td>
			<td>MBS834415</td>
			<td></td>
		</tr>
		<tr>
			<td>Mouse dengue NS1 monoclonal antibody&#160;</td>
			<td>MyBiosource (USA)</td>
			<td>MBS834236</td>
			<td></td>
		</tr>
		<tr>
			<td>NS1 serotype 2 antigens</td>
			<td>MyBiosource (USA)</td>
			<td>MBS 568697</td>
			<td></td>
		</tr>
		<tr>
			<td>PBS 1X containing 0.1% Tween 20 was used as t</td>
			<td></td>
			<td></td>
			<td>elution buffer</td>
		</tr>
		<tr>
			<td>Plastic backing card 10x30 cm</td>
			<td>Pacific Biotech Co., Ltd. (Thailand)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Poly-L-lysine (PLL)</td>
			<td>Sigma Aldrich</td>
			<td>P4832</td>
			<td></td>
		</tr>
		<tr>
			<td>Potassium Chloride</td>
			<td>Merck</td>
			<td>104936</td>
			<td></td>
		</tr>
		<tr>
			<td>Potassium monophosphate</td>
			<td>Merck</td>
			<td>104877</td>
			<td></td>
		</tr>
		<tr>
			<td>Sodium Chloride</td>
			<td>Merck</td>
			<td>7647-14-5</td>
			<td></td>
		</tr>
		<tr>
			<td>Sodium tetraborate&#160;</td>
			<td>Sigma Aldrich</td>
			<td>1303-96-4</td>
			<td></td>
		</tr>
		<tr>
			<td>Sucrose</td>
			<td>Merck</td>
			<td>57-50-1</td>
			<td></td>
		</tr>
		<tr>
			<td>Tween 20</td>
			<td>Sigma Aldrich</td>
			<td>9005-64-5</td>
			<td></td>
		</tr>
		<tr>
			<td>Instruments</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>CytationTM 5 multimode reader</td>
			<td>BioTek</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Mobile phones</td>
			<td>Huawei Y7, iPhone 11, Samsung a20</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Photo scanner</td>
			<td>Epson Perfection V30</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Oven</td>
			<td>Memmert</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Wax printer&#160;</td>
			<td>Xerox ColorQube 8880-PS</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Software</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Could AutoML Vision Object Detection documentation</td>
			<td>Google Cloud</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>ImageJ</td>
			<td>National Institute of Health, Bethesda, MD, USA</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Inkscape 0.91 Software</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
	</tbody>
</table>

portable paper-based immunoassay combined with smartphone application for colorimetric and quantitative detection of dengue ns1 antigen

Dengue virus (DENV) infection, which is transmitted by Aedes mosquitoes, is a major public health concern in tropical and subtropical countries. With an annual incidence of approximately 10 million cases and 20,000-25,000 deaths, particularly among children, there is an urgent need for practical diagnostic tools. The presence of dengue non-structural protein 1 (NS1) during early infection has been linked to cytokine release, vascular leakage, and endothelial dysfunction, making it a potential marker for severe dengue. 
Paper-based immunoassays such as lateral flow assays (LFAs) and microfluidic paper-based analytical devices (PADs) have gained popularity as diagnostic tests due to their simplicity, rapidity, inexpensiveness, specificity, and ease of interpretation. However, conventional paper-based immunoassays for dengue NS1 detection typically rely on visual inspection, yielding only qualitative results. To address this limitation and enhance sensitivity, we proposed a highly portable NS1 dengue detection assay on a Paper-based Analytical Device (PAD), namely, DEN-NS1-PAD, that integrates a smartphone application as a colorimetric and quantitative reader. The development system enables direct quantification of NS1 concentrations in clinical samples. 
Serum and blood samples obtained from patients were utilized to demonstrate the system prototype performance. The results were obtained immediately and can be employed for clinical assessment, both in well-equipped healthcare facilities and resource-limited settings. This innovative combination of a paper-based immunoassay with a smartphone application offers a promising approach for enhanced detection and quantification of dengue NS1 antigen. By augmenting sensitivity beyond the capabilities of the naked eye, this system holds great potential for improving clinical decision-making in dengue management, particularly in remote or underserved areas.

Author Spotlight: Development of a Smartphone-Enhanced Paper-Based Device for Rapid Dengue NS1 Detection

Portable Paper-Based Immunoassay Combined with Smartphone Application for Colorimetric and Quantitative Detection of Dengue NS1 Antigen

The scope of the research is detecting dengue using a specific biomarker NS1, with a paper-based analytical device. Our research focus on developing a simple, cheap, sensitive, and specific device to detect dengue. We also develop smartphone application to improve sensitivity and measurement of NS1 in real sample.Our sensitivity is developing a method to fabricate microfluidic method based on our coning analysis called PEC. Using formula is in our team, were some high techniques, aimed to embody commercialization of various PEC designs, offering scalable solutions for more supply applications. Our research combines PEC with smartphone application, which is very promising for dengue NS1 measurement.The mobile application significantly enhanced sensitivity and efficiency in qualifying NS1 in serum samples, compared to observation with the naked eye. This protocol is simple, cheap and reproducible to produce PEC to detect dengue NS1. The PEC is sensitive and specific for detecting dengue NS1, and will increase the sensitivities after combining with their smartphone.Our goal is to improve dengue NS1 past performance with the blood samples, evaluating the cross-infections cell types, and incubating it into the paper-based devices for multiple diseases. Addressing technology culture trend is crucial for successful commercialization in bring all the disease screening and medical care.

Selecting a fabrication method is pivotal to ensure reproducible assay performances in paper-based immunoassay devices. In our study, we explored various manufacturing processes and materials in the context of demonstrating a paper-based immunoassay. Our chosen method utilizes a wax printing system to create hydrophobic barriers within paper-based microfluidic devices. This approach stands out due to its simplicity, speed, and consistent results. Of note, it offers the advantage of avoiding the use of photoresist chemica...

Watch this Scientific Journal Video about Portable Paper-Based Immunoassay Combined with Smartphone Application for Colorimetric and Quantitative Detection of Dengue NS1 Antigen at JoVE.com

Addressing urgent dengue diagnostic needs, here we introduce a smartphone app-integrated Dengue NS1 Paper-based Analytical Device (DEN-NS1-PAD) for quantifying Dengue NS1 antigen concentration in clinical serum/blood samples. This innovation enhances dengue management by aiding clinical decision-making in various healthcare settings, even resource-limited ones.

Dengue virus (DENV) infection, which is transmitted by Aedes mosquitoes, is a major public health concern in tropical and subtropical countries. With an ...

portable-paper-based-immunoassay-combined-with-smartphone-application

Research

JoVE Journal

Immunology and Infection

720 Views.  Universitas Islam Indonesia. The scope of the research is detecting dengue using a specific biomarker NS1, with a paper-based analytical device. Our research focus on developing a simple, cheap, sensitive, and specific device to detect dengue. We also develop smartphone application to improve sensitivity and measurement of NS1 in real sample.Our sensitivity is developing a method to fabricate microfluidic method based on our coning analysis called PEC. Using formula is in our team, were some high techniques, aimed...