20.500.12556/DKUM-15687 Uporabnost NIR spektroskopije za napovedovanje kemične sestave mesa in mesnih izdelkov Ability of NIR spectroscopy to predict chemical composition of meat and meat products Ugotavljali smo zanesljivost bližnje infrardeče spektroskopije (NIRS) za napovedovanje kemijske sestave mesa in mesnih izdelkov. Vzorcem smo kemijsko določili vsebnost intramuskularne maščobe (IMM) z ekstrakcijo po Soxhletu s predhodno hidrolizo (SIST ISO 1443, 2001), vsebnost vode (ISO 6496, 1999), vsebnost beljakovin (ISO 5983-2, 2005) ter izračunali razmerje med vodo in beljakovinami (VB). Z aparatom NIR Systems model 6500 (Silver Spring, MD, USA) smo posneli spektre vzorcev na območju valovnih dolžin vidne in NIR svetlobe (400-2500 nm) ter pripravili umeritvene enačbe (WinISI II programski paket) za različna spektralna območja (NIR, vidni ali celoten spekter) znotraj mišice, za različne mišice ter na vseh vzorcih skupaj. Točnost kalibracij smo ocenili s statističnimi parametri: determinacijski koeficient navzkrižne validacije (R2CV) in predikcije (R2P), standardna napaka navzkrižne validacije (SECV) in predikcije (SEP) ter razmerjem med standardnim odklonom referenčnih vrednosti in standardno napako (RPD). Točnost napovedovanja je bila pri vseh enačbah odlična za vsebnost IMM (RCV: 0,91- 0,98; RPD: 3,40-6,74). Za ostale lastnosti je bila točnost (z izjemo slabe točnosti kalibracij znotraj mišice longissimus dorsi) le malo slabša (za beljakovine RCV: 0,80-0,91, RPD: 2,26-3,29; za vodo RCV: 0,73-0,98, RPD: 1,92-7,55; za VB: RCV: 0,65-0,92, RPD: 1,68-3,43). Glede na različne kazalce kalibracijske statistike lahko zaključimo, da lahko z NIRS zelo točno napovedujemo kemično sestavo mesa in mesnih izdelkov (z izjemo kalibracij za vsebnost beljakovin znotraj mišice longissimus dorsi). The aim of this study was to determine the accuracy of predicting chemical composition of meat and meat products by near infrared spectroscopy (NIRS). For each sample we chemically determined intramuscular fat (IMF) content using Soxhlet extraction with hydrolysis (SIST ISO 1443, 2001), moisture content (ISO 6496, 1999), protein content (ISO 5983-2, 2005) and calculatedratio moisture-to-protein. Sample spectra were recorded with spectrophotometer NIR Systems model 6500 (Silver Spring, MD, USA) at the spectral range of visible and NIR wavelengths (400-2500 nm). Calibration equations were prepared using WinISI II software for various spectral ranges (NIR, visible or the whole spectrum) for a single muscle, several muscles and all samples together. The accuracy of calibration was assessed by the following statistical parameters: coefficient of determination in cross-validation (R2CV) and prediction (R2P), standard error of cross-validation (SECV) and prediction (SEP) and the ratio between standard deviation of reference values and standard error (RPD). All equations showed very good accuracy to predict IMF content (RCV: 0,91-0,98; RPD: 3,40-6,74). For other constituents the accuracy (with the exception of the low accuracy of calibrations within longissimus dorsi muscle sample group ) was slightly lower (for proteins RCV: 0,80-0,91, RPD: 2,26-3,29; moisture: RCV: 0,73-0,98, RPD: 1,92-7,55; moisture and protein ratio: RCV: 0,65-0,92, RPD: 1,68-3,43). Based on parameters of calibration statistics we can conclude that it is possible to accurately predict the chemical composition of meat and meat products with NIRS (with the exception of protein content in the longissimus dorsi sample group). NIR spektroskopija meso mesni izdelki kemijska sestava NIR spectroscopy meat meat products chemical composition true true false Slovenski jezik Angleški jezik Diplomsko delo 2010-09-16 19:42:33 2010-10-13 13:02:05 2022-04-13 07:30:25 2023-12-19 03:27:20 2010 0 Maribor 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 URN:SI:UM:DK:BXFHEC25 UNI_Drevensek_Andreja_2010.pdf UNI_Drevensek_Andreja_2010.pdf 1 E8D4650010D38D914EA9BC53622AB1B3 4b5426b5691f273ab11bc8c125122b500550da8af5fa3e47f5192d34677924f5 7524b77a-7c04-11eb-bb7a-00155d0001ca 20.500.12556/dkum/51fe3877-c531-4e91-aaa7-23e42ccda968 https://dk.um.si/Dokument.php?lang=slv&id=17354 Fakulteta za kmetijstvo in biosistemske vede 0 0 0