Millet is an unexploited cereal with potential in the food industry due to its nutritional value and resistance to harsh climate conditions. Nutritious millet byproducts have a potential application in the development of functional cereal products, but require processing in order to improve their physical and nutritional quality. Therefore, we investigated high intensity ultrasound as a pretreatment to increase the amount of freely available bioactives from proso millet bran. We also analysed the effect of high intensity ultrasound on enzymatic browning, water retention and protein digestibility, which are crucial for the utilization in the bakery and pasta industry. A 15 % millet bran suspension in water was treated with 400-W ultrasound probe for 5, 12.5 or 20 min, with the 60, 80 or 100 % amplitude. High intensity ultrasound treatment with 80 % amplitude for 12.5 min improved most significantly the nutritive value; the antioxidant activity measured by FRAP test increased by 15 % (p<0.05), and total phenolic content by 16 % (p<0.05). Still, the impact on the increase of water-soluble and ethanol-insoluble dietary fibre by 38 % was evident after the treatment for 20 min at 100 % amplitude. High intensity ultrasound treatment at 100 % amplitude for 5 min caused the largest improvements in water retention and limited browning of the sample. High intensity ultrasound treatment activated polyphenol oxidase, regardless of the applied heating of the sample. Due to its ambiguous impact on proso millet bran characteristics, the treatment required an optimization, which showed that the optimal pretreatment of a 15 % millet bran suspension in water is at 100 % amplitude for 9.3 min.

Different lactic acid bacteria starters were used to prepare sourdough to make partially-baked frozen wholemeal wheat bread. The sourdough was prepared with a pure culture of Lactobacillus plantarum or with commercial starters containing Lactobacillus brevis combined with Saccharomyces cerevisiae var. chevalieri (LV4), Lactobacillus fermentum (PL1), or Lactobacillus fermentum with phytase (PL3). We determined the acetic and lactic acid concentrations in sourdough, bread chemical composition, total phenolics content and glycemic index (GI) in vivo. Depending on the starter, the lactic to acetic acid ratio in the sourdough was significantly different. The GI of control bread without sourdough (70) was significantly higher than that of bread containing sourdough prepared with LV4 starter (50), PL1 starter (56) or PL3 starter (56), but not from bread with L. plantarum sourdough (60). The addition of 10% sourdough with a lower molar ratio of lactic to acetic acid ( ≤ 4) and higher total phenolics content is preferable for generating bread with medium and low GI.