Integrated analysis and pattern recognition of Swiss cheese aroma by SPME/GC, SPME/GC/MS and electronic noses

Two objective methods: (a) Electronic Noses, (b) Solid Phase Micro Extraction (SPME)/GC, were utilized to differentiate the aroma portion of Swiss cheese volatile compounds and these results were cross compared to sensory evaluation. Investigation was conducted in two trials. In both trials, Alpha M.O.S. Fox 2000 and Fox 3000 electronic noses differentiated cheeses by discriminant analysis, which could be related to flavor characteristics of Swiss cheese.; SPME/GC was found to provide for quantitative analysis of the head space volatiles of Swiss cheese. Thirty predominant and common peaks found in all cheeses were integrated by the internal HP ChemStation software and selected for further data analysis.; The GC data were further analyzed by discriminant analysis. The discriminant plots of major GC peaks were compared with that of the Electronic Nose sensory discriminant patterns. Similar results were observed with both instruments. Acetoin and C-2 to C-6 fatty acids were primary compounds that affected the differentiation of the cheeses by the electronic nose.; An Artificial Neural Networks (ANN) program with back propagation learning mechanism was programmed in PASCAL computer language. In trial one, ANN was trained with the first batch of data comprising of five replicates of each cheese sample. An optimum performance was determined by the training time and number of hidden units. Another batch of GC data obtained at different dates were fed into the ANN to test the performance of this ANN. ANN was able to recognize all cheeses perfectly. By adding noise to ANN, the relationship of different cheeses to one another could be determined.; ANN was not successful in recognizing all of the cheeses when it as applied to Electronic Nose data.; Overall the results suggest that discriminant analysis of Electronic Nose odor maps and pattern recognition by ANN on GC peaks complement human sensory evaluation and can provide an objective approach to aroma analysis. The combination of both Electronic Nose and SPME/GC provided more information than either method alone.