Year 2025

• Application of UV-VIS Spectroscopy and Machine Learning Methods in Glucosuria Diagnostics: A Phantom Study

  Publication

  • M. Babińska
  • A. Władziński

  - Photonics Letters of Poland - Year 2025

  In this study, UV-VIS spectroscopy was used as a tool for detecting low glucose concentrations in urine. Measurements were performed on artificial urine samples and solutions with 0.1% and 0.2% glucose, covering both normal and pathological thresholds. Among the evaluated models, Random Forest reached 0.887for the 0.1% glucose sample, while Logistic Regression achieved 0.7796 for the 0.2% glucose sample,...

  Full text available to download

• Machine Learning Enhanced Optical Fiber Sensor For Detection Of Glucose Low Concentration In Samples Mimicking Tissue

  Publication

  • M. Babińska
  • A. Władziński
  • T. Talaśka
  • M. Szczerska

  - Photonics Letters of Poland - Year 2025

  This study presents anoptical fiber sensor for detecting low glucose concentrations in a sample mimicking urine. Our research focused on designing a sensor capable of detecting 0.5% glucose concentrations in artificial urine. Algorithms were applied to analyze and accurately classify the data andidentify the principal components of the collected data. The Random Forest and XGBoostmodel achieved the highest...

  Full text available to download