Related Experiment Video
Updated: Jun 18, 2026
A Murine Orthotopic Bladder Tumor Model and Tumor Detection System
Published on: January 12, 2017
Machine-learning-assisted artificial olfactory colorimetric sensor array for bladder cancer early detection from
Yiquan Xiong1, Jun Xiao2, Lanyu Jing3
1Dongguan Key Laboratory of Interdisciplinary Science for Advanced Materials and Large-Scale Scientific Facilities, School of Physical Sciences, Great Bay University, Dongguan, Guangdong, 523000, China; Guangdong Provincial Key Laboratory of Optical Information Materials and Technology & Institute for Advanced Materials, South China Academy of Advanced Optoelectronics, South China Normal University, Guangzhou, 510006, China.
A novel artificial olfactory sensor array detects bladder cancer (BC) non-invasively using urinary volatile organic compounds (VOCs). Machine learning analysis of the sensor’s color changes shows high accuracy for early BC screening.
Area of Science:
- Biomedical Engineering
- Analytical Chemistry
- Oncology
Background:
- Non-invasive bladder cancer (BC) detection remains a significant clinical challenge.
- Current diagnostic methods often lack sensitivity for early-stage disease or are invasive.
Purpose of the Study:
- To develop and validate a machine-learning-assisted artificial olfactory colorimetric sensor array (CSA) for non-invasive BC screening.
- To analyze urinary volatile organic compound (VOC) profiles for BC detection.
Main Methods:
- Fabrication of a CSA using electrospun PVDF nanofibrous membranes integrated with eight cross-reactive dyes.
- Testing of urine samples from 134 subjects (81 BC patients, 53 controls) using static headspace exposure.
- Colorimetric data analysis via Euclidean distance and classification using a multilayer perceptron (MLP) model.
Main Results:
- The MLP model achieved high performance metrics: 0.806 accuracy, 0.913 recall, 0.851 F1-score, and 0.861 AUC.
- The CSA demonstrated reduced humidity interference and improved colorimetric reproducibility.
- Bromocresol green (Dye 6) was identified as the most influential sensing unit.
Conclusions:
- Electrospun CSA-based urinary VOC fingerprinting offers a promising low-cost, non-invasive strategy for BC screening.
- The developed sensor array and machine learning approach show significant potential for improving early cancer detection.