Fluorescence Lifetime Endoscopy with a Nanosecond Time-Gated CAPS Camera with IRF-Free Deep Learning Method

Area of Science:

• Biomedical Optics
• Medical Imaging
• Endoscopy

Background:

• Fluorescence imaging is crucial in preclinical and clinical settings.
• Advancements in imaging technology and fluorescent labels are driving the adoption of fluorescence lifetime imaging.
• Distinguishing between malignant and benign cells is vital in medical diagnostics.

Purpose of the Study:

• To develop and demonstrate an endoscopic fluorescence lifetime imaging system for improved cell differentiation.
• To achieve real-time fluorescence lifetime imaging in the near-infrared (NIR) region (700-900 nm).
• To evaluate a novel deep learning-based method for fluorescence lifetime imaging with minimal time points.

Main Methods:

• Development of the tauCAM system utilizing the Current-Assisted Photonic Sampler.
• Demonstration using a rigid endoscope with various phantoms.
• Application of an IRF-free deep learning algorithm requiring only 6 time points.

Main Results:

• The endoscopic lifetime imaging system successfully demonstrated its capabilities.
• The deep learning-based method achieved fluorescence lifetime imaging with high precision and uniformity.
• The system using 6 time points outperformed conventional methods in lifetime uniformity and precision.

Conclusions:

• Endoscopic fluorescence lifetime imaging holds significant potential for enhanced medical diagnostics.
• The developed tauCAM system and deep learning approach offer a precise and efficient method for real-time NIR fluorescence lifetime imaging.
• This technology can improve the differentiation of malignant and benign cells, paving the way for advanced endoscopic procedures.