Baghdad Science Journal

The present research includes a comparative study of preparing mesoporous silica using two different methods: the chemical method and the green method. The sol-gel method was used using sodium silicate (glass water) as a cheap source of silica and the cationic-type surfactant (Hydroxylethyl, 2-Hydroxylcetyl (2-Dimethyl ammonium chloride) in the chemical method, and Zyziphus Spina Christi (Sider tree) tree leaf extract in the green method, as a soft template. The mesoporous silica prepared using the two methods were compared by characterizing them using different methods: measuring surface area by BET method and porosity by BJH, XRD, SAXRD, AFM, FTIR, SEM, and EDX. The results showed that the silica prepared by the green method had the best surface area, recording 800 m2g-1 and an average pore diameter equal to 2.5 nm, while XRD measurements showed that both methods produced non-crystalline (amorphous) silica, while SAXRD showed that the green method produced more regular silica. SEM images displayed gr-SiO2 particles were comparatively smaller than the particles of ch--SiO2 with spherical shape particles and agglomeration as a flower shape, while EDX results confirmed the presence of Si and O, with small amount of C which may be due to non-complete decomposition of surfactants. The Langmuir, Freundlich, and Temkin isotherm models were fitted to the data obtained from the experiments and the results of adsorption study shows that the chemical prepared silica has higher capacity (63.694 mg g-1) toward amlodipine adsorption according to Freundlich parameters obtained by linear fitting.

The purpose of this study is to find the optimal conditions for producing wood vinegar from cacao pod shells using Response Surface Methodology (RSM). Box Behnken Design (BBD) was used to optimize the variables of pyrolysis times, temperature, and particle size. Wood vinegar was prepared by slow pyrolysis technology. The RSM with three design variables including pyrolysis duration at 50, 75, and 100 minutes; pyrolysis temperatures at 300°C, 340°C, and 380°C; and particle size of 3, 5 and 7 cm with 15 runs were applied. Chemical characterization of wood vinegar was performed by Gas Chromatography-Mass Spectrometry (GC-MS) analysis. The results showed that phenolic, furan, and ketone compounds were the main component in the cacao pod shells vinegar. The optimum production of wood vinegar from cacao pod shells based on RSM results was identified at a pyrolysis time of 100 minutes, a temperature of 353°C, and a particle size of 3 cm.

Biodiesel produced through transesterification reactions with homogeneous and heterogeneous base catalysts is a promising alternative energy source. This study uses Precipitated Calcium Carbonate (PCC) enriched with TiO2 to produce biodiesel from waste vegetable oil. The synthesized catalyst was comprehensively identified through XRD, XRF, SEM, BET, and FT-IR analyses and then applied in the transesterification reaction. The results showed that PCC enriched with TiO2 is an effective catalyst, producing biodiesel with an efficiency of 95.09% under optimal conditions: 5 wt% catalyst, a methanol to oil ratio of 6:1, a reaction temperature of 65°C, and a reaction time of 4 hours. Enhancing catalytic properties, increased surface area, and strengthening of basic sites significantly improved biodiesel yield.

The method of removing nitrates from water using zero-valent metals is one of the methods that proves to be highly effective day after day, especially zero-valent iron. However, the most prominent obstacles that cause an obstacle to continuing its use are the exposure of its surface to passivation during the reaction and the production of ammonium ions as a main product for the reduction of nitrates under conditions of treatment.
In this research, the effect of clay during the removal of nitrate by Fe0 was studied. Raw clay, acid-activated clay, and clay modified with iron nanoparticles at a ratio of (1:1) were used. The results showed that 80% of nitrate ions in water can be removed by using Fe0 at a concentration of 40 g/L with 4 ≥ pH for 120 min at room temperature, adding raw clay reduced the efficiency due to its alkaline properties, which led to the inhibition reaction of Fe0 with nitrates. while Acid-activated clay led to an increase in nitrate removal by increasing dosage.
Modified clay has given greater effectiveness than raw clay and acid-activated clay, removal rate increased f to 95.33% under conditions of treatment, while ammonium was decreased with increasing the added clay in all its forms (raw, activated and modified). Finally, its noted that adding 5 gr/L of clay (raw, activated, modified) was sufficient to reduce the percentage of nitrate and ammonium within the permissible limits according to Syrian Specification No. 45/2007 for drinking water.

The current study was designed to investigate the impact of the Single Nucleotide Polymorphism (SNP), rs5918 (T>C), of integrin beta 3 genes (ITGB3) and rs1801106 (G>A) of integrin alpha 2 genes ( ITGB3 ) in peripheral blood samples of Iraqi female; 71 infertile females under IVF protocol ( divided into two subgroups, 29 success implantation and 42 failure implantation) and 50 fertile females as control. From March 2021 to August 2021, blood samples were taken from the patients and control groups at Roh al-Hayat center in Baghdad, Iraq. The genotyping of SNPs was carried out by using high-resolution melting real-time polymerase chain reaction (HRM real-time PCR) of the purified genomic DNA obtained from blood samples. The findings demonstrated that fertile females had significantly greater levels of the TT and GG genotypes of rs5918 at ITGB3 and rs1801106 at ITGA2, respectively, than infertile females. that shows the GG genotype functions as a protective factor, while the TC genotype of rs5918 at ITGB3 and AA genotype of rs1801106 at ITGA2 act as risk factors. In conclusion, these SNPs in the ITGA2 and ITGB3 genes may have an influence on the predisposition of a certain group to infertility.

Over five decades, eastern eelgoby, Taenioides mordax (De Vis, 1883) remained native to Cairns to Townville (Queensland) and Sydney to Newcastle (New South Wales) Australia. In this study, two specimens, measuring 19.8 cm in total length and weighing 20.25 grams, were collected from Nareri lagoon, Badin, Sindh. The specimens were collected with an encircled small gillnet installed over the night in shallow water. The body shape of the eelgoby is elongated, scaleless, tiny eyes, mouth upturned with large-sharp canine teeth 12/10 on the upper and lower jaws, respectively. The single continuous dorsal fin, anal fin connected with caudal fin and reaching to opening ventrally, pectoral fin and pelvis are short. The shape of the sagitta is oblong, margin smooth, and prominent rostrum, whereas the anti-rostrum is understated. On the posterior side, a distinct grove can be seen. The sulcus acusticus, ostium and cauda are invisible. Based on these features, the specimens were confirmed eastern eelgoby. To the best of our knowledge, this is the second confirmed record of this species found in the world.

Anatomical changes in internal tissue of stem and leaf when seed and plant treated with acids to enhance growth and development in maize was studied during the spring seasons of 2019 and 2020. Randomized complete block design was used with three replications. Main plots received foliar nutrition treatments, including ascorbic acid (AA), citric acid (CA), and humic acid (HA) at concentrations of 100 mg L-1, alongside HA at 1 ml L-1, with distilled water as the control. Sub-plots underwent corresponding treatments for seed soaking. Results indicated variations in vascular bundle size among treatments, with foliar CA treatment showing superior results in both years, as well as seed soaking in CA and HA. Interaction effects were observed, notably in 2019 with the combination of foliar CA and seed soaking with distilled water, and in 2020 with HA. Effects on leaf epidermis were minimal, with slight distortions in stomatal shapes observed with AA and CA treatments compared to the control. AA and HA treatments led to larger ordinary epidermal cells with straighter cell walls than the control, along with an increase in cork and silica cell size in treated plants. This study contributes to understanding anatomical modifications in maize leaves and stems during the growing season, shedding light on the potential impacts of acid treatments on plant physiology.

Anatomical changes in internal tissue of stem and leaf when seed and plant treated with acids to enhance growth and development in maize was studied during the spring seasons of 2019 and 2020. Randomized complete block design was used with three replications. Main plots received foliar nutrition treatments, including ascorbic acid (AA), citric acid (CA), and humic acid (HA) at concentrations of 100 mg L-1, alongside HA at 1 ml L-1, with distilled water as the control. Sub-plots underwent corresponding treatments for seed soaking. Results indicated variations in vascular bundle size among treatments, with foliar CA treatment showing superior results in both years, as well as seed soaking in CA and HA. Interaction effects were observed, notably in 2019 with the combination of foliar CA and seed soaking with distilled water, and in 2020 with HA. Effects on leaf epidermis were minimal, with slight distortions in stomatal shapes observed with AA and CA treatments compared to the control. AA and HA treatments led to larger ordinary epidermal cells with straighter cell walls than the control, along with an increase in cork and silica cell size in treated plants. This study contributes to understanding anatomical modifications in maize leaves and stems during the growing season, shedding light on the potential impacts of acid treatments on plant physiology.

The polymorphism in the vitamin D receptor gene FokI position is used to evaluate the polymorphism impact on the levels of vitamin D, testosterone and prolactin hormones in the sera of patients with prostate cancer and benign prostatic hyperplasia vs. healthy controls. The vitamin D receptor gene Fok1 restriction site was amplified and examined by TaqMan RT-PCR technique. It was found that the TT genotype played a protective effect in 70% and 50% in prostate cancer and benign prostatic hyperplasia patients respectively. While, the CC genotype was found to be 100% disease-attributed genotype in both prostate cancer and benign prostate hyperplasia. Also, the distribution of genotypes (TT, TC and CC) was not consistent with Hardy Weinberg equation in the patients with prostate cancer as a significant difference was found by chi-square test (X2 ˃3.84) at P ≥0.05 between the observed and expected frequencies. But wasn’t seen in patients with BPH or control group. The level of vitamin D was significantly affected by the genotype CC of VDR-FOK I in prostate cancer patients compared with TT and TC genotypes. There were no significant differences in Vit. D level among the three genotypes in the patients with BPH and the healthy control group. In association with genotypes, the levels of testosterone and prolactin did not differ significantly among the studied groups. It could be concluded that the vitamin D receptor FokI polymorphism is associated with Iraqi prostate cancer patients more than in benign prostate hyperplasia with vitamin D deficiency in blood serum

Many parts of the world are facing water scarcity problems due to the limitation of water resources coinciding with the increasing population. Rivers are vital freshwater systems that are critical for the sustenance of life. This work deals with the monitoring and assessment of water quality of the Euphrates River between Karbala and Babylon provinces from ten sites near the discharge of industrial facilities, drainage of agricultural lands, wastewater and other facilities, samples were taken monthly and the results were converted into four seasons, it’s divided into two seasons (dry- wet). Various physico-chemical and microbial analyses were performed based on standard methods. The following parameters were measured (pH, EC,BOD, DO,WATERTEM, TDS, TOTAL COLIFORM, TH, SO4, NO3, CL, Na, PO4, RSC, SAR, SSP, MH, B,NO2 and Mg). By using A Novel Iraqi water quality index that fits the Iraqi aquatic systems and reflects the environmental reality of Iraqi water, the developed IWQI includes above components, The results showed that the water quality of the Euphrates River in the studied sites are ranged from ( poor - very poor) for fresh water. As for the suitability of the water for agriculture and living for aquatic organisms, and its suitability for drinking, the results for the sites ranged from (good - unsuitable) during the wet and dry seasons. Based on these indices, it is concluded that industrial facilities, city wastewater and intensive communities that live along the river bank are negatively affecting the water quality of rivers.

Because of industrial and human activities, lead is a hazardous heavy metal that is widely present in the environment and poses a threat to public health and ecosystems. Garlic extract (GE) demonstrates protective impacts against many disorders. Its effects on heavy metal toxicity require more research. A nanoparticle formulation could increase its effectiveness even more. This investigation examined the repro-protection impact of tocopherol polyethylene glycol succinate-coated garlic selenium (TPGS-GSNP) against lead acetate (LA) toxicity. For this purpose, 64 mature male rabbits were equally divided into eight groups (8 each). Male rabbits were treated three times a week for 12 weeks. The males received drinking water (G1), 30 mg of LA/ kg. BW (G2), 800 mg of GE/kg. BW (G3), GE and LA (G4), 1 mg of TPGS-Selenium/kg. BW (G5), TPGS-S and LA (G6), 1 mg of TPGS-GSNP/kg. BW (G7), and TPGS-GSNP and LA (G8). After treatment, serum levels of follicle stimulating hormone, luteinizing hormone, testosterone, and gonadotropin releasing hormone were assessed. Testicular histopathological findings and the expression levels of testicular FSH and LH receptors, ABP, 3β-HSD, 17β-HSD, and Inh-α, and pituitary LHβ and FSHβ genes were evaluated. Obvious testicular histopathological changes, decreased genital organ weights, reduced serum hormonal levels, and declined testicular and pituitary gene expression levels, were exhibited in LA group. The TPGS-GSNP group, however, displayed improvement in these findings. Finally, synthesized TPGS-GSNP showed a greater repro-protective effect against LA toxic-induced male rabbits. Further research is required to explore the mechanism of TPGS-GSNP action.

Rainfall time series are considered a parameter that can provide clues to climate change. Based on data from 39 stations, monthly rainfall to cover all regions of Iraq for a period (1980–2021) was analyzed statistically by using Microsoft Office Excel 2013. Rainfall patterns and the factors affecting them is determined using (GIS).The results show that Iraqi rainfall is highly dynamic and their pattern are not a cyclical pattern. The amount of rain decreases as we orient from the north towards the south, with the highest mean reaching in Ducan Station then decreasing as it goes southward, reaching Nukaib Station. There is a decreasing trend in rainfall for all stations, except for the stations (Baghdad, Kut, Hilla, Amarah, and Samawah). According to the calculation, the 2000s were drier than normal, the driest year was observed in 2021 over 42 years. The stations (Ducan, Amadiyah, Sulaymaniyah, Zakho, Dohuk and Salah al-Din) recorded the highest amounts of rain among the 39 stations, have the highest standard deviation values compared to the rest of the stations, have recorded a decrease from the averages greater than the other stations. We saw signs of a shift to a drier climate, especially in the stations that represent Northeast. The Pearson coefficient (R) was used to test the relationship between rainfall and elevation, and it was found that a strong relationship found to be 0.71. This indicates that elevation is not the only variable that affects the rainfall process, but it is one of the important factors.

Rainfall time series are considered a parameter that can provide clues to climate change. Based on data from 39 stations, monthly rainfall to cover all regions of Iraq for a period (1980–2021) was analyzed statistically by using Microsoft Office Excel 2013. Rainfall patterns and the factors affecting them is determined using (GIS).The results show that Iraqi rainfall is highly dynamic and their pattern are not a cyclical pattern. The amount of rain decreases as we orient from the north towards the south, with the highest mean reaching in Ducan Station then decreasing as it goes southward, reaching Nukaib Station. There is a decreasing trend in rainfall for all stations, except for the stations (Baghdad, Kut, Hilla, Amarah, and Samawah). According to the calculation, the 2000s were drier than normal, the driest year was observed in 2021 over 42 years. The stations (Ducan, Amadiyah, Sulaymaniyah, Zakho, Dohuk and Salah al-Din) recorded the highest amounts of rain among the 39 stations, have the highest standard deviation values compared to the rest of the stations, have recorded a decrease from the averages greater than the other stations. We saw signs of a shift to a drier climate, especially in the stations that represent Northeast. The Pearson coefficient (R) was used to test the relationship between rainfall and elevation, and it was found that a strong relationship found to be 0.71. This indicates that elevation is not the only variable that affects the rainfall process, but it is one of the important factors.

In this study, CQDs were synthesized by the green method using orange juice and ethanol, confused at a relatively low temperature by hydrothermal carbonization process, TiO2 were prepared by a facile sol-gel method, and TiO2-CQDs nanocomposites with different weight ratios were prepared by slipe mixing and heat treatment. XRD showed that the CQDs exhibited a broad peak at (002) with hexagonal structure, and TiO2 (anatase phase) had a polycrystalline nature with tetragonal structure. The FESEM results showed the formation of nanostructures with different shapes and small average particle size. High-resolution transmission electron microscopy revealed that the TiO2 (anatase phase) agglomerated in mostly spherical shapes and sizes less than 15 nm. The CQDs had a relatively uniform diameter, a spherical shape with a highly crystalline structure, and a size below 5 nm. . The FTIR spectra of TiO2 NPs, CQDs, and TiO2-CQDs nanocomposite showed the presence of a broad band at 450–4000 cm-1, which corresponded to the stretching vibration of terminating hydroxyl groups in samples. The results of UV-visible spectroscopy showed that the absorbance of TiO2-CQDs nanocomposite increased with the increase in the CQDs rate, and the optical energy band gap of TiO2 and CQDs was 3.14 ev and 3.07 eV, respectively. The energy band gap values of TiO2-CQDs nanocomposite decreased with the increase in the CQDs rate in the range of (2.72-1.85) eV. The performance of the Photocatalytic was shown by decreasing methylene blue (MB) and methyl orange (MO) under UV irradiation. The results showed that the TiO2-CQDs nanocomposite with different weight ratios had higher photocatalytic efficiency than TiO2 NPs, and the photocatalytic efficiency increased with the increase in the CQDs rate. The degradation efficiencies of MB and MO were high at 84% and 39% within 240 min, respectively.

Promoting sustainable living in the future involves efficiently managing waste to reclaim resources and convert them into new raw materials. There is a global focus on accelerating the utilization and recycling of waste to create valuable biomaterials, driving significant research efforts in this area. In this research, pure silica was isolated from glass waste and used in the manufacture of a type of bioceramic for medical applications. Several sequential stages were followed, starting with collecting samples of crushed glass and treating them physically (grinding, granular sorting), mixing the ground glass with NaOH in the following proportions (1:1), (2:1), and (3:1), then heat treatment was carried out at different temperatures (1000–900–800) ˚C, then the silica was isolated by treating the sample with HCl solution, followed by heat treatment at 900 degrees to obtain SiO2 with a yield of 68% and purity reaching 99.4%. Isolated pure silica was used to prepare standard bioglass 45S5. Specifications of prepared bioglass were determined using XRD, IR, and DTA, then the bioglass was converted into bioceramic through heat treatment according to DTA to ultimately obtain a product whose properties, composition, and susceptibility to biodegradation were studied. Results indicate that it was usable in medical applications.

This paper presents an innovative approach for thyroid disease detection leveraging the capabilities of Laser-Induced Breakdown Spectroscopy (LIBS). Thyroid diseases are a global health issue with a significant impact on millions of people worldwide. Traditional diagnostic methods, such as blood tests, can be invasive, costly, and time-consuming. In contrast, LIBS provides a quick, non-destructive, and cost-effective approach to illness identification. The study uses LIBS to analyze human hair's elemental composition, revealing elemental signatures linked to thyroid disease through laser ablation and spectral emissions. This method provided data that demonstrated the distinction between the control and infected patients by precisely identifying elemental changes in hair samples. The spectra lines of copper (Cu), zinc (Zn) and selenium (Se) were determined to identify thyroid disorders. The diagnosis and therapy of thyroid problems may be completely changed with LIBS. This non-invasive method not only enhances patient outcomes but also creates opportunities for future developments in personalized medicine and medical spectroscopy.

In this study, nano textile fibers were prepared from poly vinyl alcohol (PVA) and multi wall carbon nano tubes (MWCNTs) via electro spinning method. Different proportions of the multi wall carbon nano tubes were used, including 2, 4, and 6 % weight percent. The optical, electrical, and thermal properties were studied. Scanning electron microscopy was used to study the morphology of samples. Fourier-transform infrared spectroscopy (FTIR) analysis was used to study the bonds of the prepared samples. Differential scanning calorimetry (DSC) was used to investigate the thermal properties including the glass transition temperature and melting the point. Ultraviolet spectroscopy was used to study the optical properties. The results showed that the addition of carbon nano tubes in small proportions of 6% wt leads to an improvement in the thermal properties represented by increasing the glass transition temperature value from 85.17 o C of poly vinyl alcohol to reach 107 o C with a strengthening percentage of 6 weight percent. The electrical conductivity improves with the addition of 6 weight percent multi wall carbon nanotubes from 14.28 Ω-1cm-1 to 20 Ω-1. cm-1. The results of the UV spectrum indicate that the inclusion of multi wall carbon nano tubes control the energy band gap. Absorbance and coefficient of absorbance also increase with increasing the carbon nano tubes contents. The best sample had a band energy gap of roughly 1.75 eV which consisted of poly vinyl alcohol with 6 weight percent of multi wall carbon nano tubes.

The duo module plays an important role in the module theory. Many researchers generalized this concept such as Ozcan AC, Hadi IMA and Ahmed MA. It is known that in a duo module, every submodule is fully invariant. This paper used the class of St-closed submodules to work out a module with the feature that all St-closed submodules are fully invariant. Such a module is called an Stc-duo module. This class of modules contains the duo module properly as well as the CL-duo module which was introduced by Ahmed MA. The behaviour of this new kind of module was considered and studied in detail, for instance, the hereditary property of the St-duo module was investigated, as the result; under certain conditions, every St-closed submodule of an St-duo module is also St-duo. Another characterization of the Stc-duo module was given. Additionally, the relationships of St-duo among some types of modules were investigated and discussed, for example; In the class of semi-extending modules, every weak duo module is an Stc-duo module. Also, the authors gave a case in which St-duo, duo, CL-duo and weak duo are equivalent. Furthermore, the St-duo module was used to make the concepts semi-extending and FI-extending equivalent.

Calculating the reliability of complex systems is an urgent matter that deserves attention due to its wide applications, ranging from engineering and economic sciences to applications in medical fields Traditional methods often suffer from computational complexity when dealing with large systems when calculating their reliability. This paper presents a new method to calculate and improve the reliability of highly complex systems. Furthermore, the proposed methodology combines the principles of system reliability analysis and optimization techniques to determine the optimal means for a system that increases reliability. A mathematical optimization approach was employed to undertake the task of improving the reliability of the system without complicating the calculations. By formulating the problem as an optimization task, the reliability of vehicles is optimized to meet specified reliability constraints. The effectiveness of this approach is demonstrated by experimental evaluations on various complex systems, demonstrating significant improvements in system reliability. This new approach was tested on an entire highly complex network of 1,225 vehicles, and the results were very acceptable. Finally, the proposed method was applied and numerical optimization results were obtained using the programming language Python version 3.12.2.

Cryptography and steganography play critical roles in ensuring the security of network communications. Combining these methods holds great potential for safeguarding information transmitted over the internet. DNA Cryptography, a modern and robust technique, leverages the unique properties of DNA for secure data handling. The increasing of cyber threats has made it necessary to propose a new secure communication method to block unauthorized access to the sending information. This paper introduces an innovative system that integrates cryptography, steganography, and machine learning techniques to enhance data transfer security over the Internet. The system unfolds in six steps to encrypt and hide text. The initial step involves using the Caesar cipher to encode the original text. This is followed by the conversion of the text into DNA bases. Subsequent steps include the conversion of DNA bases to ASCII format and further transformation into binary numbers. The fifth step introduces a dynamic element by shifting binary numbers using a random key. The final step involves covertly embedding these binary numbers (ciphertext) within an image. Beyond traditional metrics, machine learning elements have been incorporated to elevate the system's performance. Performance evaluation was conducted across three standard images with varying data sizes, demonstrating the system's effectiveness. The proposed system showcased rapid cryptography times, with encryption and decryption times of 2.802 ms and 3.388 ms, respectively. The integration of machine learning techniques enriches the system's capabilities, presenting a compelling solution for secure and efficient data transfer over the Internet.

There has been a COVID-19 pandemic since December 2019, and successful medical treatment for COVID-19 patients requires rapid and accurate diagnosis. Fighting the COVID-19 pandemic requires an automated system that uses deep transfer learning to diagnose the virus on chest X-ray (CXR). CXR are frequently utilized in healthcare because they offer the potential for rapid and accurate disease diagnosis. Automated computer-aided diagnosis (CAD) systems incorporate ML or deep learning to enhance efficiency and accuracy, hence reducing future problems. Numerous AI systems based on deep learning can be employed for diagnosis; among the most widely used is the CNN, which was first developed and has demonstrated encouraging accuracy in identifying COVID-19 confirmed patients using CXR pictures. Through using X-ray images, this work will design ML and deep learning to provide faster diagnostics for Covid-19 infection. As a result, the deep transfer learning technique uses an existing model first, then applies the needed data to it again. Where a Densenet201transfer learning model was utilized, which is one of a DL techniques, as feature extraction and its combination with multilayer perceptron algorithm; these technique were applied to a data set of a National Institute Health (NIH), where several performance measures were utilized, such as precision, precision, specificity and sensitivity, as an experiment proved the efficiency of the algorithm used in terms of accuracy by 98.82%. These outcomes are encouraging when compared to other DL models that were trained on the identical dataset.

Identification of medicinal plant leaves is very crucial as their cultivation and production are essential for the medicine industry. Many different classes of medicinal leaves look identical but serve different purposes in the medicine industry and have different remedies for different diseases. Hence it is imperative to use methods that are automated, faster, and produce good accuracy. Cutting-edge models have been trained to discern the subtle distinctions between various species of leaves, accounting for a myriad of factors such as leaf texture, shape, and color variations, which are often imperceptible to the human eye. In this research, Transfer learning (TL) based VGG16 and Vision Transformer (ViT) models such as ConvMixer and Compact Convolutional Transformer (CCT) are implemented for the classification of medicinal leaf images using a dataset of 38066 leaf images having 10 different classes. The proposed customized Convolutional Neural Network (CNN) and hybrid CNN-ViT models both have a very low number of parameters compared to the other models in comparison making them light and capable of being less computationally expensive. In the experimental evaluation, all the results are collected for 30 epochs. VGG16, CCT, and ConvMixer produce AUC scores of 0.50, 0.79, and 0.50, respectively for the dataset while the proposed CNN and hybrid model gave AUC scores of 0.83 and 0.74, respectively. In addition, a hybrid denoising approach with Wavelet thresholding and Gaussian blurring is utilized to minimize the noises in the images by retaining the original image quality.

Loss functions are used for estimating the accuracy of the classification models in Machine Learning. The error loss value measures whether the predicted labels match the true labels or are close to them. Handwritten character recognition is a pattern recognition problem. The progression of pattern recognition applications for offline image classification is rapidly improved by using deep learning techniques. Convolution Neural Networks (CNN) and the activation function are the most commonly used in deep learning. In addition, different loss functions are used to evaluate the performance of the model such as Categorical Cross-Entropy (CCE) and Mean Square Error (MSE). In deep learning techniques, the size of the datasets plays an important role in obtaining high performance, but with traditional MSE, the loss values reach zero in the very early stages of the training process when the dataset size is large, and yet the model accuracy is still in progress. This study proposes a developed loss function via enhancing MSE. The proposed improved MSE is based on dividing the square error by the sum of the predicted label probabilities instead of the total of the sample number. Five datasets are used to test the performance of the proposed modified MSE with the proposed CNN pipeline model in addition to the modified VGG16. The datasets are AHCD, AIA9K, HIJJA, Self-collected, and MNIST. The loss rates of the proposed loss function showed a significant improvement in the accuracy rates for all datasets with the improved MSE in comparison to the CCE.