Identification of Microorganism in Infected Wounds by Positively Charged Selective Sensor Array and Deep Learning...

Area of Science:

• Microbiology
• Biochemistry
• Materials Science

Background:

• Microorganisms are crucial to human health and disease.
• Accurate and rapid identification of pathogens is vital for infection diagnosis.
• Existing methods for microorganism identification can be time-consuming or lack specificity.

Purpose of the Study:

• To develop a novel fluorescence sensor array for rapid and accurate microorganism identification.
• To investigate the interaction mechanism between synthesized tetraphenylethylene derivatives and microbial cell components.
• To utilize a deep learning model for classifying various microorganisms and their mixtures.

Main Methods:

• Synthesis of three tetraphenylethylene derivatives (S-TDs) with varying properties.
• Creation of a fluorescence sensor array utilizing S-TDs for microorganism imaging.
• Calculation of binding energies between S-TDs and microbial cell wall constituents.
• Application of a deep learning residual network (ResNet) model for image analysis and classification.

Main Results:

• The S-TDs sensor array successfully imaged diverse microorganisms via distinct fluorescent signatures.
• Binding energy calculations elucidated the interaction mechanisms between S-TDs and microbial cell components.
• The fluorescence sensor array combined with ResNet achieved high accuracy in differentiating Gram-negative bacteria, Gram-positive bacteria, and fungi.
• Accuracies exceeded 92.8% for Gram species and antibiotic-resistant species, and 90.35% for mixed microorganism detection in infected wounds.

Conclusions:

• A novel, rapid, and accurate method for microbial classification was developed using a fluorescence sensor array and deep learning.
• This approach holds significant potential for improving the diagnosis and treatment of infectious diseases.
• The study demonstrates the utility of tailored fluorescent probes and AI in microbial diagnostics.