Digital technologies to assess yoghurt quality traits and consumers acceptability

Area of Science:

• Food Science
• Sensory Science
• Biotechnology

Background:

• Sensory biometrics offer objective measurement of consumer responses during product tasting, overcoming limitations of self-reported data.
• This study investigated emotional and physiological reactions to various dairy and plant-based yoghurts.
• Physicochemical properties of yoghurts were analyzed alongside consumer feedback.

Purpose of the Study:

• To quantify consumer emotional and physiological responses to different commercial yoghurts.
• To correlate these responses with the yoghurts' physicochemical properties.
• To develop predictive models for yoghurt acceptability based on sensory and physical data.

Main Methods:

• Consumer liking was assessed using a nine-point hedonic scale (n=62).
• Participant emotions and heart rate were recorded via video analysis using the Bio-Sensory application.
• Physicochemical properties (Brix, pH, density, color, firmness, NIR spectroscopy) were measured.
• Machine learning models predicted physicochemical parameters and consumer liking using NIR data.

Main Results:

• Heart rate correlated positively with firmness, head movement, and overall liking, particularly for the 'Cookies' yoghurt sample.
• Two machine learning models demonstrated high predictive accuracy (R=0.98 for physicochemical parameters, R=0.99 for consumer liking).
• Near-infrared (NIR) spectroscopy effectively predicted yoghurt properties and consumer acceptance.

Conclusions:

• The developed sensory biometric and machine learning methods are highly accurate and reliable.
• These techniques provide a valuable tool for the food industry to assess yoghurt quality and consumer acceptability.
• Objective measurement of consumer responses can enhance product development and quality control in the food sector.