Although yeasts have been used as indicator micro-organisms of
quality deterioration of the yoghurt with fruit (spoilage of the
product when the population was 105 CFU/g) (Suriyarachchi and
Fleet, 1981), the present work showed that the quality impairment
of the product may be determined more precisely based on
the evident spoilage. An alternative or an extension of the current
model would be the modelling of the shelf-life of the product based
on the occurrence of mycelium on the product surface owned to
spore germination (Gougouli and Koutsoumanis, 2010). In this way,
the development of stochastic models to predict the spoilage of
these products is possible. The occurrence of mycelium on the
surface of the yoghurt is dependent on germination time of spores
and growth rate of mycelium. Temperature conditions during
distribution and retail storage affect these two parameters resulting
in a distribution of the time of spore germination and mycelium
occurrence. Therefore, the mathematical description of the kinetic
behaviour of spore germination and mycelium growth can be used
to predict the shelf-life of the product and also to allow the development
of decision-making tools for quality improvement. The end
of shelf-life could be the time of spore germination just before the
occurrence ofmycelium on the surface of the product (Gougouli and
Koutsoumanis, 2010). Another alternative method to control the
expiration date of yoghurt is the use of Artificial Neural Networks
(Sofu and Ekinci, 2007). The authors used as inputs variables of the
network the measurements of pH, total aerobic, yeasts, mould,
coliform counts and color analysis values measured by a machine
vision system at different days of storage (e.g. 1, 7 and 14), whereas
the output variable was the storage time of the yoghurt