The authors of this study developed the use of attenuated total reflectance Fourier transform infrared spectroscopy (ATR–FTIR) combined with machine learning as a point-of-care (POC) diagnostic platform, considering neonatal respiratory distress syndrome (nRDS), for which no POC currently exists, as an example. nRDS can be diagnosed by a ratio of less than 2.2 of two nRDS biomarkers, lecithin and sphingomyelin (L/S ratio), and in this study, ATR–FTIR spectra were recorded from L/S ratios of between 1.0 and 3.4, which were generated using purified reagents. The calibration of principal component (PCR) and partial least squares (PLSR) regression models was performed using 155 raw baselined and second derivative spectra prior to predicting the concentration of a further 104 spectra. A three-factor PLSR model of second derivative spectra best predicted L/S ratios across the full range (R²: 0.967; MSE: 0.014). The L/S ratios from 1.0 to 3.4 were predicted with a prediction interval of +0.29, −0.37 when using a second derivative spectra PLSR model and had a mean prediction interval of +0.26, −0.34 around the L/S 2.2 region. These results support the validity of combining ATR–FTIR with machine learning to develop a point-of-care device for detecting and quantifying any biomarker with an interpretable mid-infrared spectrum.

Reducing measurement uncertainty is crucial to enable the adoption of rapid point-of-use techniques for clinical and industrial applications. Diagnosis of neonatal respiratory distress syndrome and liposome formulation quality control are two applications for which measuring the ratio of the lecithin to sphingomyelin composition of liposomes is important, for which no rapid measurement currently exists. Raman and infrared spectroscopies are two complementary approaches to examine characteristic molecular vibrations that can spectroscopically measure liposomes and, when combined with machine learning, predict their composition. We show that employing a data-fusion approach the uncertainty in the predicted compositions compared to the individual modalities (IR R²: 0.902 and Raman R²: 0.951) can be reduced to obtain more accurate and precise measurements (low-level fused model R²: 0.973, mean squared error: 0.024, prediction interval width: 0.303, high-level weighted fusion model R²: 0.970, mean squared error: 0.027, prediction interval width: 0.268).

Rapid point of care assessment of pulmonary surfactant composition by measuring the lecithin/sphingomyelin (L/S) ratio could improve management of patients with neonatal respiratory distress syndrome (nRDS). Attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR) offers a practical route to making such measurements, but the influence of the sample solvent prior to drying on measurement repeatability is poorly understood. We compare films dried from dichloromethane (DCM) and water (AQ) solvents (DCM-dry route vs. AQ-dry route) by ATR-FTIR and show that spectra from the AQ-dry route increased the signal-to-noise ratio (SNR) of a representative (2920 cm⁻¹) absorption peak for the mixture from 20.13 to 128.20 and for human endotracheal aspirate (ETA) from 6.33 to 8.13. A mixed nested analysis of variance (ANOVA) showed that drying route accounted for 89.52% of mixture peak height variance and reduced percent relative standard deviation (%RSD) from 23.5% to 16.2%, corroborated by multivariate analysis for ETA. We further demonstrate that partial least squares regression (PLSR) models trained on AQ-dry mixture spectra predicted L/S (R² = 0.91; root mean square error (RMSE) = 0.31) with 95% prediction interval grey-zone interpretation around L/S = 2.2, complemented by a receiver operating characteristic area under the curve (ROC-AUC) of 0.978.