Quantification and profiling of urine cells by integrated cytocentrifugation and infrared spectroscopy

Autores de IIS La Fe

• Birgit Felderer

  Autor

• Julia Kuligowski

  Autor

Participantes ajenos a IIS La Fe

• Navarro-Esteve, V
• Quintás, G
• Wood, BR
• Perez-Guaita, D

Grupos

• Perinatología

Abstract

The presence of cells in urine and in particular White Blood Cells (WBCs) is often associated with Urinary Tract Infections (UTIs) and other diseases. Non-invasive screening of WBCs requires the development of cost-effective point of care diagnostic tools. Infrared (IR) spectroscopy has the potential to identify and quantify cells in urine. However, the quantification of cells by compact IR spectrophotometers can be hindered by the presence of highly concentrated interfering biomolecules. The use of separation procedures can assist in identifying and quantifying cells but reduces the point of care capabilities of the technology. In this study, we propose coupling cytocentrifugation with transflection IR spectroscopy for the isolation and quantification of cells in urine. Urine samples were spiked with monocytes and T-lymphocytes, cyto-centrifuged onto low-e slides and measured in transflection mode. An optional cell clean-up step, either performed before (by resuspending in PBS) or after the cytocentrifugation (by soaking the slide in water), was evaluated. In a first experiment using monocytes, IR band areas were linear (R-2 = 0.98) in the 8 x 10(3)-2 x 10(5) cells mL(-1) range, thus demonstrating the detection of cells at pathological numbers (pyuria, i.e., >10(4) WBCs mL(-1)). Secondly, to mimic real samples with varying cell types, urine samples containing both monocytes and T-lymphocytes were analysed to determine their concentration simultaneously. Partial Least Squares (PLS) regression enabled the simultaneous quantification of two types of different cells, yielding prediction errors of 2 x 10(4) cells mL(-1) for monocytes and 4 x 10(4) cells mL(-1) for T-lymphocytes. The results suggest that the technique has the potential to be implemented as a fast, simple, versatile, and cost-effective method for quantifying and profiling cells in urine.

Copyright © 2025 Elsevier B.V. All rights reserved.

Keywords

• Infrared spectroscopy; Point of care; Urinary tract infection; Urine; Leukocytes; Machine learning