Sandi Baressi Baressi Šegota, S. Lysdahlgaard, S. Hess, R. Antulov
The fact that Artificial Intelligence (AI) based algorithms exhibit a high performance on image classification tasks has been shown many times. Still, certain issues exist with the application of machine learning (ML) artificial neural network (ANN) algorithms. The best known is the need for a large amount of statistically varied data, which can be addressed with expanded collection or data augmentation. Other issues are also present. Convolutional neural networks (CNNs) show extremely high performance on image-shaped data. Despite their performance, CNNs exhibit a large issue which is the sensitivity to image orientation. Previous research shows that varying the orientation of images may greatly lower the performance of the trained CNN. This is especially problematic in certain applications, such as X-ray radiography, an example of which is presented here. Previous research shows that the performance of CNNs is higher when used on images in a single orientation (left or right), as opposed to the combination of both. This means that the data needs to be differentiated before it enters the classification model. In this paper, the CNN-based model for differentiation between left and right-oriented images is presented. Multiple CNNs are trained and tested, with the highest performing being the VGG16 architecture which achieved an Accuracy of 0.99 (+/- 0.01), and an AUC of 0.98 (+/- 0.01). These results show that CNNs can be used to address the issue of orientation sensitivity by splitting the data in advance of being used in classification models.
{"title":"USE OF CONVOLUTIONAL NEURAL NETWORKS FOR X-RAY IMAGE ORIENTATION DETERMINATION","authors":"Sandi Baressi Baressi Šegota, S. Lysdahlgaard, S. Hess, R. Antulov","doi":"10.46793/iccbi21.263bs","DOIUrl":"https://doi.org/10.46793/iccbi21.263bs","url":null,"abstract":"The fact that Artificial Intelligence (AI) based algorithms exhibit a high performance on image classification tasks has been shown many times. Still, certain issues exist with the application of machine learning (ML) artificial neural network (ANN) algorithms. The best known is the need for a large amount of statistically varied data, which can be addressed with expanded collection or data augmentation. Other issues are also present. Convolutional neural networks (CNNs) show extremely high performance on image-shaped data. Despite their performance, CNNs exhibit a large issue which is the sensitivity to image orientation. Previous research shows that varying the orientation of images may greatly lower the performance of the trained CNN. This is especially problematic in certain applications, such as X-ray radiography, an example of which is presented here. Previous research shows that the performance of CNNs is higher when used on images in a single orientation (left or right), as opposed to the combination of both. This means that the data needs to be differentiated before it enters the classification model. In this paper, the CNN-based model for differentiation between left and right-oriented images is presented. Multiple CNNs are trained and tested, with the highest performing being the VGG16 architecture which achieved an Accuracy of 0.99 (+/- 0.01), and an AUC of 0.98 (+/- 0.01). These results show that CNNs can be used to address the issue of orientation sensitivity by splitting the data in advance of being used in classification models.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"59 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89167866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The discovery that palladium complexes possess a wide range of biological activities (from antitumor, -viral, -malarial, -fungal to antimicrobial activities) encourages further research in this scientific field. Herein we describe the synthesis and characterization of a novel palladium (II) complex, using [Pd(dien)Cl]Cl and 2-(azidomethyl)cyclopropane-1,1-dicarboxylic acid (azmcpda) as a ligand. [Pd(dien)Cl]Cl was selected as a starting material taking into consideration its importance as a model for the investigation of the substitution reactions in coordination chemistry and a deeper understanding of the biological activities of some structurally similar compounds. The ligand compound was synthesized by the procedure described in the literature. It is noteworthy to mention that 2- (azidomethyl)cyclopropane-1,1-dicarboxylic acid presents the precursor for the synthesis of 2- (aminomethyl)cyclopropane-1,1-dicarboxylic acid, as an example of the constrained γ-amino dicarboxylic acids. The synthesis was achieved by the conversion of the ligand compound into the corresponding sodium dicarboxylate salt and subsequent treatment with [Pd(dien)Cl]Cl (pH maintained between 6-7). The IR and NMR spectra, as well as elemental analysis have confirmed that the Na[Pd(dien)(azmcpda)]. H2O species was formed and that coordination of the ligand compound to the metal ion was established through carboxylate oxygen donor atom.
{"title":"SYNTHESIS AND CHARACTERIZATION OF PALLADIUM (II)–2- (AZIDOMETHYL)CYCLOPROPANE-1,1-DICARBOXYLIC ACID COMPLEX","authors":"M. Kostić, V. Divac, Sven Mangelinckx","doi":"10.46793/iccbi21.297k","DOIUrl":"https://doi.org/10.46793/iccbi21.297k","url":null,"abstract":"The discovery that palladium complexes possess a wide range of biological activities (from antitumor, -viral, -malarial, -fungal to antimicrobial activities) encourages further research in this scientific field. Herein we describe the synthesis and characterization of a novel palladium (II) complex, using [Pd(dien)Cl]Cl and 2-(azidomethyl)cyclopropane-1,1-dicarboxylic acid (azmcpda) as a ligand. [Pd(dien)Cl]Cl was selected as a starting material taking into consideration its importance as a model for the investigation of the substitution reactions in coordination chemistry and a deeper understanding of the biological activities of some structurally similar compounds. The ligand compound was synthesized by the procedure described in the literature. It is noteworthy to mention that 2- (azidomethyl)cyclopropane-1,1-dicarboxylic acid presents the precursor for the synthesis of 2- (aminomethyl)cyclopropane-1,1-dicarboxylic acid, as an example of the constrained γ-amino dicarboxylic acids. The synthesis was achieved by the conversion of the ligand compound into the corresponding sodium dicarboxylate salt and subsequent treatment with [Pd(dien)Cl]Cl (pH maintained between 6-7). The IR and NMR spectra, as well as elemental analysis have confirmed that the Na[Pd(dien)(azmcpda)]. H2O species was formed and that coordination of the ligand compound to the metal ion was established through carboxylate oxygen donor atom.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"32 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85116796","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nikola Ristivojević, D. Dimić, Marko Đošić, Stefan Mišić, Anđela Gavran, Jelena Đorović Jovanović, Jasmina M. Dimitrić Marković
Anabolic steroids are a group of commonly counterfeit substances used by individuals who want to gain weight and muscles. Testosterone propionate (TP), an ester analog of testosterone, belongs to this group and its spectroscopic analysis is important especially when it is improperly labeled and misused. In this contribution quantum chemical methods, at the B3LYP/6- 311++G(d,p) level of theory, were applied for the prediction of the vibrational (IR and Raman) and UV-VIS spectra of TP. The applicability of the chosen level of theory was proven based on the comparison between experimental and theoretical bond lengths and angles. The most prominent bands in the IR and Raman spectra were assigned and correlated with the calculated ones. The electronic spectra were also analyzed and the assignments were made based on the Time-Dependent Density Functional Theory (TD-DFT) calculations. The orbitals included in the most intense transitions were visualized and possible solvent effects were discussed. The presented results proved the applicability of the DFT methods for the prediction of spectra that could lead to the counterfeit substances determination.
{"title":"SPECTROSCOPIC AND QUANTUM-CHEMICAL INVESTIGATION OF TESTOSTERONE PROPIONATE, A COMMONLY MISUSED ANABOLIC STEROID","authors":"Nikola Ristivojević, D. Dimić, Marko Đošić, Stefan Mišić, Anđela Gavran, Jelena Đorović Jovanović, Jasmina M. Dimitrić Marković","doi":"10.46793/iccbi21.450r","DOIUrl":"https://doi.org/10.46793/iccbi21.450r","url":null,"abstract":"Anabolic steroids are a group of commonly counterfeit substances used by individuals who want to gain weight and muscles. Testosterone propionate (TP), an ester analog of testosterone, belongs to this group and its spectroscopic analysis is important especially when it is improperly labeled and misused. In this contribution quantum chemical methods, at the B3LYP/6- 311++G(d,p) level of theory, were applied for the prediction of the vibrational (IR and Raman) and UV-VIS spectra of TP. The applicability of the chosen level of theory was proven based on the comparison between experimental and theoretical bond lengths and angles. The most prominent bands in the IR and Raman spectra were assigned and correlated with the calculated ones. The electronic spectra were also analyzed and the assignments were made based on the Time-Dependent Density Functional Theory (TD-DFT) calculations. The orbitals included in the most intense transitions were visualized and possible solvent effects were discussed. The presented results proved the applicability of the DFT methods for the prediction of spectra that could lead to the counterfeit substances determination.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"79 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84089759","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
P. Stanić, N. Vukićević, V. Cvetković, M. Pavlovic, S. Dimitrijević, Biljana M Šmit
Four 2-thiohydantoin derivatives were synthesized and their corrosion inhibition properties on mild steel (MS) in 0.5M HCl solution was evaluated using usual gravimetric and electrochemical methods (weight loss, potentiodynamic polarization, electrochemical impedance spectroscopy (EIS). Morphology of the metal surface was characterized by scanning electron microscopy (SEM) and atomic force microscopy (AFM). The study has shown that these compounds provide good protection for mild steel against corrosion in the acidic medium.
{"title":"ELECTROCHEMICAL INVESTIGATION OF 2-THIOHYDANTOIN DERIVATIVES AS CORROSION INHIBITORS FOR MILD STEEL IN ACIDIC MEDIUM","authors":"P. Stanić, N. Vukićević, V. Cvetković, M. Pavlovic, S. Dimitrijević, Biljana M Šmit","doi":"10.46793/iccbi21.157s","DOIUrl":"https://doi.org/10.46793/iccbi21.157s","url":null,"abstract":"Four 2-thiohydantoin derivatives were synthesized and their corrosion inhibition properties on mild steel (MS) in 0.5M HCl solution was evaluated using usual gravimetric and electrochemical methods (weight loss, potentiodynamic polarization, electrochemical impedance spectroscopy (EIS). Morphology of the metal surface was characterized by scanning electron microscopy (SEM) and atomic force microscopy (AFM). The study has shown that these compounds provide good protection for mild steel against corrosion in the acidic medium.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"110 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86277721","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nevena Milivojević, D. Caballero, M. Carvalho, M. Zivanovic, N. Filipovic, Rui R Reis, J. Oliveira
Further technological advances are in great need for improving our understanding about critical biological and fundamental pathological processes, such as tissue development and cancer progression, or for the discovery and screening of novel pharmacological drugs. Preclinical experimentation demands for highly reliable and physiologically-relevant systems capable of recapitulating the complex human physiology. Traditional in vitro models, albeit widely employed, fail to reproduce the complexity of the native scenario with cells displaying aberrant gene expressions. Similarly, in vivo animal models, such as mice, poorly mimic the human condition and are ethically questionable. During the last decades, a new paradigm in preclinical modelling has emerged aiming to solve the limitations of the aforementioned methods. The combination of advanced tissue engineering, nanotechnology, and cell biology has resulted in the development of cutting-edge microfluidics-based models with an unprecedented ability to recreate within a microfluidic device the native habitat of cells within a microengineered chip. A diverse variety of micro- and bio-fabrication techniques is available for the development of microfluidic devices. Among all them, UV-photolithography and soft lithography is the considered the gold-standard method for the fabrication of chips due to its simplicity, versatility, and rapid prototyping. In this work, we describe the step-by-step fabrication procedure of a microfluidic chip by UV-photolithography and replica molding and discuss about their potential applications in the biomedical field.
{"title":"ENGINEERING A MICROFLUDIC PLATFORM AS A PRE-CLINICAL MODEL FOR BIOMEDICAL APPLICATIONS","authors":"Nevena Milivojević, D. Caballero, M. Carvalho, M. Zivanovic, N. Filipovic, Rui R Reis, J. Oliveira","doi":"10.46793/iccbi21.259m","DOIUrl":"https://doi.org/10.46793/iccbi21.259m","url":null,"abstract":"Further technological advances are in great need for improving our understanding about critical biological and fundamental pathological processes, such as tissue development and cancer progression, or for the discovery and screening of novel pharmacological drugs. Preclinical experimentation demands for highly reliable and physiologically-relevant systems capable of recapitulating the complex human physiology. Traditional in vitro models, albeit widely employed, fail to reproduce the complexity of the native scenario with cells displaying aberrant gene expressions. Similarly, in vivo animal models, such as mice, poorly mimic the human condition and are ethically questionable. During the last decades, a new paradigm in preclinical modelling has emerged aiming to solve the limitations of the aforementioned methods. The combination of advanced tissue engineering, nanotechnology, and cell biology has resulted in the development of cutting-edge microfluidics-based models with an unprecedented ability to recreate within a microfluidic device the native habitat of cells within a microengineered chip. A diverse variety of micro- and bio-fabrication techniques is available for the development of microfluidic devices. Among all them, UV-photolithography and soft lithography is the considered the gold-standard method for the fabrication of chips due to its simplicity, versatility, and rapid prototyping. In this work, we describe the step-by-step fabrication procedure of a microfluidic chip by UV-photolithography and replica molding and discuss about their potential applications in the biomedical field.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85072492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sonja Stojanov, M. Jovičić, I. Bobinac, Olga M. Govedarica, J. Pavličević, Julijana Blagojević, Dragan D. Govedarica, O. Bera
This paper aims to investigate the effect of the addition of mineral oil and pyrolytic carbon black on crosslinking the natural rubber and the mechanical properties of the crosslinked products. A rheometer determined curing characteristics at a temperature of 150 °C. The mechanical properties of prepared vulcanized composites were determined. By adding mineral oil to rubber compounds, the vulcanization reaction starts later, and it takes slightly more time to achieve the optimal vulcanization time. The addition of mineral oil to the rubber mixture achieves better dispersion of pyrolytic carbon blacks in the matrix and thus increases the physical interaction between the filler and rubber. Pyrolytic carbon black (pCB) is obtained by recycling waste products and contains a higher proportion of impurities. Due to impurities, PCB has a smaller surface area for the physical adsorption of rubber molecules than standard carbon black, and it can be assumed that this has led to a decrease in the crosslinking density. The addition of mineral oil to rubber compounds results in a slight reduction in mechanical properties. The type of carbon black has a much more significant influence on the mechanical properties of vulcanized composites based on natural rubber.
{"title":"RHEOLOGICAL BEHAVIOR AND MECHANICAL PROPERTIES OF RUBBER COMPOSITES BASED ON NATURAL RUBBER LOADED WITH MINERAL OILS AND PYROLYTIC CARBON BLACK","authors":"Sonja Stojanov, M. Jovičić, I. Bobinac, Olga M. Govedarica, J. Pavličević, Julijana Blagojević, Dragan D. Govedarica, O. Bera","doi":"10.46793/iccbi21.173s","DOIUrl":"https://doi.org/10.46793/iccbi21.173s","url":null,"abstract":"This paper aims to investigate the effect of the addition of mineral oil and pyrolytic carbon black on crosslinking the natural rubber and the mechanical properties of the crosslinked products. A rheometer determined curing characteristics at a temperature of 150 °C. The mechanical properties of prepared vulcanized composites were determined. By adding mineral oil to rubber compounds, the vulcanization reaction starts later, and it takes slightly more time to achieve the optimal vulcanization time. The addition of mineral oil to the rubber mixture achieves better dispersion of pyrolytic carbon blacks in the matrix and thus increases the physical interaction between the filler and rubber. Pyrolytic carbon black (pCB) is obtained by recycling waste products and contains a higher proportion of impurities. Due to impurities, PCB has a smaller surface area for the physical adsorption of rubber molecules than standard carbon black, and it can be assumed that this has led to a decrease in the crosslinking density. The addition of mineral oil to rubber compounds results in a slight reduction in mechanical properties. The type of carbon black has a much more significant influence on the mechanical properties of vulcanized composites based on natural rubber.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"31 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88152786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jelisaveta Ignjatović, T. Šušteršič, S. Cvijić, Aleksandar Bodić, Jelena Đuriš, N. Filipovic
Computational fluid dynamics (CFD) coupled with discrete phase modeling (DPM) appeared as an alternative approach to the commonly used in vitro methods for the assessment of dry powders for inhalation (DPI) aerodynamic properties. The aim of this study was to compare the parameters that describe DPI aerodynamic performance, obtained computationally by CFD-DPM and in vitro by next generation impactor (NGI). The analyzed parameters included: emitted fraction (EF), fine particle fraction (FPF), mass median aerodynamic diameter (MMAD) and geometric standard deviation (GSD). The results showed that CFD-DPM simulated EF values were generally comparable to the NGI obtained values, but there were some differences between the results obtained by these two methods. On the other hand, CFD-DPM predicted MMAD values were almost twice bigger than the NGI determined values, while the predicted GSD values were lower than NGI obtained values. In addition, CFD-DPM predicted values indicated larger differences between MMAD for different formulations in comparison to the NGI results. The largest difference between CFD-DPM and NGI results was observed for FPF values. Namely, CFD-DPM predicted FPF values were markedly lower than the NGI determined values for four of five tested formulations. Overall, although the designed CFD-DPM model and NGI measurements provided comparable data on the DPI EF values, the other relevant parameters obtained by these two approaches largely diverged indicating the need for further refinement of computational models to fully capture DPI aerodynamic performance.
{"title":"COMPUTATIONAL VS. IN VITRO APPROACH TO PREDICT AERODYNAMIC PERFORMANCE OF DRY POWDERS FOR INHALATION","authors":"Jelisaveta Ignjatović, T. Šušteršič, S. Cvijić, Aleksandar Bodić, Jelena Đuriš, N. Filipovic","doi":"10.46793/iccbi21.096i","DOIUrl":"https://doi.org/10.46793/iccbi21.096i","url":null,"abstract":"Computational fluid dynamics (CFD) coupled with discrete phase modeling (DPM) appeared as an alternative approach to the commonly used in vitro methods for the assessment of dry powders for inhalation (DPI) aerodynamic properties. The aim of this study was to compare the parameters that describe DPI aerodynamic performance, obtained computationally by CFD-DPM and in vitro by next generation impactor (NGI). The analyzed parameters included: emitted fraction (EF), fine particle fraction (FPF), mass median aerodynamic diameter (MMAD) and geometric standard deviation (GSD). The results showed that CFD-DPM simulated EF values were generally comparable to the NGI obtained values, but there were some differences between the results obtained by these two methods. On the other hand, CFD-DPM predicted MMAD values were almost twice bigger than the NGI determined values, while the predicted GSD values were lower than NGI obtained values. In addition, CFD-DPM predicted values indicated larger differences between MMAD for different formulations in comparison to the NGI results. The largest difference between CFD-DPM and NGI results was observed for FPF values. Namely, CFD-DPM predicted FPF values were markedly lower than the NGI determined values for four of five tested formulations. Overall, although the designed CFD-DPM model and NGI measurements provided comparable data on the DPI EF values, the other relevant parameters obtained by these two approaches largely diverged indicating the need for further refinement of computational models to fully capture DPI aerodynamic performance.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"19 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76033728","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Bladder cancer is one of the most common malignancies of the urinary tract. It is characterized by high metastatic potential and a high recurrence rate, which significantly complicates diagnosis and treatment. In order to increase the accuracy of the diagnostic procedure, algorithms based on artificial intelligence are introduced. This paper presents the principle of selection of convolutional neural network (CNN) models based on a multi-objective approach that maximizes classification and generalization performance. Model selection is performed on two standard CNN architectures, AlexNet and VGG-16. Classification performances are measured by using ROC analysis and the resulting AUC value. On the other hand, generalization performances are evaluated by using a 5-fold cross-validation procedure. By using these two metrics, a multi-objective fitness function, used in meta-heuristic algorithms, is designed. The multi-objective search was performed using a Genetic algorithm (GA) and a Discrete Particle Swarm (D-PS) algorithm. From obtained results, it can be noticed that such an approach has resulted in CNN models that are defined with high classification and generalization performances. When a GA-based approach is used, fitness values up to 0.97 are achieved. On the other hand, by using the D-PS approach, fitness values up to 0.99 are achieved pointing towards the conclusion that such an approach has provided models with higher classification and generalization performances.
{"title":"A META-HEURISTIC MULTI-OBJECTIVE APPROACH TO THE MODEL SELECTION OF CONVOLUTION NEURAL NETWORKS FOR URINARY BLADDER CANCER DIAGNOSIS","authors":"I. Lorencin, Klara Smolić, D. Markić, J. Španjol","doi":"10.46793/iccbi21.243l","DOIUrl":"https://doi.org/10.46793/iccbi21.243l","url":null,"abstract":"Bladder cancer is one of the most common malignancies of the urinary tract. It is characterized by high metastatic potential and a high recurrence rate, which significantly complicates diagnosis and treatment. In order to increase the accuracy of the diagnostic procedure, algorithms based on artificial intelligence are introduced. This paper presents the principle of selection of convolutional neural network (CNN) models based on a multi-objective approach that maximizes classification and generalization performance. Model selection is performed on two standard CNN architectures, AlexNet and VGG-16. Classification performances are measured by using ROC analysis and the resulting AUC value. On the other hand, generalization performances are evaluated by using a 5-fold cross-validation procedure. By using these two metrics, a multi-objective fitness function, used in meta-heuristic algorithms, is designed. The multi-objective search was performed using a Genetic algorithm (GA) and a Discrete Particle Swarm (D-PS) algorithm. From obtained results, it can be noticed that such an approach has resulted in CNN models that are defined with high classification and generalization performances. When a GA-based approach is used, fitness values up to 0.97 are achieved. On the other hand, by using the D-PS approach, fitness values up to 0.99 are achieved pointing towards the conclusion that such an approach has provided models with higher classification and generalization performances.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"7 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72817629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Emilija Milović, Nenad Ž. Janković, Jelena Petronijevic, Nenad Joksimović
Tetrahydropyrimidines (THPMs) attracted attention as a very important class of aza heterocycles with broad pharmacological activities during the past years. In many studies have been proven that THPMs have anticancer, anti-inflammatory, antimicrobial, antioxidant, antifungal, anti-HIV activity. Bearing in mind our interest in medicinal and Biginelli chemistry, we investigated interaction with important biomacromolecules (DNA, BSA) and our earlier synthetized THPMs derivatives with proven very good cytotoxic activity.[1] Investigation of affinity of compounds A and B (Figure 1) to bind to bovine serum albumin (BSA) is based on the fact that the efficiency of drugs depends on their ability to bind for carrier protein. Binding properties were investigated by using the fluorescence emission titration of BSA with A and B. The obtained values of Ka, which are in optimum range which is considered to be 106-107M-1 indicate that both compounds have great ability to bind to BSA. In addition, Ka values for A-BSA and B-BSAshow that both compounds are suitable for drug-cell
{"title":"CHEMICO-BIOLOGICAL INTERACTION OF SELECTED TETRAHYDROPYRIMIDINES","authors":"Emilija Milović, Nenad Ž. Janković, Jelena Petronijevic, Nenad Joksimović","doi":"10.46793/iccbi21.347m","DOIUrl":"https://doi.org/10.46793/iccbi21.347m","url":null,"abstract":"Tetrahydropyrimidines (THPMs) attracted attention as a very important class of aza heterocycles with broad pharmacological activities during the past years. In many studies have been proven that THPMs have anticancer, anti-inflammatory, antimicrobial, antioxidant, antifungal, anti-HIV activity. Bearing in mind our interest in medicinal and Biginelli chemistry, we investigated interaction with important biomacromolecules (DNA, BSA) and our earlier synthetized THPMs derivatives with proven very good cytotoxic activity.[1] Investigation of affinity of compounds A and B (Figure 1) to bind to bovine serum albumin (BSA) is based on the fact that the efficiency of drugs depends on their ability to bind for carrier protein. Binding properties were investigated by using the fluorescence emission titration of BSA with A and B. The obtained values of Ka, which are in optimum range which is considered to be 106-107M-1 indicate that both compounds have great ability to bind to BSA. In addition, Ka values for A-BSA and B-BSAshow that both compounds are suitable for drug-cell","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"74 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73510831","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. Grujović, K. Mladenović, Z. Simić, Simona R. Đuretanović
This research aimed to investigate the quality of drinking water from the rural area of village Pajsijević (Šumadija, central Serbia). The water is consumed as raw since it is not purified or chlorinated before consumption. The water was collected at three sampling points – in the spring of Kotlenik Mountain stream (W1 sample), in the local reservoir (W2 sample), and from the tap (W3 sample). Also, the sediment samples (soil and sand) were analyzed, too. The health risks related to the presence and concentration of some major and trace elements (Ca, Cr, Cu, Fe, Mg, Mn, Ni, Zn) and N, NO, NN3, NH4, P, P2O5, and PO4 were evaluated. Additionally, the presence and the number of total coliform bacteria and Escherichia coli (as an indicator of fecal contamination) were evaluated. The concentrations of analyzed major and trace elements in all water samples were below those at which toxic effects may occur. The exception was the concentrations of Fe (2.02 – 2012 mg/L), which were higher than is allowed. The origin of Fe in water is from sediment (soil and sand), which also showed high content of Fe (3006.0 mg/g and 2229.9 mg/g, respectively). The results of the Colorimetric test indicated the presence of coliform bacteria as well as the presence of E. coli in all water samples. Further research needs to include characterization of isolated coliform bacteria and serological investigation of E. coli strains in order to evaluate the risks of consumption related to waterborne illness.
{"title":"CONSUMPTION OF RAW WATER – THE HEALTH RISKS RELATED TO THE PRESENCE OF HEAVY METALS AND „ESCHERICHIA COLI“","authors":"M. Grujović, K. Mladenović, Z. Simić, Simona R. Đuretanović","doi":"10.46793/iccbi21.210g","DOIUrl":"https://doi.org/10.46793/iccbi21.210g","url":null,"abstract":"This research aimed to investigate the quality of drinking water from the rural area of village Pajsijević (Šumadija, central Serbia). The water is consumed as raw since it is not purified or chlorinated before consumption. The water was collected at three sampling points – in the spring of Kotlenik Mountain stream (W1 sample), in the local reservoir (W2 sample), and from the tap (W3 sample). Also, the sediment samples (soil and sand) were analyzed, too. The health risks related to the presence and concentration of some major and trace elements (Ca, Cr, Cu, Fe, Mg, Mn, Ni, Zn) and N, NO, NN3, NH4, P, P2O5, and PO4 were evaluated. Additionally, the presence and the number of total coliform bacteria and Escherichia coli (as an indicator of fecal contamination) were evaluated. The concentrations of analyzed major and trace elements in all water samples were below those at which toxic effects may occur. The exception was the concentrations of Fe (2.02 – 2012 mg/L), which were higher than is allowed. The origin of Fe in water is from sediment (soil and sand), which also showed high content of Fe (3006.0 mg/g and 2229.9 mg/g, respectively). The results of the Colorimetric test indicated the presence of coliform bacteria as well as the presence of E. coli in all water samples. Further research needs to include characterization of isolated coliform bacteria and serological investigation of E. coli strains in order to evaluate the risks of consumption related to waterborne illness.","PeriodicalId":9171,"journal":{"name":"Book of Proceedings: 1st International Conference on Chemo and BioInformatics,","volume":"90 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78427102","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：