Related Experiment Video
Updated: Aug 25, 2025
A Rapid Method for Multispectral Fluorescence Imaging of Frozen Tissue Sections
Published on: March 30, 2020
Early Detection of Cervical Cancer by Fluorescence Lifetime Imaging Microscopy Combined with Unsupervised Machine
Mingmei Ji1, Jiahui Zhong2, Runzhe Xue1
1Department of Optical Science and Engineering, Shanghai Engineering Research Center of Ultra-Precision Optical Manufacturing, Key Laboratory of Micro and Nano Photonic Structures (Ministry of Education), School of Information Science and Technology, Fudan University, 220 Handan Road, Shanghai 200433, China.
This study introduces a new method using fluorescence lifetime imaging microscopy (FLIM) and machine learning (ML) for cervical cancer screening. FLIM-ML offers higher accuracy than traditional cytology for detecting cervical cancer risk and recurrence.
Area of Science:
- Biomedical Optics
- Medical Imaging
- Machine Learning in Healthcare
Background:
- Cervical cancer poses a significant global health burden with high mortality rates.
- Current screening methods like cytology lack optimal sensitivity and specificity for early detection.
- Accurate screening is crucial for timely intervention and effective follow-up care.
Purpose of the Study:
- To evaluate the potential of fluorescence lifetime imaging microscopy (FLIM) combined with unsupervised machine learning (ML) for cervical cancer risk assessment.
- To compare the diagnostic performance of the FLIM-ML method against conventional cytology.
- To explore the utility of FLIM-ML in predicting cancer recurrence for improved patient follow-up.
Main Methods:
- Exfoliated cervical cells from 71 participants were imaged using FLIM to detect endogenous reduced nicotinamide adenine dinucleotide (phosphate) [NAD(P)H].
- FLIM data were analyzed using an unsupervised machine learning approach to develop a cervical cancer risk prediction model.
- The performance of the developed FLIM-ML model was compared with cytology results.
Main Results:
- The FLIM-ML method achieved a sensitivity of 90.9% and a specificity of 100%.
- These performance metrics significantly surpassed those of the standard cytology approach.
- The FLIM-ML method successfully predicted one case of cancer recurrence several months earlier than conventional clinical methods.
Conclusions:
- FLIM-ML demonstrates significant clinical applicability as a novel detection method for cervical cancer screening.
- The FLIM-ML approach offers superior diagnostic accuracy compared to current cytology methods.
- This technique shows promise as a valuable tool for monitoring patients and managing follow-up cancer care.
