Electronic Nose: A Non-destructive Technology to Screen Tomato Fruit with Internal Bruising

Tomato (Lycopersicon esculentum Mill) fruit, cv. Solar Set, were harvested at the mature-green stage and gassed with 100 mg·kg–1 of ethylene at 20 °C. At the breaker stage, fruit were held by vacuum to avoid fruit rotation and dropped from a 40 cm height on a metallic, solid, smooth surface. Following impact, fruit were stored at 20 °C and 85% to 95% relative humidity until table-ripe stage. Bruised and unbruised fruit were then placed individually inside the electronic nose-sampling vessel and the 12 conducting polymer sensors were lowered into the vessel and exposed to the volatile given off by the fruit. Data were analyzed employing multivariate discriminant analysis (MVDA), which maximizes the variance between treatments. The degree of dissimilarity was defined using the Mahalanobis distance and posterior probabilities were calculated to accurate re-classification of cases. The differences found between bruised and unbruised fruit were highly significant (P < 0.0041). The Mahalanobis distance between groupings (28.19 units) was a dramatic indicative of the differences between the two treatments. The re-classification of bruised and unbruised fruit using a single linear discriminant function was highly accurate, being 1.0 for both bruised and unbruised fruit. The electronic nose proved to be a useful tool to nondestructively identify and classify tomato fruit exposed to harmful postharvest practices such as mechanical injuries. However, there are still some factors that must be investigated, including system stability and the development of specific sensors for specific commodities.