Selected medicinal plants leaves identification: a computer vision approach
- Authors: Deyi, Avuya
- Date: 2023-10-13
- Subjects: Deep learning (Machine learning) , Machine learning , Convolutional neural network , Computer vision in medicine , Medicinal plants
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/424552 , vital:72163
- Description: Identifying and classifying medicinal plants are valuable and essential skills during drug manufacturing because several active pharmaceutical ingredients (API) are sourced from medicinal plants. For many years, identifying and classifying medicinal plants have been exclusively done by experts in the domain, such as botanists, and herbarium curators. Recently, powerful computer vision technologies, using machine learning and deep convolutional neural networks, have been developed for classifying or identifying objects on images. A convolutional neural network is a deep learning architecture that outperforms previous advanced approaches in image classification and object detection based on its efficient features extraction on images. In this thesis, we investigate different convolutional neural networks and machine learning algorithms for identifying and classifying leaves of three species of the genus Brachylaena. The three species considered are Brachylaena discolor, Brachylaena ilicifolia and Brachylaena elliptica. All three species are used medicinally by people in South Africa to treat diseases like diabetes. From 1259 labelled images of those plants species (at least 400 for each species) split into training, evaluation and test sets, we trained and evaluated different deep convolutional neural networks and machine learning models. The VGG model achieved the best results with 98.26% accuracy from cross-validation. , Thesis (MSc) -- Faculty of Science, Mathematics, 2023
- Full Text:
- Date Issued: 2023-10-13
- Authors: Deyi, Avuya
- Date: 2023-10-13
- Subjects: Deep learning (Machine learning) , Machine learning , Convolutional neural network , Computer vision in medicine , Medicinal plants
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/424552 , vital:72163
- Description: Identifying and classifying medicinal plants are valuable and essential skills during drug manufacturing because several active pharmaceutical ingredients (API) are sourced from medicinal plants. For many years, identifying and classifying medicinal plants have been exclusively done by experts in the domain, such as botanists, and herbarium curators. Recently, powerful computer vision technologies, using machine learning and deep convolutional neural networks, have been developed for classifying or identifying objects on images. A convolutional neural network is a deep learning architecture that outperforms previous advanced approaches in image classification and object detection based on its efficient features extraction on images. In this thesis, we investigate different convolutional neural networks and machine learning algorithms for identifying and classifying leaves of three species of the genus Brachylaena. The three species considered are Brachylaena discolor, Brachylaena ilicifolia and Brachylaena elliptica. All three species are used medicinally by people in South Africa to treat diseases like diabetes. From 1259 labelled images of those plants species (at least 400 for each species) split into training, evaluation and test sets, we trained and evaluated different deep convolutional neural networks and machine learning models. The VGG model achieved the best results with 98.26% accuracy from cross-validation. , Thesis (MSc) -- Faculty of Science, Mathematics, 2023
- Full Text:
- Date Issued: 2023-10-13
Pharmaco-chemical investigation of Erythrina caffra: extracts, isolated compounds and their biological activities
- Authors: Nogqala, Simnikiwe
- Date: 2023-03-29
- Subjects: Coast coral tree , Traditional medicine South Africa , Antibacterial agents , Antineoplastic agents , Organic compounds
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422459 , vital:71944
- Description: In this study, secondary metabolites isolated from Erythrina caffra, a medicinal plant indigenous to South Africa, were investigated. E. caffra is well-known for its healing properties and it is traditionally used for treating bacterial infections like tuberculosis (TB), abscesses, tooth aches and ear infections. Its extracts have also been used to treat cancer. Though many studies have been done on this plant, most of them tended to focus solely on the isolated compounds. In the present study however, extracts, fractions and isolated compounds from E. caffra were evaluated for their anticancer, anti-oxidant, anti-enzymatic, antibacterial and cytotoxicity. The methanol crude extract (B1) from the stem bark of E. caffra was used to extract alkaloidic fractions (B2 and B3) using ethyl acetate and n-butanol respectively, a third fraction (B4) was also extracted using ethyl acetate this fraction was called a neutral fraction. The neutral fraction (B4) was fractionated and through a sequence of column chromatography three active secondary metabolites were isolated. The isolated compounds included Lupeol (1), stigmasterol (2) and 5,7-Dihydroxy-4'-methoxy-3',5'-diprenylflavanone (3). These isolated compounds were characterized and identified using spectroscopic techniques including IR, NMR and high-resolution Mass Spectrometry. Using the cell line HCC-70, isolated from a primary ductal carcinoma, in vitro anticancer assays were carried out on the crude extract from the bark, fractions, isolated compounds and an unseparated mixture of two compounds. These samples were also evaluated for their anti-oxidant, anti-enzymatic, antibacterial and cytotoxicity activities. The crude extract inhibited the cell viability by over 30% and had no effect on the HeLa cells at concentrations of 20μM. Abyssinone V’ 4-methyl-ether (3) and the mixture of stigmasterol (2) and an unidentified compound exhibited potent anticancer activity against the HCC-70 cell line with IC50 of 18.05μM and 9.04μM respectively. Antibacterial assays were also carried out on the crude extracts, fractions and concoctions made from the fractions with the best activity combined with the ones that performed poorly. The concoctions were prepared as two separate series (S and N series). The crude extract inhibited more than 80% of the Staphylococcus aureus cells at a concentration of 20μM with only minimal damage to the HeLa cells. In the concoctions however, the N series managed to inhibit over 96% of the S. aureus while exhibiting no cytotoxicity towards HeLa cells. The extract and its fractions also showed good anti-oxidant activities. Molecular docking of these compounds was done on the Human estrogen receptor (PDB ID:3ERT) and Abyssinone V’ 4-methyl-ether (3) showed the best docking score of -6.6 Kcal/mol, for the simulation against Epidermal growth factor receptor (PDB ID: 1M17) Stigmasterol (2) showed the best docking score of -3.8 Kcal/mol. In silico docking on 3ERT and 1M17 were done to test the binding affinity of the isolated compounds to the proteins which are well known to be overexpressed in some types of cancer. Flavonoids isolated from Erythrina species have been reported to possess good antiplasmodial activity. However, due to the minute amounts isolated in the present study in-vitro assays could not be carried out. Nevertheless, in-silico assays were conducted on the most prominent protozoal parasite which causes malaria in the majority of African countries. In-silico simulations were done against Plasmodium falciparum protein (PDB ID: 7KJH), of the tested compounds Abyssinone V’ 4-methyl-ether (3) was found possess the best docking score of -4.4 Kcal/mol. The molecular docking of 7KJH was done to assess the inhibitory potential of the isolated compounds on protozoal parasites. Pharmacokinetic properties of the isolated compounds were also assessed in silico to assist in evaluating the drug likeness of these compounds. The compounds showed a percent human oral absorption of 100% except for Abyssinone V’ 4-methyl-ether (3), which showed 93.83%, this indicates a remarkable oral bioavailability. Stigamsterol (2) exhibited a Caco-2 cell permeability (QPPCaco) greater than 500 which indicates outstanding results for good intestinal absorption. The compounds also displayed a blood-brain partition co-efficient (QPlogBB) ranging from -1.433 to 0.128 suggesting they will have less potential to cross the blood-brain barrier, thus reducing any CNS related toxicity. Molecular networking of the crude extracts and the fractions was done through GNPS which allowed the identification of known compounds including one isolated in the present study, Abyssinone V’ 4-methyl-ether (3). Possible derivatives that have not been isolated from this plant before were also putatively identified. , Thesis (MSc) -- Faculty of Science, Chemistry, 2023
- Full Text:
- Date Issued: 2023-03-29
- Authors: Nogqala, Simnikiwe
- Date: 2023-03-29
- Subjects: Coast coral tree , Traditional medicine South Africa , Antibacterial agents , Antineoplastic agents , Organic compounds
- Language: English
- Type: Academic theses , Master's theses , text
- Identifier: http://hdl.handle.net/10962/422459 , vital:71944
- Description: In this study, secondary metabolites isolated from Erythrina caffra, a medicinal plant indigenous to South Africa, were investigated. E. caffra is well-known for its healing properties and it is traditionally used for treating bacterial infections like tuberculosis (TB), abscesses, tooth aches and ear infections. Its extracts have also been used to treat cancer. Though many studies have been done on this plant, most of them tended to focus solely on the isolated compounds. In the present study however, extracts, fractions and isolated compounds from E. caffra were evaluated for their anticancer, anti-oxidant, anti-enzymatic, antibacterial and cytotoxicity. The methanol crude extract (B1) from the stem bark of E. caffra was used to extract alkaloidic fractions (B2 and B3) using ethyl acetate and n-butanol respectively, a third fraction (B4) was also extracted using ethyl acetate this fraction was called a neutral fraction. The neutral fraction (B4) was fractionated and through a sequence of column chromatography three active secondary metabolites were isolated. The isolated compounds included Lupeol (1), stigmasterol (2) and 5,7-Dihydroxy-4'-methoxy-3',5'-diprenylflavanone (3). These isolated compounds were characterized and identified using spectroscopic techniques including IR, NMR and high-resolution Mass Spectrometry. Using the cell line HCC-70, isolated from a primary ductal carcinoma, in vitro anticancer assays were carried out on the crude extract from the bark, fractions, isolated compounds and an unseparated mixture of two compounds. These samples were also evaluated for their anti-oxidant, anti-enzymatic, antibacterial and cytotoxicity activities. The crude extract inhibited the cell viability by over 30% and had no effect on the HeLa cells at concentrations of 20μM. Abyssinone V’ 4-methyl-ether (3) and the mixture of stigmasterol (2) and an unidentified compound exhibited potent anticancer activity against the HCC-70 cell line with IC50 of 18.05μM and 9.04μM respectively. Antibacterial assays were also carried out on the crude extracts, fractions and concoctions made from the fractions with the best activity combined with the ones that performed poorly. The concoctions were prepared as two separate series (S and N series). The crude extract inhibited more than 80% of the Staphylococcus aureus cells at a concentration of 20μM with only minimal damage to the HeLa cells. In the concoctions however, the N series managed to inhibit over 96% of the S. aureus while exhibiting no cytotoxicity towards HeLa cells. The extract and its fractions also showed good anti-oxidant activities. Molecular docking of these compounds was done on the Human estrogen receptor (PDB ID:3ERT) and Abyssinone V’ 4-methyl-ether (3) showed the best docking score of -6.6 Kcal/mol, for the simulation against Epidermal growth factor receptor (PDB ID: 1M17) Stigmasterol (2) showed the best docking score of -3.8 Kcal/mol. In silico docking on 3ERT and 1M17 were done to test the binding affinity of the isolated compounds to the proteins which are well known to be overexpressed in some types of cancer. Flavonoids isolated from Erythrina species have been reported to possess good antiplasmodial activity. However, due to the minute amounts isolated in the present study in-vitro assays could not be carried out. Nevertheless, in-silico assays were conducted on the most prominent protozoal parasite which causes malaria in the majority of African countries. In-silico simulations were done against Plasmodium falciparum protein (PDB ID: 7KJH), of the tested compounds Abyssinone V’ 4-methyl-ether (3) was found possess the best docking score of -4.4 Kcal/mol. The molecular docking of 7KJH was done to assess the inhibitory potential of the isolated compounds on protozoal parasites. Pharmacokinetic properties of the isolated compounds were also assessed in silico to assist in evaluating the drug likeness of these compounds. The compounds showed a percent human oral absorption of 100% except for Abyssinone V’ 4-methyl-ether (3), which showed 93.83%, this indicates a remarkable oral bioavailability. Stigamsterol (2) exhibited a Caco-2 cell permeability (QPPCaco) greater than 500 which indicates outstanding results for good intestinal absorption. The compounds also displayed a blood-brain partition co-efficient (QPlogBB) ranging from -1.433 to 0.128 suggesting they will have less potential to cross the blood-brain barrier, thus reducing any CNS related toxicity. Molecular networking of the crude extracts and the fractions was done through GNPS which allowed the identification of known compounds including one isolated in the present study, Abyssinone V’ 4-methyl-ether (3). Possible derivatives that have not been isolated from this plant before were also putatively identified. , Thesis (MSc) -- Faculty of Science, Chemistry, 2023
- Full Text:
- Date Issued: 2023-03-29
- «
- ‹
- 1
- ›
- »