AI-assisted colorimetric sensor lowers nitrite detection limit for home UTI screening
Investigators developed a portable, smartphone-integrated colorimetric sensor for urinary nitrite detection using an electrospun PAN/PVP membrane to stabilize the Griess reagent for 30 days. A deep convolutional neural network processes the colorimetric readout to quantify nitrite across a 5.00 to 200.0 µM range, achieving 95.20 to 104.00% recovery and 99.50% accuracy in artificial urine. The platform reduces the detection limit from 25.0 µM to 4.65 µM compared to commercial test strips and delivers results in 2 minutes without complex preprocessing. Validated in real urine samples, the system offers a reliable, low-cost strategy for decentralized point-of-care screening of urinary tract infections.
The original study
Rapid and home-testable integrated colorimetric sensor for nitrite detection in urine.
- Authors
- Liu FJ, Li HL, Ke-Shi, Li SX, Luo JY, Huang ZJ, et al.
- Journal
- Mikrochimica acta
- Type
- Journal Article
- PMID
- 42834246
Original abstract
A portable, rapid and convenient intelligent detection platform for nitrite is proposed based on a polyacrylonitrile NO2- colorimetric sensor (PAN-NP) integrated with a smartphone for household urinary nitrite detection. The PAN-NP membrane was prepared from polymer polyacrylonitrile (PAN) and polyvinylpyrrolidone (PVP) via electrospinning technology to protect the Griess reagent, resulting in prolonged stability for 30 days. Deep Convolutional Neural Network (DCNN) was utilized for color recognition to determine NO2- content, and it can detect NO2- in artificial urine over a concentration range 5.00-200.0 µM, with recoveries of 95.20-104.00% and accuracy of 99.50%. Compared to commercial test strip reflectance analysis, it lowers the detection limit for NO2- from 25.0 µM to 4.65 µM, facilitating earlier detection and treatment of urinary tract infections, and reducing the entire testing time to 2 min without complex preprocessing. This method has been successfully applied to detect nitrite in real urine samples, providing a reliable strategy for point-of-care testing of urinary nitrite and intelligent home health monitoring.