Related Experiment Videos
Infrared analysis of urinary calculi by a single reflection accessory and a neural network interpretation algorithm.
M Volmer1, J C de Vries, H M Goldschmidt
1Department of Pathology and Laboratory Medicine, University Hospital Groningen (UHG), PO Box 30001, 9700 RB Groningen, The Netherlands.
Clinical Chemistry
|June 28, 2001
Summary
A new Fourier transform infrared (FT-IR) method using a Golden Gate Attenuated Total Reflection system and artificial neural networks (ANN) offers a faster, more robust analysis of urinary calculus composition compared to traditional KBr tablet methods.
Area of Science:
- Analytical Chemistry
- Biomedical Engineering
- Spectroscopy
Background:
- Traditional Fourier transform infrared (FT-IR) analysis of urinary calculus using KBr tablets is laborious and prone to sample breakage.
- A novel FT-IR approach was developed to overcome these limitations in urinary calculus analysis.
Purpose of the Study:
- To introduce and evaluate a new, rapid FT-IR method for analyzing urinary calculus composition.
- To compare the performance of the new method against the conventional KBr tablet technique.
Main Methods:
- Utilized a Golden Gate Single Reflection Diamond Attenuated Total Reflection (ATR) sample holder for sample analysis.
- Developed a computer library and an artificial neural network (ANN) for spectral interpretation and quantitative analysis.
- Trained and validated the ANN with a comprehensive library of pure components and mixtures, and tested on patient samples.
Main Results:
- The optimized ANN model achieved low root mean square errors (1.5% training, 2.3% validation).
- Comparison with the KBr tablet method showed good agreement for most urinary calculus components, with minor biases for specific minerals.
- The new method demonstrated a 95% level of agreement for all results.
Conclusions:
- The Golden Gate FT-IR method is superior due to its smaller sample requirement, ease of use, robustness, and speed.
- The combination of library search and ANN prediction reduces the need for extensive expert interpretation, though visual inspection is still advised.