Comprehensive Summary
This study explored if multispectral fluorescence imaging combined with machine learning can detect both visible and invisible fecal contamination on chicken carcasses, which is a main source of Salmonella. Researchers applied fecal material from different parts of the digestive tract onto chicken skin, rinsed the carcasses to mimic reprocessing, then imaged them with a handheld fluorescence camera. Computer models were trained to classify the images, and swabs were tested with PCR to confirm Salmonella. The system identified contamination even after rinsing, reaching more than 97 percent accuracy. The authors concluded that visual checks alone can miss hidden contamination and that imaging methods through machine learning may improve food safety inspection. This method provides a reliable way to target carcasses that need further cleaning. It also highlights the potential of combining imaging and artificial intelligence tools to improve accuracy and consistency in food safety.
Outcomes and Implications
This study matters because Salmonella from poultry is a main cause of food illness. Being able to detect contamination that cannot be seen by inspectors helps prevent bacteria from reaching consumers. While not meant for clinical diagnosis, the method has clear public health value by reducing infections. The authors note that with more testing in real production settings, this approach could be used in processing plants in the near future. It could also lower the number of recalls and improve consumer confidence in food safety. The study shows how machine learning and imaging can be used in practical ways to protect public health. In the future, this type of technology may also be used for other food industries to control hidden contamination.