Smartphone-based ELISA for rapid pathogen detection in low-resource settings

Thesis Abstract

In low-resource settings, timely and affordable pathogen detection remains a critical bottleneck for outbreak prevention and patient management, due to limited laboratory infrastructure, supply chain constraints, and skilled personnel deficits. This study investigates a smartphone-based ELISA system designed to perform rapid, quantitative detection of focal pathogens at the point of need, leveraging a colorimetric readout and cloud-enabled data analytics to bridge the gap between field testing and clinical decision-making. The aim is to evaluate the diagnostic performance, operational feasibility, and data utility of the smartphone-ELISA platform under real-world conditions. The specific objectives are (1) to optimize a low-cost, disposable ELISA microplate-compatible plate holder and optical module that integrates with a standard smartphone for absorbance or reflectance readings; (2) to calibrate the device against conventional laboratory ELISA using serial dilutions of known antigen standards to establish analytical sensitivity, linear range, limit of detection, and dynamic range; (3) to assess diagnostic accuracy (sensitivity, specificity, positive and negative predictive values) in field samples collected from community health centers (n=420) across two rural districts; (4) to evaluate user acceptability, usability, and workflow integration among community health workers using a mixed-methods approach; and (5) to model the potential public health impact through cost-effectiveness and outbreak surveillance simulations. The methodology adopts a pragmatic, mixed-methods research design underpinned by the Technology Acceptance Model and the Diffusion of Innovations framework to capture adoption dynamics. The population comprises health workers and patients presenting with febrile syndromes or suspected infections, with a purposive sample of 60 health workers and 420 patient samples. Data collection instruments include a standardized smartphone-ELISA protocol, laboratory reference ELISA kits, field data collection forms, user usability questionnaires (System Usability Scale and NASA-TLX), semi-structured interviews, and field performance logs. Validity and reliability of the instruments are ensured through cross-validation against reference ELISA results (R2 > 0.98 for calibration curves), test-retest reliability (intraclass correlation > 0.85 for reading stability), and triangulation of qualitative and quantitative data. Data analysis comprises multiple approaches diagnostic accuracy metrics are computed using ROC curves and Youden’s index; regression analyses (logistic and linear) quantify factors influencing readout accuracy and time-to-result; Bland-Altman analysis assesses agreement between smartphone-based measurements and laboratory ELISA; thematic analysis is applied to interview transcripts to elucidate barriers and facilitators to adoption; and health economic evaluation employs a tentative cost-effectiveness model comparing per-test costs and projected disease averted under different uptake scenarios. A conceptual framework linking sensor performance, user interaction, and data ecosystems is articulated to guide interpretation. Expected findings include a high analytical performance with a strong correlation (R2 ? 0.95) between smartphone readings and laboratory ELISA across the dynamic range, sensitivity and specificity exceeding 90% in field samples, and acceptable usability scores by trained health workers (system usability score > 70). It is anticipated that the platform will reduce turn-around time to result from 6–24 hours to under 90 minutes in most field cases and lower per-test costs by 40–60% relative to centralized ELISA workflows, assuming scaled production of disposable kits. The study contributes to knowledge by advancing an ICT-enabled, portable diagnostic paradigm for infectious diseases in resource-limited contexts, integrating smartphone sensing with cloud-based analytics, and providing empirical evidence on feasibility, diagnostic performance, and cost-effectiveness. The main conclusion is that smartphone-based ELISA can deliver reliable, rapid, and scalable pathogen detection in low-resource settings when combined with user-centered workflow design and robust data integration. Recommendations include standardizing kit manufacturing, developing region-specific data governance policies, expanding multi-pathogen panels, and conducting longitudinal trials to evaluate impact on outbreak control and health system resilience.