Modelling and Classification of Apple Textural Attributes Using Sensory, Instrumental and Compositional Analyses

Area of Science:

• Food Science
• Agricultural Science
• Sensory Science

Background:

• Fruit texture is crucial for quality, shelf-life, and consumer appeal.
• Instrumental texture analysis offers a potentially reliable alternative to subjective sensory evaluations.

Purpose of the Study:

• To validate instrumental texture measurements against sensory perceptions in apples.
• To develop a classification system for apples based on their textural properties.
• To create predictive models linking instrumental and compositional data to sensory attributes.

Main Methods:

• Evaluated textural characteristics (crispness, hardness, juiciness, skin toughness) of 12 apple cultivars.
• Utilized a TA.XTplus Texture Analyzer for instrumental measurements and performed compositional analyses.
• Applied Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) for data interpretation and classification.

Main Results:

• Two principal components explained 95.88% of instrumental variation, correlating with flesh firmness and skin strength.
• Four distinct textural apple groups were identified with 94.44% classification accuracy using LDA.
• Predictive models explained over 85% of variation for hardness and crispness; juiciness and skin toughness models were more complex.

Conclusions:

• Instrumental texture analysis can reliably predict key sensory attributes of apples.
• A robust classification scheme for apples based on texture can be established.
• This approach can significantly reduce reliance on time-consuming and expensive sensory evaluations in the industry.