Hanbat National University Study Advances Machine Learning Calibration of Biosensors for Microcystin Toxin Monitoring in Freshwater, Enhancing On-Site Accuracy

July 19, 2026

PR Newswire (Hanbat National University) South Korea

Overview

Researchers from Hanbat National University and the University of Central Florida developed a machine learning framework to improve the calibration accuracy of portable screen-printed carbon electrode biosensors for detecting microcystin-lysine-arginine (MC-LR) toxin in freshwater. By incorporating water quality parameters such as electrical impedance, conductivity, pH, and UV absorbance, the AI model significantly enhanced the precision of on-site toxin detection. This framework aims to reduce the need for repeated sample-specific calibration, thereby cutting down on time, labor, and sensor consumption, leading to more efficient environmental monitoring.

In Depth

Key Findings

A collaborative research team from Hanbat National University in Korea and the University of Central Florida in the U.S. has developed a machine learning (ML) framework that significantly enhances the calibration accuracy of portable screen-printed carbon electrode (SPCE) biosensors for monitoring microcystin-lysine-arginine (MC-LR) toxin in freshwater. This innovative approach integrates various water quality parameters—including electrical impedance, conductivity, pH, and UV absorbance—into an AI model, thereby improving the reliability of on-site toxin detection. The framework is designed to reduce the need for frequent, sample-specific calibration, leading to substantial savings in time, labor, and sensor consumption, and ultimately boosting the efficiency of environmental monitoring.

Technical & Clinical Details

Target Toxin and Biosensor: This study focuses on microcystin toxins, particularly MC-LR, which are produced by toxic cyanobacteria (blue-green algae) and frequently pose problems in freshwater bodies, with severe implications for human and animal health. Low-cost, portable SPCE biosensors are utilized for detection. SPCEs function by electrochemically detecting specific target substances after surface immobilization with specific recognition molecules, such as antibodies or enzymes.
Machine Learning Framework: Traditional biosensors often suffer from calibration drift due to changing environmental conditions. To address this, the research team implemented a machine learning model. The developed AI model incorporates not only electrochemical signals from the SPCE biosensor but also multiple auxiliary water quality parameters, including water temperature, pH, conductivity, total dissolved solids (TDS), and UV absorbance, as inputs.
Enhanced Calibration Accuracy: The AI model comprehensively analyzes this multi-variate data, learning the complex, non-linear relationships between the biosensor’s response and actual MC-LR concentrations. As a result, it significantly improves MC-LR detection accuracy by compensating for environmental factors, offering a notable advantage over conventional methods. This enables reliable data acquisition even in diverse field environments.
Efficiency and Cost Reduction: By reducing the necessity for per-sample calibration, the framework streamlines and accelerates the measurement process. It also leads to a reduction in reagent and sensor consumption, contributing to lower operational costs.

Background & Industry Context

Microcystins, produced by harmful algal blooms (HABs) prevalent in freshwater lakes and reservoirs worldwide, can cause severe health impacts on humans and animals. Contamination of drinking water sources is a major public health concern, necessitating rapid and accurate monitoring systems. However, for remote or widespread monitoring, specialized laboratory analyses are often time-consuming and expensive. While portable biosensors offer the potential for rapid on-site detection, issues with accuracy due to environmental variations have been a challenge. The integration of AI/ML technology is set to overcome these issues, fundamentally changing the paradigm of environmental monitoring.

Strategic Significance & Outlook

This machine learning framework is applicable not only to MC-LR toxin monitoring but also to biosensors detecting other environmental pollutants (e.g., heavy metals, pesticides, other algal toxins). In the future, it is expected to be integrated into smart sensor networks and IoT devices, forming the core of real-time, wide-area automated environmental monitoring systems. This will significantly enhance water resource safety, protect public health, and contribute to ecosystem conservation. Further advancements in AI-driven self-calibration capabilities will accelerate the realization of next-generation environmental sensors that are maintenance-free and capable of long-term operation.

Source: https://www.prnewswire.com/news-releases/hanbat-national-university-study-advances-machine-learning-calibration-of-biosensors-for-microcystin-toxin-monitoring-in-freshwater-302300062.html

Get our weekly technology intelligence — free

Receive an infographic that lets you judge at a glance whether each field’s analysis report is worth reading.

Subscribe Free — Weekly Tech Intelligence

By subscribing, you’ll receive Troy-Technical’s weekly technology intelligence newsletter.