The squared distance approach to frequency domain time-resolved fluorescence analysis.

Area of Science:

• Biophotonics
• Medical Diagnostics
• Spectroscopy

Background:

• Frequency-domain (FD) analysis of fluorescence lifetime (FLT) offers rapid cellular classification for medical diagnostics.
• Current FLT data analysis relies on nonlinear fitting, which can reduce classification accuracy.
• There is a need for robust FLT analysis methods unaffected by data processing distortions.

Purpose of the Study:

• To develop and validate a novel sample classification technique for FD-FLT data.
• To overcome limitations of nonlinear fitting in FLT data analysis.
• To assess the diagnostic potential of the new method in simulated and experimental data.

Main Methods:

• Utilized a squared distance (D²) metric applied directly to raw frequency response data (FRD).
• Validated the D² technique using two simulated datasets with varying FLT spacings.
• Tested the method on experimental data from bacterial and viral infected patients.

Main Results:

• Achieved over 95% classification accuracy in simulated datasets across six groups.
• Correctly classified 41 out of 43 patient samples compared to previous reports.
• Demonstrated high agreement with preliminary physician diagnoses for infected samples.

Conclusions:

• The D² approach provides a robust and analysis-independent method for FLT-based sample classification.
• This technique significantly improves classification accuracy compared to traditional methods.
• FD-FLT measurements using the D² method show promise for sensitive and specific medical diagnostics.