Cell-phone-based measurement of TSH using Mie scatter optimized lateral flow assays

Semi-quantitative thyr oid stimulating hormone (TSH) lateral flow immunochromatographic assays (LFA) are used to screen for serum TSH concentration >5 mIUL(-1) (hypothyroidism). The LFA format, however, is unable to measure TSH in the normal range or detect suppressed levels of TSH (<0.4 mIU L(-1); hyperthyroidism). In fact, it does not provide quantitative TSH values at all. Obtaining quantitative TSH results, especially in the low concentration range, has until now required the use of centralized clinical laboratories which require specimen transport, specialized equipment and personnel, and result in increased cost and delays in the timely reporting of important clinical results. We have conducted a series of experiments to develop and validate an optical system and image analysis algorithm based upon a cell phone platform. It is able to provide point-of-care quantitative TSH results with a high level of sensitivity and reproducibility comparable to that of a clinical laboratory-based third-generation TSH immunoassay. Our research approach uses the methodology of the optimized Rayleigh/Mie scatter detection by taking into consideration the optical characteristics of a nitrocellulose membrane and gold nanoparticles on an LFA for quantifying TSH levels. Using a miniature spectrometer, LED light source, and optical fibers on a rotating benchtop apparatus, the light intensity from different angles of incident light and angles of detection to the LFA were measured. The optimum angles were found that the minimized Mie scattering from nitrocellulose membrane, consequently maximizes the Rayleigh scatter detection from the gold nanoparticles in the LFA bands. Using the results from the benchtop apparatus, a cell-phone-based apparatus was designed which utilized the embedded flash in the cell phone camera as the light source, piped the light with an optical fiber from the flash through a collimating lens to illuminate the LFA. Quantification of TSH was performed in an iOS application directly on the phone and verified using the code written in MATLAB. The limit of detection of the system was determined to be 0.31 mIU L(-1) (never achieved before on an LFA format), below the commonly accepted minimum concentration of 0.4 mIU L(-1) indicating clinical significance of hyperthyroidism. The system was further evaluated using human serum showing an accurate and reproducible platform for rapid and point-of-care quantification of TSH using a cell phone.