A predictive model for the evaluation of flavor attributes of raw and cooked beef based on sensor array analyses

Area of Science:

• Food Science
• Analytical Chemistry
• Sensory Science

Background:

• Objective evaluation of beef flavor attributes is lacking, particularly for comparing raw and cooked beef.
• Beef flavor is a critical factor influencing consumer purchasing decisions.
• Current methods lack standardization for objective beef flavor assessment.

Purpose of the Study:

• To develop a predictive model for objective beef flavor attribute evaluation.
• To reduce the complexity and detection time of flavor analysis.
• To determine if raw beef can be used for reliable flavor assessment.

Main Methods:

• Utilized a 12-ion-sensor array combined with sensory properties to create a predictive model.
• Reduced the sensor array to six sensors and incorporated a saturated calomel reference electrode.
• Employed multivariate statistical methods, including cluster analysis (CA) and genetic algorithm (GA), for data processing and comparison with sensory evaluation.

Main Results:

• Achieved 100% accuracy in classifying five basic beef flavors (acidity, sweetness, bitterness, saltiness, freshness) with the optimized sensor array.
• Genetic algorithm (GA) analysis showed high consistency with sensory evaluation, yielding accuracy rates of 85.0% for raw beef and 90.0% for cooked beef.
• Reduced sensor count decreased data dimensionality and detection time, enabling efficient flavor profiling.

Conclusions:

• The developed sensor array model provides an objective and repeatable method for evaluating beef flavor attributes.
• Raw beef can be effectively used for flavor attribute evaluation, simplifying the process.
• This approach offers a significant advancement over subjective sensory evaluations for beef quality control.