Deep Learning-Based Kinetic Analysis in Paper-Based Analytical Cartridges Integrated with Field-Effect Transistors

Area of Science:

• Biosensing and Analytical Chemistry
• Electrical Engineering
• Computational Biology

Background:

• Paper-based analytical devices often suffer from low sensitivity.
• Field-effect transistor (FET) sensors offer electrical measurement capabilities but face challenges like sample matrix interference.
• Integrating these technologies can overcome existing limitations in biosensing.

Purpose of the Study:

• To develop a novel quantitative biosensing platform by combining FET sensors, paper-based cartridges, and deep learning (DL).
• To enhance sensitivity and accuracy in biosensing through kinetic data analysis.
• To create a cost-effective and user-friendly diagnostic tool for point-of-care applications.

Main Methods:

• Utilized a field-effect transistor (FET) integrated with a paper-based analytical cartridge for biosensing.
• Employed deep learning (DL) algorithms to analyze kinetic data from bioreactions.
• Performed quantitative analysis of cholesterol concentration using the developed platform.

Main Results:

• The integrated system demonstrated improved sensitivity compared to traditional paper analytical devices.
• Deep learning analysis effectively mitigated sample matrix interference, a common challenge in FET biosensors.
• Cholesterol testing showed a coefficient of variation of <6.46% and a strong correlation (r² >0.976) with clinical laboratory results.

Conclusions:

• The fusion of FET sensors, paper cartridges, and DL offers a powerful approach for quantitative biosensing.
• This technology has the potential to significantly advance point-of-care diagnostics and at-home testing.
• The developed platform provides enhanced accessibility, ease-of-use, and accuracy in diagnostic measurements.