Microwave cavity antenna for automatic detection of chicken breast muscles affected by wooden breast defect

Over the past years, artificial selection for meat-type (broiler) chickens has resulted in improved production efficiency, but also in an increased incidence of skeletal muscle abnormalities. In particular, wooden breast (WB) abnormality causes abrupt changes in meat quality and enormous economic losses. Thus, the poultry processing industry requires a reliable method to detect the abnormality in production lines in a non-destructive and contactless way. The present research aimed to evaluate the effect of WB on dielectric properties to develop a potential online technique to distinguish unaffected from affected breasts. Sixty-five pectoralis major muscles were picked 3 h post-mortem from the same flock (42-day-old broilers, 2.8 kg average live weight) and classified by visual inspection according to their phenotype as normal (N, not showing not any signs of the WB condition; n = 33) or WB (exhibiting extensive hardened areas and stiffness perceived by manual palpation throughout the entire fillet; n = 32). WB muscles exhibited remarkably higher values of dielectric constant and loss factor in a wide range of the explored frequencies (200 MHz-14 GHz), suggesting higher water mobility and a higher solvate capacity; this makes the electromagnetic technique for classification promising. Based on the above evidence, a rapid microwave electric technique in the range 1.5 GHz-3 GHz was developed for the online detection of WB. Indeed, combining the use of a cavity antenna (vector network analyzer instrumental chain with partial least square - discriminant analysis), this technique correctly classified 100% of the validation test set.