Predikcija fizikalno-kemijskih parametara vodenih ekstrakata samoniklog bilja primjenom blisko infracrvene spektroskopije u kombinaciji s kemometrijskim metodama

master's thesis

• Title (english): Prediction of physical and chemical parameters of medicinal plant extracts using near infrared spectroscopy in combination with chemometrics
• Author: Doris Dabić
• Mentor: Davor Valinger (mentor)
• Granter: University of Zagreb; Faculty of Food Technology and Biotechnology; (Department of Process Engineering); (Laboratory for Measurement, Regulation and Automatisation); Zagreb
• Defense date and country: 2019-09-17, Croatia
• Scientific / art field, discipline and subdiscipline: BIOTECHNICAL SCIENCES; Nutrition

Abstract

Prehrambena i farmaceutska industrija prepoznali su značaj ljekovitog i začinskog bilja kao izvor bioaktivnih funkcionalnih sastojaka sa antimikrobnom, protuupalnom, antialergijskom i antioksidativnom ulogom u pružanju zdravstvenih benefita za ljude. Epidemiološka istraživanja ukazuju da fitokemikalije usporavaju degenerativne procese i smanjuju rizik od raka i kardiovaskularnih bolesti, a veliki značaj pridaje se implementaciji tih sastojaka u nove proizvode s dodanom vrijednošću. U ovom radu... More provedeno je istraživanje na vodenim ekstraktima 15 biljaka: kamilice, koprive, maslačka, stolisnika, trputca, melise, nevena, mente, preslice, kadulje, majčine dušice, lavande, masline, gospine trave i lipe. Ekstraktima su testirana fizikalna svojstva (pH, električna vodljivosti, TDS, udio suhe tvari, vode i pepela) i kemijski parametri (ukupni polifenoli, antioksidacijski kapacitet izmjeren trima metodama: ABTS, FRAP i DPPH). Cilj je bio provjeriti može li se pomoću blisko infracrvene spektroskopije napraviti predikcija 10 fizikalno-kemijskih parametara, a kako bi to utvrdili provedena je analiza glavnih komponenata (PCA) nakon čega su napravljeni modeli parcijalnom regresijskom metodom najmanjih kvadrata (PLS) i umjetnih neuronskih mreža (ANN). Pokazalo se da NIR spektroskopija u kombinaciji sa PCA i ANN može dati vrlo dobru predikciju fizikalno-kemijskih parametara, a utvrđeno je i da su rezultati ANN modela pokazali bolje slaganje sa eksperimentalnim podatcima nego PLS model. Less

Abstract (english)

Food industry and pharmaceutical industry have recognized the importance of herbs and spices as a source of bioactive functional ingredients with an antimicrobial, anti-inflammatory, anti-allergic and antioxidant role in providing health benefits to humans. Epidemiological studies indicate that phytochemicals are slowing down degenerative processes and can reduce the risk of cancer and cardiovascular disease. The implementation of these ingredients in new, value-added products is of great... More importance to the industry. In this study, an examination was conducted on aqueous extracts of 15 plants: chamomile, nettle, dandelion, yarrow, plantain, lemon balm, calendula, mint, horsetail, sage, thyme, lavender, olive, St. John'swortand linden. Physical properties of extracts were tested (pH, electrical conductivity, TDS, and dry matter, water and ash content) as well as chemical parameters (total polyphenols and antioxidant capacity measured by three methods: ABTS, FRAP and DPPH).The aim was to test whether near-infrared spectroscopy can predict 10 physicochemical parameters, and to determine this, principal component analysis (PCA) was performed, followed by partial least squares regression (PLS) and artificial neural networks (ANN).The study showed that NIR spectroscopy in combination with PCA and ANN can give a very good prediction of physicochemical parameters, and it was found that the ANN model results showed better concurrence with experimental data than the PLS model. Less

• Keywords: samoniklo bilje; blisko-infracrvena spektroskopija; kemometrija; umjetne neuronske mreže; parcijalna linearna regresija
• Keywords (english): medicinal plants; near infrared spectroscopy; chemometrics; artificial neural networks; partial least squares regression
• Language: croatian
• URN:NBN: urn:nbn:hr:159:892970
• Study programme: Title: Nutrition; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra nutricionizma (magistar/magistra nutricionizma)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2020-02-13 14:52:55