Browsing by Author "Fatema, Kaniz"

Now showing 1 - 20 of 31

A comprehensive study on irritable bowel syndrome using the IBS-36 quality of life measurement tool and patients’ attitudes, concerns, and knowledge
(Daffodil International University, 2024-03-31) Fatema, Kaniz
This cross-sectional study conducted at Mugda Medical College and Hospital in Dhaka, Bangladesh, explores various aspects related to Irritable Bowel Syndrome (IBS), including knowledge, attitudes, concerns, and QoL. The study population comprises adults aged 20 years and above, either diagnosed with IBS or experiencing symptoms consistent with the condition. A sample size of 200 participants was determined using a standard formula. Data were collected through a structured questionnaire, and pre-constructed questionnaires by Lacy et al. (2007) and IBS-36 by Groll et al. (2002) were utilized to assess knowledge, attitudes, concerns, and quality of life (QoL). The mean age of the participants was 38.27 years, with a balanced distribution across age groups, ranging from 20 to 59 years. The gender distribution indicated a slight predominance of females (62.5%). 64.5% of respondents were overweight, while 14.5% were classified as obese. 70.5% of respondents resided in urban areas, 77.0% were married. Among the identified IBS signs and symptoms, majority of the patients had constipation (34.5%). Colonoscopy being the most frequently (48.0%) employed in diagnosing of IBS. 32.0% experiencing symptoms for 3-5 years and 94.0% of respondents reported using IBS medications. Additionally, dietary restrictions were common, with 58.0% restricting dairy products. 23.5% of participants fall into the Poor category, indicating high IBS severity and A majority of respondents, comprising 56.5%, exhibit a good category, denoting low IBS severity. 62.0% of the respondents, falls under the Good Knowledge category. The findings reveal a statistically significant association between knowledge levels and IBS severity (p value = 0.001). Only the Social Functioning scale demonstrates lower reliability (α = 0.104), as does the Pain scale (α = 0.284), suggesting potential limitations in patients’ quality of life. The study also revealed significant associations
A Computer-Aided Diagnostic System to Identify Diabetic Retinopathy, Utilizing a Modified Compact Convolutional Transformer and Low-Resolution Images to Reduce Computation Time
(MDPI, 2023-05-28) Khan, Inam Ullah; Raiaan, Mohaimenul Azam Khan; Fatema, Kaniz; Azam, Sami; Rashid, Rafi Ur; Mukta, Saddam Hossain; Jonkman, Mirjam; Boer, Friso De
Diabetic retinopathy (DR) is the foremost cause of blindness in people with diabetes worldwide, and early diagnosis is essential for effective treatment. Unfortunately, the present DR screening method requires the skill of ophthalmologists and is time-consuming. In this study, we present an automated system for DR severity classification employing the fine-tuned Compact Convolutional Transformer (CCT) model to overcome these issues. We assembled five datasets to generate a more extensive dataset containing 53,185 raw images. Various image pre-processing techniques and 12 types of augmentation procedures were applied to improve image quality and create a massive dataset. A new DR-CCTNet model is proposed. It is a modification of the original CCT model to address training time concerns and work with a large amount of data. Our proposed model delivers excellent accuracy even with low-pixel images and still has strong performance with fewer images, indicating that the model is robust. We compare our model’s performance with transfer learning models such as VGG19, VGG16, MobileNetV2, and ResNet50. The test accuracy of the VGG19, ResNet50, VGG16, and MobileNetV2 were, respectively, 72.88%, 76.67%, 73.22%, and 71.98%. Our proposed DR-CCTNet model to classify DR outperformed all of these with a 90.17% test accuracy. This approach provides a novel and efficient method for the detection of DR, which may lower the burden on ophthalmologists and expedite treatment for patients.
A Lightweight Robust Deep Learning Model Gained High Accuracy in Classifying a Wide Range of Diabetic Retinopathy Images
(IEEE, 2023-05-01) Raiaan, Mohaimenul Azam Khan; Fatema, Kaniz; Khan, Inam Ullah; Azam, Sami; Rashid, Md. Rafi Ur; Mukta, Md. Saddam Hossain; Jonkman, Mirjam
Diabetic retinopathy (DR) is a common complication of diabetes mellitus, and retinal blood vessel damage can lead to vision loss and blindness if not recognized at an early stage. Manual DR detection using large fundus image data is time-consuming and error-prone. An effective automatic DR detection system can be significantly faster and potentially more accurate. This study aims to classify fundus images into five DR classes, using deep learning methods, with the highest possible accuracy and the lowest possible computational time. Three distinct DR datasets, APTOS, Messidor2, and IDRiD, are merged, resulting in 5,819 raw images. Before training the model, various image preprocessing techniques are applied to remove artifacts and noise from the images and improve their quality. Three augmentation techniques: geometric, photometric, and elastic deformation, are used to create a balanced dataset. A shallow convolutional neural network (CNN) is developed using three blocks of convolutional layers and maxpool layers with a categorical cross-entropy loss function, Adam optimizer, 0.0001 learning rate, and 64 batch size as a base model, and this is also employed to determine the best data augmentation method for further processing. A study to optimize the performance is then conducted by changing different components and hyperparameters of the base model, resulting in our proposed RetNet-10 model. Six cutting-edge models are employed for comparison. Our proposed RetNet-10 model performed the best, with a testing accuracy of 98.65%. MobileNetV2, VGG16, Xception, VGG19, InceptionV3 and ResNet50 achieved testing accuracies of 91.42%, 90.16%,89.57%, 88.21%, 87.68% and 87.23%, respectively. The model is also trained with several k values to assess its robustness. After image processing and data augmentation, using the combined dataset, and fine-tuning the base model, our proposed RetNet-10 model outperformed other automated methods for DR diagnosis.
A New Experimental Study on Fly Ash, Lime and Dredged Sediment Mixed Block as an Alternative Eco-Friendly Building Materials
(Scopus, 2021) Fatema, Kaniz; Rahman, Mahbubur; Sarker, Md. Akhter Hossain; Haque, M. Aminul; Islam, Md. Shihabul
Fly ash is a byproduct in the thermal power plant station that is produced from combustion of coal and identified as hazardous material. Many global researchers have been attempted over the long years for effective utilization of fly ash a sustainable construction material. In Bangladesh, only 10%-15% fly ash is used in cement manufacturing industry for producing the Portland composite cement (PCC) and concrete industry to be cost effective as well as minimize the costly disposal processes. This paper reported the experimental investigations that were carried out to study the effectiveness of fly ash, lime and dredges sediment on the strength development of block as an alternative building material. Fly ash was partially replaced with four percentages (30%, 35%, 40% and 45%) and again was replaced with other four percentages (25%, 30%, 35% and 40%) including 5% of lime by weight. The study observed the greater strength and water absorption resistant performance on the specimen made by 35% fly ash and 5% lime at 28 days curing age. The outcomes of the study can be applied in the construction sectors for making decision to use the studied building materials for cost-effective and eco-friendly manner.
A Robust Framework Combining Image Processing and Deep Learning Hybrid Model to Classify Cardiovascular Diseases Using a Limited Number of Paper-Based Complex ECG Images
(Scopus, 22-11-07) Fatema, Kaniz; Montaha, Sidratul; Rony, Md. Awlad Hossen; Azam, Sami; Hasan, Md. Zahid; Jonkman, Mirjam
Heart disease can be life-threatening if not detected and treated at an early stage. The electrocardiogram (ECG) plays a vital role in classifying cardiovascular diseases, and often physicians and medical researchers examine paper-based ECG images for cardiac diagnosis. An automated heart disease prediction system might help to classify heart diseases accurately at an early stage. This study aims to classify cardiac diseases into five classes with paper-based ECG images using a deep learning approach with the highest possible accuracy and the lowest possible time complexity. This research consists of two approaches. In the first approach, five deep learning models, InceptionV3, ResNet50, MobileNetV2, VGG19, and DenseNet201, are employed. In the second approach, an integrated deep learning model (InRes-106) is introduced, combining InceptionV3 and ResNet50. This model is developed as a deep convolutional neural network capable of extracting hidden and high-level features from images. An ablation study is conducted on the proposed model altering several components and hyper parameters, improving the performance even further. Before training the model, several image pre-processing techniques are employed to remove artifacts and enhance the image quality. Our proposed hybrid InRes-106 model performed best with a testing accuracy of 98.34%. The InceptionV3 model acquired a testing accuracy of 90.56%, the ResNet50 89.63%, the DenseNet201 88.94%, the VGG19 87.87%, and the MobileNetV2 achieved 80.56% testing accuracy. The model is trained with a k-fold cross-validation technique with different k values to evaluate the robustness further. Although the dataset contains a limited number of complex ECG images, our proposed approach, based on various image pre-processing techniques, model fine-tuning, and ablation studies, can effectively diagnose cardiac diseases.
An IoT Intensive AI-integrated System for Optimized Surface Water Quality Profiling
(Independent University, Bangladesh, 2023-05) Syeed, M M Mahbubul; Karim, Md. Rajaul; Hossain, Md Shakhawat; Fatema, Kaniz; Uddin, Mohammad Faisal; Khan, Razib Hayat
Surface water is heavily exposed to contamination as this is the ubiquitous source for the majority of water needs. This situation is exaggerated by excessive population, heavy industrialization, rapid urbanization, and ad-hoc monitoring. Comprehensive measurement and knowledge extraction of surface water pollution is therefore pivotal for ensuring safe and hygienic water use. However, current process of surface water quality profiling involves laboratory-based manual sample collection and testing, which is tardy, expensive, error-prone, and untraceable. This paper, therefore presents the design and development of an IoT integrated water quality profiling system that possesses a novel plug-and-play physical layer for the sensor actuation, and an AI powered fog computing based cloud application layer for remote water quality parameter measurement and data acquisition, remote data logging, monitoring and control, with data analytic for critical reasoning and decision making
Antigen-specific memory B-cell responses in Bangladeshi adults after one or two dose oral killed cholera vaccination, and comparison with responses following natural cholera
(2011) Alam, Mohammad Murshid; Riyadh, M. Asrafuzzaman; Fatema, Kaniz; Rahman, Mohammad Arif; Akhtar, Nayeema; Ahmed, Tanvir; Chowdhury, Mohiul Islam; Chowdhury, Fahima; Calderwood, Stephen B.; Harris, Jason B.; Ryan, Edward T.; Qadri, Firdausi
National Institutes of Health
Attitude towards Psychiatric Medication among the Providers and Recipients of Mental Health Services
(2020-12-02) Fatema, Kaniz
People often do not seek or adhere to psychiatric treatment for mental health problem. Among many factors including accessibility, stigma and side effects, attitudinal barriers toward psychiatric medication is one of the major reasons behind this. The attitudinal aspect of both the service providers and recipients can contribute to this. However, study findings on attitude towards psychiatric medication are limited in the worldwide context and almost nonexistent in the Bangladesh context. This study was therefore designed to explore the attitude towards psychiatric medication among the providers and recipients of mental health services. Qualitative exploration using grounded theory approach was used in this study. 18 individuals from different categories of providers and recipients participated in this study. The recipients were selected purposively from mental health service centers (National Institute of Mental Health, Bangabandhu Sheikh Mujib Medical University, Nasirullah Psychotherapy Unit) and from community to ensure maximum variation. Ethical approval was taken before starting data collection. Data collection involved in-depth interview done face to face using a predesigned topic guide to cover the research objectives. The topic guide was developed through mind-map exercise and was pilot tested before starting the interview. Written consent was obtained from the participants before the interview except for one illiterate participant. All the interviews were audio recorded with participants‟ permission. Recorded interviews were transcribed in the form of text document. Data was analyzed using computer based qualitative analysis program NVivo-10. Data collection, coding and analysis were carried out simultaneously. The data from the interview revealed several distinct themes which were classified into six broad categories namely, pre-conceived ideas, experience, attitude, behavior, facilitators of behavior and burden. Although the primary concern was the attitude towards medication, the other categories seem to contribute in widening the context in which attitude towards psychiatric medication develops and interact. Data revealed presence of several pre-conceived ideas among the participants, which include equating psychiatric medicine with psychiatric illness, reliance on medical model, low priority of mental health problem and misconceptions. They also reported personal and vicarious experiences of using psychiatric medication that have exposed them to direct and indirect impacts of psychiatric medication. The pre-conceived ideas and experiences seemed to contribute to the formation of attitude towards psychiatric medication. Attitude was conceptualized with a bi-factorial cognitive-affective model, which leads towards medication related behavior. The findings also indicated six distinct themes namely knowledge, psycho-education, communication, faith on expertise, multidisciplinary teamwork, and family influence that seemed to facilitate behavior related to psychiatric medication. Several types of burdens also revealed, which thought to be associated with behavior related to psychiatric medication. The findings have enhanced understanding of the connections between attitude and practice around the use of psychiatric medication. This may be useful in increasing early uptake of psychiatric intervention and in reducing non-compliance to medical treatment. Overall, findings of this study may contribute in policymaking and devising strategies to reduce mental health service gap in Bangladesh.
Classification of chronic kidney disease (ckd) using data mining techniques
(Daffodil International University, 2018-05-05) Arafat, Faisal; Islam, Shajedul; Fatema, Kaniz
In the past decade rapid growth of digital data and global accessibility of it through modern internet has seen a massive rise in machine learning research. In proportion to it, the medical data has also seen a massive serge of expansion. With the availability of structured clinical data, it has attracted scores of researchers to study on the automation of clinical disease detection with machine learning and data mining. Chronic Kidney disease (CKD) also known as renal disorder has been such a field of study for quite some time now. So, our research aims to study the automated detection of chronic kidney disease with clinical data using several machine learning classifier. This research particularly focuses on Random Forest classifier, Naïve Bayes and decision tree in the purpose of classifying the intended dataset. Observational and comparative studies will be conducted on the each of the classifier’s accuracy. The correlation and importance of each of the attributes to achieve the intended classification has been also explored in this study. Overall our endeavor has been to achieve a sustainable and feasible model to detect the chronic kidney disease with comprehensive clinical accuracy.
Comparison of automated system D2 Mini with conventional method and VITEK-2 for identification and antibiotic susceptibility pattern of gram-negative fermentative bacteria
(BRAC University, 2025-02) Fatema, Kaniz; Rahman, Tasnim; Jilani, Md. Shariful Alam; Hossain, Mahboob
This study evaluates and compares the performance of the automated D2 Mini system with conventional microbiological methods and the VITEK-2 system for the identification and antibiotic susceptibility testing (AST) of gram-negative fermentative bacteria. A total of 34 clinical isolates were analyzed, including Escherichia coli (n=9), Klebsiella pneumoniae (n=7), Klebsiella oxytoca (n=1), Salmonella Typhi (n=7), Salmonella Paratyphi (n=1), Serratia marcescens (n=3), Aeromonas hydrophila (n=2), Morganella morganii (n=1), Proteus hauseri (n=1), Proteus mirabilis (n=1), and Edwardsiella hoshinae (n=1). The isolates were recovered from various clinical specimens, including blood, urine, pus, wound swabs, catheter tips, and bronchial wash. Identification was performed using conventional methods, including Gram staining, biochemical tests, and antibiotic susceptibility testing (AST) by the disk diffusion method, followed by automated systems (VITEK-2 and D2 Mini). The results showed that both the VITEK-2 and conventional methods achieved 100% concordance for genus and species identification. However, the D2 Mini system demonstrated high genus-level concordance (100%) for most isolates, except for Klebsiella pneumoniae (85.7%) and Serratia marcescens (66.7%) at the species level. The D2 Mini system failed to identify Salmonella species at the species level. Antibiotic susceptibility testing revealed that both automated systems (VITEK-2 and D2 Mini) exhibited high concordance with the disk diffusion method for several antibiotics, including amoxicillin-clavulanate, ceftazidime, ciprofloxacin, gentamicin, and amikacin. However, discrepancies were observed for antibiotics such as netilmicin and colistin, with low concordance values. The D2 Mini demonstrated a restricted antibiotic panel, lacking profiles for antibiotics such as ceftriaxone colistin, and TZP, while VITEK-2 showed higher concordance but also displayed limitations for certain antibiotics. This study highlights the strengths and limitations of automated systems in microbial diagnostics, emphasizing the need for further improvements in their antibiotic testing capabilities, particularly for last-resort antibiotics and non-fermenting bacteria.
Design & Development of a Microcontroller Based on measuring Real Time Data and Interface with Mobile through Bluetooth
(East West University, 8/1/2015) Hosain, Shah Md. Manjir; Fatema, Kaniz
We know smart phone is a big blessing in our daily life. Through the smart phone we can do several things. That simplifies our work. We can also use it as technology purposes. In this project, we have designed and developed an Arduino based measuring system. Here we use real time clock, Arduino uno, sensor, Bluetooth, signal conditioning circuit, and Android mobile. The system can measure radiation, time and temperature. Arduino read data from radiation sensor and save data in EEPROM. When we need those data then we paired our Android mobile with Bluetooth. Consequently we can read data in our mobile phone. Here, an Arduino board has been used as the main controlling unit, HC-05 module as the Bluetooth sensor. A program has been developed and burnt into the microcontroller of the Arduino module. The system is constructed and tested and found to work satisfactorily.
Detection of Autism Spectrum Disorder of Children Using Machine Learning Techniques
(Daffodil International University, 2024-11-20) Fatema, Kaniz
Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition characterized by a wide range of symptoms that affect an individual's social interaction, communication, and behavior. Early diagnosis and intervention play a crucial role in improving the quality of life for individuals with ASD. This thesis presents a comprehensive investigation into the analysis and detection of ASD using advanced machine learning techniques. The primary objective of this research is to develop a robust and accurate ASD screening tool that can assist in early identification. To achieve this goal, we leverage machine learning algorithms and a diverse range of data sources, including behavioral assessments, clinical records, and demographic information. The research explores the application of supervised learning, feature selection, and data preprocessing techniques to enhance the performance of ASD detection models. This thesis also delves into the development of a prototype application that combines sophisticated machine learning models with an intuitive user interface. The application enables caregivers, educators, and healthcare professionals to conduct preliminary ASD assessments efficiently and receive real-time feedback. Furthermore, the study examines the significance of feature selection and engineering in improving the interpretability of ASD detection models. It explores the potential of neural networks, support vector machines, and other state-of-the-art algorithms in achieving high diagnostic accuracy. The findings presented in this thesis contribute to the growing body of research in the field of autism spectrum disorder detection. The research outcomes have practical implications for early intervention and support for individuals with ASD, ultimately promoting better outcomes and enhanced quality of life. In conclusion, this thesis serves as a valuable resource for researchers, healthcare professionals, and stakeholders in the field of ASD. It underscores the potential of machine learning techniques in advancing the accuracy and efficiency of ASD detection and sets the stage for further research and innovation in this critical area of healthcare.
Development of a Iron Removal Plant for Ground Water Based Supply System in Apartment Buildings
(Daffodil International University, 2021-01-23) Haque, Md. Anamul; Fatema, Kaniz; Khalil, Mohammad Ibrahim
Now a day’s groundwater is one of the major sources of water supply in Bangladesh, though it has an excessive iron content. Iron exists naturally in, underground water in Bangladesh. Presence of iron has negative impacts on aesthetic features of supplied water. Besides this iron precipitations cause damage to sanitary and water supply fittings and fixtures. To avoid such nuisance and to improve the quality of life for the residences of an apartment building, a treatment plant to fit the demand of a typical apartment building is designed in this research project. This project considered the different water flow pattern for buildings as well as the space constraints and made adjustments for those matters. A small physical model is created and tested in a small scale to understand and demonstrate its efficiency.
Development of an Automated Optimal Distance Feature-Based Decision System for Diagnosing Knee Osteoarthritis Using Segmented X-Ray Images
(Elsevier, 2023-11-03) Fatema, Kaniz; Rony, Md Awlad Hossen; Azam, Sami; Mukta, Md Saddam Hossain; Karim, Asif; Hasan, Md Zahid; Jonkman, Mirjam
Knee Osteoarthritis (KOA) is a leading cause of disability and physical inactivity. It is a degenerative joint disease that affects the cartilage, cushions the bones, and protects them from rubbing against each other during motion. If not treated early, it may lead to knee replacement. In this regard, early diagnosis of KOA is necessary for better treatment. Nevertheless, manual KOA detection is time-consuming and error-prone for large data hubs. In contrast, an automated detection system aids the specialist in diagnosing KOA grades accurately and quickly. So, the main objective of this study is to create an automated decision system that can analyze KOA and classify the severity grades, utilizing the extracted features from segmented X-ray images. In this study, two different datasets were collected from the Mendeley and Kaggle database and combined to generate a large data hub containing five classes: Grade 0 (Healthy), Grade 1 (Doubtful), Grade 2 (Minimal), Grade 3 (Moderate), and Grade 4 (Severe). Several image processing techniques were employed to segment the region of interest (ROI). These included Gradient-weighted Class Activation Mapping (Grad-Cam) to detect the ROI, cropping the ROI portion, applying histogram equalization (HE) to improve contrast, brightness, and image quality, and noise reduction (using Otsu thresholding, inverting the image, and morphological closing). Besides, the focus filtering method was utilized to eliminate unwanted images. Then, six feature sets (morphological, GLCM, statistical, texture, LBP, and proposed features) were generated from segmented ROIs. After evaluating the statistical significance of the features and selection methods, the optimal feature set (prominent six distance features) was selected, and five machine learning (ML) models were employed. Additionally, a decision-making strategy based on the six optimal features is proposed. The XGB model outperformed other models with a 99.46 % accuracy, using six distance features, and the proposed decision-making strategy was validated by testing 30 images.
English Classes Observation and Teaching Scheme
(Daffodil International University, 22-12-06) Fatema, Kaniz
This complete project paper is based on personal experience, my class taking, class observation, extra curriculum development, communication, behavior, self evaluation in one word teaching skills are expressed. While observing the class, I noticed that the teachers follow the Communicative language teaching (CLT) method, communication between instructors and learners and class formatting system. There is also Developmental Learning Program ( DLP) so that the syllabus can be completed easily. This system is arranged for every class. The system of each class is to take a class test on the topic of the previous class at the beginning. Then explain the new topic in class work (CW) in a systematic way. Each class had Advanced English Learner's Communicative English Grammar and Composition books. During the three month of internship at the school, I observed 4 classes and conducted many classes. Among all these I have written about the conduct of 5 classes. This internship helped me gain experience on how to conduct a class with professionalism and how to maintain all the students in the classroom. While I conducted the class I tried to build a passionate relationship between students and teachers so that I can understand their thoughts. This project paper included many more problems which were quite challenging for me but I was able to overcome. Moreover, I have already mentioned what I have learned from this journey which is very useful for my future professional care.
Functional overview of finance department: Avery Dennison Bangladesh
(BRAC University, 12/4/2017) Fatema, Kaniz; Habib, Dr. Md. Mamun
Avery Dennison Bangladesh is the leading organization in the industry of garments packaging and labeling and it is being serving a huge range of customers for many years.Avery Dennison Bangladesh is working with their consumers by adding value for them, creating customer’s brand image, building customer’s brands, generating new ideas, creating new products, managing their data and distributing their merchandise throughout supply chain. Getting opportunity to work in core finance function at ADBD Finance department, I get to learn about fixed asset verification, tax challan, certificates preparation, invoice payment preparation, and payment summary reconciliation processes practiced in Avery Dennison Bangladesh Finance Department. The title of the report is “Functional Overview of Finance Department: Avery Dennison Bangladesh”. The report is based on both primary and secondary data analysis. In the overall report I have tried to focus mainly on how Avery Dennison BangladeshFinance department functions distinctively. Basically the department has four wings. One is Finance also known as Finance core, Finance planning, Corporate & Regulatory affairs and Credit Control. This report is a detailed representation of all the mentioned topics and contains a preliminary discussion about Avery Dennison Finance Department. This company is the market leader in the packaging and labeling sector in Bangladesh. The normal day to day finance department functional activities are illustrated in order to make the processes look easy in the eyes of readers. Finally, I have tried to give some recommendations from my learning that I have got from my day to day job at Avery Dennison and from my university lectures as well.
Image processing based human detection and social distancing measurements with monitoring via fine tuned deep learning and computer vision
(BRAC University, 2022-05) Datta, Anurag; Fatema, Kaniz; Tasnim, Nowshin; Sitara, Faria; Das, Mrithik Kanti; Chakrabarty, Amitabha
The COVID-19 pandemic has significantly affected day to day lifestyle all over the planet by disequilibrating social order. It moreover added anxiety about the capability of the world’s democracies to cope with the vital and crucial emergencies. We should urgently restore a necessary methodology to fathom the emergency and rigorously depict a course forward. The Division of Public Health authorities have recommended everyone to uphold social distancing with a view to diminishing the number of physical encounters. To keep a record of social distancing from an overhead standpoint, we established a computer vision deep learning framework. Our schemed system utilized the object recognition paradigm to spot and identify people in video sequences or frames. In our research, we assess the classification performance of two distinct multilayer neural network models named YOLO using OpenCV and TensorFlow which are used in the implementation process of an automatic recognition system. Amongst these using SSD, CUDA, and CUDNN we achieved a success rate in the classification. Neural networks were trained on a dataset where we used COCO dataset methods. At a time when neural networks are increasingly being utilized for a spectrum of uses, it is essential to select the proper model for the classification process that can attain the ultimate accuracy with the least amount of training duration. The demonstration created by us allows the insertion of images and the creation of their datasets, this allows the user to train a model using their chosen parameters. The models can then be saved and used in other systems. Moreover, to prevent future crucial situations and by keeping in the head about COVID affected situations on various global aspects this work will become an integral part of contributing to the term “Social Distancing” by implementing this sustainably and with one of the best results outcomes in our proposed image processing based human detection and social distancing measurements with monitoring via fine tuned deep learning and computer vision. Because coronavirus sickness has had such a negative influence on the world economy, this research tries to reduce the further impacts while minimizing resource loss. Also, create a very accurate detection mechanism to aid in the tracking of social distancing. In these types of serious situations, adequate actions must be taken and help to assist further research and work as an example for future works on this segment.
Isolation of Acinetobacter baumannii from patient and hospital environment: analyzing their antibiotic resistance, serum resistance & biofilm formation
(BRAC University, 2024-07) Jinnah, Mst. Maskera; Fatema, Kaniz; Haque, Fahim Kabir Monjurul
The Acinetobacter baumannii is a gram-negative coccobacillus also known as opportunistic bacterial pathogens currently creating great concern in clinical aspects due to their capacity to endure for extended periods of time in the environment and ability to cause multi-drug resistant infections. The objective of the study was to identify A. baumannii from hospital environments and admitted patients in hospital to analyze their antibiotic resistance, serum resistance and biofilm formation. Total 450 samples were collected from different wards of Rajshahi Medical College and Hospital. Among them we obtained 53 isolates of A. baumannii confirmed by polymerase chain reaction by targeting of the blaOXA-51 gene. From 53 isolates, 20 were from patient’s samples (7 endotracheal aspirates, 4 blood, 5 wound swab, 2 throat swab and 2 catheter tube ) and 33 isolates were from environment’s samples (bed sheet, surface of furniture, nebulizer machine, floor, nurse’s hand swab, food cart, medicine cart and trolleys). Subsequently, an antibiotic susceptibility test was done. Isolates from patient specimens were resistant to gentamicin (90%), amikacin (90%), cefepime, piperacillin-tazobactam (85%), ceftazidime(85%), and tetracycline (80%). A significant proportion of the isolates, 70%, displayed resistant against levofloxacin, imipenem, meropenem. Most importantly, 90% of all patient isolates were MDR. On the other hand, hospital environment’s isolates were resistant against ceftazidime(100%), imipenem(87.9%,), piperacillin-tazobactam (78.8%), and cefepime(78.8%). A significant proportion of the isolates, 69.7% and 66.7% were resistant against meropenem and gentamicin. Among them 80% of all isolates were MDR. The result of the serum bactericidal assay showed that almost 31% of isolates were serum resistant and 35% were sensitive and 34% were intermediate. Isolates from the environment's samples were 10% more resistant than the isolates from the patient's samples. Among environment’s isolates 36% isolates were resistant whereas among patient’s isolates only 26% were resistant. According to quantitative biofilm formation results, among 33 environmental isolates 17% of isolates formed strong biofilm, 14% formed moderate film, and 30% formed weak and 39% isolates did not form biofilm.
Isolation of Staphylococcus spp. from hospital wastewater and adjacent community household water: Special focus on their antibiotic resistance
(BRAC University, 2023-01) Fatema, Kaniz; Ahmed, Akash
Staphylococcus spp. is a leading cause of human bacterial infections. These infections can damage the skin, soft tissues, bones, circulation, and respiratory system. It has the unusual capacity to rapidly develop resistance to any antibiotic deployed against it. Antibiotic-resistant S. aureus strains are rising at an alarming rate, which not only limits treatment options but also makes it impossible to calculate the economic deprivation caused by this superbug. In this research, the antimicrobial resistance patterns of Staphylococcus spp. isolated from hospital wastewater and community household water samples were investigated for 15 different antibiotics. The antibiotic resistance pattern of these Staphylococcus spp. isolates was determined using the disc diffusion method. Staphylococcus spp. was particularly resistant to antibiotics in the penicillin category, such as Penicillin-G, Oxacillin, and Methicillin. 60% isolates of hospital wastewater sample were resistant to both penicillin-G and methicillin & 100% isolates of community household water were resistant to penicillin-G and methicillin; 100% isolates of both hospital and community were resistant towards oxacillin. 40% isolates of hospital wastewater were resistant to Tetracycline. 100% isolates of hospital wastewater showed resistance towards the antibiotic Ceftazidime which belong to the group cephalosporins and 60% isolates of community water were resistant to Ceftazidime. 20% isolates of hospital wastewater and 25% isolates of community water were resistant to the antibiotic Erythromycin of the macrolides group. Since Staphylococcus spp. samples were resistant to more than one class of antibiotics, it can be concluded that they exhibited multidrug resistance.