Rapid detection of viruses is necessary for early diagnosis of the disease and initiation of treatment. However, currently popular quantitative polymerase chain reaction (qPCR) techniques involve multistep protocols which consume a lot of time and also use expensive optics for yielding accurate results. Electrochemical biosensors have developed into reliable analytical tools over the past decade for sensitive, specific and rapid detection of many viruses. Electrochemical impedance spectroscopy (EIS) is being used more frequently in electrochemical sensors and biosensors for both characterizing the sensors throughout the various phases of sensor design, as well as a method for quantitative and qualitative measurements. In this study, a sensitive electronic detection scheme for viral RNA is developed based on non-faradaic EIS. Impedance measurements corresponding to biomodification on the electrode surface is carried out through an in-house micro-chip. Sensitivity is included through gold nanoparticle labeling to the DNA primers and a dielectrophoretic (DEP) concentration step. All the tests are conducted with cDNA synthesized from Hepatitis C virus (HCV) RNA samples and the Limit of detection (LOD) achieved through the currently developed chip is in the range of 30 to 35 ng/μL i.e., around 140 nM for a 327 bp cDNA molecule. Further, the device starts detection procedure within 15 minutes of initiation.
Conventional nucleic acid methods often face challenges in achieving both high sensitivity and specificity for reliable detection of viral RNA. Reverse transcription polymerase chain reaction (RT-PCR) provides excellent sensitivity but fluorescence readouts often face ambiguity of detection at low template concentrations. Reverse transcription loop mediated isothermal amplification (RT-LAMP) integrated with electrochemical impedance spectroscopy (EIS) enables easy, cost effective, label-free detection and supports portable assays, however, non-specific impedance responses often reduce assay specificity. To overcome these limitations, we apply non-negative matrix factorization (NMF) as a post-processing strategy for heterogeneous biosensing signals. NMF decomposes data into physically interpretable latent components by enforcing non-negativity, making it well suited for effective amplification and detection analysis. Two datasets are analyzed: cycle-dependent fluorescence intensity from RT-PCR and frequency-dependent impedance magnitude from RT-LAMP EIS. NMF successfully discriminates amplification-specific contributions from residual noise, improving sensitivity in fluorescence data and specificity in electrochemical measurements. The resulting component scores enable robust classification of positive and negative samples achieving an overall sensitivity and specificity of 100% and 86.7%, outperforming conventional both optical and electrochemical schemes. Using hepatitis C virus RNA as a representative case, we demonstrate that non-negative decomposition framework offers a generalizable strategy for improving the accuracy of molecular diagnostics across diverse biosensing platforms.