Related Experiment Video
Updated: Feb 26, 2026
05:47
Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems
Published on: June 13, 2025
1.7K
Machine learning-based prediction of Clostridium growth in pork meat using explainable artificial intelligence.
Volkan Ince1, Mohamed Bader-El-Den1, Jack Alderton2
1Computer Science, University of Portsmouth, Winston Churchill Ave, Portsmouth, PO1 3GY UK.
Journal of Food Science and Technology
|February 25, 2026
Summary
Machine learning and explainable AI accurately detect harmful bacteria in meat. Integrating genetic sequencing reveals spoilage bacteria can inhibit harmful types, improving food safety predictions.
Area of Science:
- Food Science
- Microbiology
- Computational Biology
Background:
- Bacterial growth in meat poses food safety risks.
- Predicting harmful bacterial growth is crucial for food quality assurance.
Purpose of the Study:
- To detect harmful bacteria in pork using machine learning (ML).
- To analyze the influence of other bacteria on harmful growth via explainable artificial intelligence (XAI).
Main Methods:
- Genetic sequencing for bacterial diversity analysis in pork samples.
- ML algorithms and XAI for bacterial growth prediction and influence analysis.
- Exclusion of low-abundance bacteria (<0.1%, 0.25%, 0.5%) to focus on dominant populations.
Main Results:
- Significant negative correlation found between meat spoilage bacteria and harmful bacteria (r=-0.385, p<0.05).
- ML model accuracy reached 0.89 with the inclusion of the 'day' variable for prediction.
- XAI indicated that spoilage bacteria can reduce harmful bacterial growth in specific media.
Conclusions:
- Integrated genetic sequencing, ML, and XAI provide novel insights into bacterial dynamics.
- This approach enhances predictive food safety, quality control, and shelf-life extension.
- The findings support improved food safety strategies and reduced industry losses.
