Automated Calculation of Liquid Crystal Sensing Images Based on Deep Learning

Area of Science:

• Biomedical Engineering
• Materials Science
• Computer Science

Background:

• Liquid crystal (LC) sensors are widely used for detecting chemical and biological events.
• Manual analysis of LC sensor images is time-consuming and prone to user-dependent errors.

Purpose of the Study:

• To develop an automated, deep learning-based method for calculating optical images from LC-based sensors.
• To improve the speed and accuracy of analyzing LC sensing data.

Main Methods:

• A convolutional neural network was trained using prepared LC sensing images and segmentation annotations.
• The method calculates the ratio of positive response area to the total area for quantitative analysis.
• Algorithm robustness was validated using test datasets and label-free Cadmium (Cd2+) detection.

Main Results:

• The deep learning method enables real-time detection of positive response areas in LC sensor images.
• The automated approach significantly outperforms manual processing in terms of speed.
• The method demonstrates high accuracy and robustness in quantitative analysis.

Conclusions:

• Deep learning provides an efficient and accurate automated solution for analyzing LC sensor images.
• This approach can be readily adapted for other label-free molecular detection assays.
• The developed method enhances the utility of LC-based sensors in various scientific applications.