Rapid quantification of CSF total protein via smartphone digital colorimetry: Enabling early diagnosis of central nervous system diseases.

OBJECTIVE

This study evaluated a smartphone-based system for measuring cerebrospinal fluid (CSF) total protein, verifying its compliance with clinical requirements and exploring its application in diagnosing central nervous system (CNS) diseases in remote/medically underserved areas.

METHODS

The system analyzed CSF total protein standards to establish a standard curve (concentration vs. R, G, Y values). Per Clinical and Laboratory Standards Institute (CLSI) guidelines, key parameters (precision, detection capability, linear range [LR], clinical reportable range [CRR]) were assessed. Forty-two clinical samples from neurological patients were tested, with results compared to the Beckman Coulter AU5800 analyzer for consistency. Accuracy at three medical decision levels (based on CSF total protein reference ranges) was verified.

RESULTS

The limit of blank (LOB), limit of detection (LOD), and limit of quantitation (LOQ) were 5.18 mg/dL (0.05 g/L), 11.78 mg/dL (0.12 g/L), and 12.59 mg/dL (0.13 g/L), respectively. The LR was 29.75-127.20 mg/dL (0.30-1.27 g/L), and the CRR was 12.59-508.80 mg/dL (0.13-5.09 g/L; via 1:3 dilution), both meeting clinical requirements. Total imprecision coefficients of variation (CVs) were 3.83% (high-concentration: 91.52 mg/dL [0.92 g/L]) and 3.59% (low-concentration: 31.32 mg/dL [0.31 g/L]), within acceptable ranges. The 42 clinical samples showed strong consistency with AU5800 results (R2 = 0.991). In the evaluation of the smartphone-based CSF total protein detection system, although the 95% confidence intervals of expected biases at 45 mg/dL (0.45 g/L) and 60 mg/dL (0.60 g/L) exceed 1/4 total allowable error per CLSI EP9-A2, the deviations are clinically irrelevant-with the maximum bias (5.29 mg/dL [0.0529 g/L]) fails to alter the pathological threshold classification of CSF total protein (≥45 mg/dL [0.45 g/L]), thus confirming the acceptable bias of the system for auxiliary CSF protein detection.

CONCLUSIONS

The system meets clinical standards, is user-friendly and cost-effective, and has high potential for aiding CNS disorder diagnosis in resource-constrained areas.