Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179874
Nada Hesham Ahmed Elsherbeny, Abdelrahman Zaian, E. Supriyanto
The most prevalent disease is type 2 diabetes mellitus (T2DM), a chronic metabolic disorder. T2DM is linked to fat buildup in the lower torso around the abdomen, which leads to fat buildup in the belly region. As a result, it’s important to categorize and forecast diabetes patients based on their dietary intake. In this study, we used the pre-trained Inception V3, Keras, and Tensorflow convolutional neural network (CNN) model to identify different food categories. Comparing the CNN model’s accuracy to other methods from earlier studies, it achieved 96.6%, which is fairly high. Additionally, there is a correlation between calories with fat, carbs, protein, and sugar related with T2DM via linear regression between nutrition classes.
{"title":"Nutritional Analysis Using Convolutional Neural Network for Type II Diabetes","authors":"Nada Hesham Ahmed Elsherbeny, Abdelrahman Zaian, E. Supriyanto","doi":"10.1109/ICHE55634.2022.10179874","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179874","url":null,"abstract":"The most prevalent disease is type 2 diabetes mellitus (T2DM), a chronic metabolic disorder. T2DM is linked to fat buildup in the lower torso around the abdomen, which leads to fat buildup in the belly region. As a result, it’s important to categorize and forecast diabetes patients based on their dietary intake. In this study, we used the pre-trained Inception V3, Keras, and Tensorflow convolutional neural network (CNN) model to identify different food categories. Comparing the CNN model’s accuracy to other methods from earlier studies, it achieved 96.6%, which is fairly high. Additionally, there is a correlation between calories with fat, carbs, protein, and sugar related with T2DM via linear regression between nutrition classes.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124546755","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179886
M. Umborowati, I. Citrashanty, I. Surono, Maylita Sari, A. Endaryanto, C. Prakoeswa
Background: Asia is located around equator with all year sun exposure. That is why Asians are more susceptible to aging caused by ultraviolet, or called as photoaging. Dyspigmentation and wrinkle is the most visible clinical sign of skin photoaging, making them the main target to treat.Objective: To investigate the efficacy of fractional Carbon Dioxide (CO2) laser-assisted amniotic membrane stem cell conditioned medium (AMSC-CM) plus vitamin C in Asian skin photoaging.Patients and methods: This were a randomized comparative clinical study in photoaged patients. Experimental group received amniotic membrane stem cell conditioned medium (AMSC-CM) plus vitamin C mixture after fractional CO2 laser, while another group received AMSC-CM after fractional CO2 laser. The formulation was applied three times with 4 weeks interval. Fractional CO2 laser was used to assist epidermal penetration. Wrinkle, pore, spot, and skin tone were evaluated before treatment and 4 weeks after last session.Results: This study included 60 women with an average age of over 50 years. Wrinkle and pore improvement after therapy on AMSC-CM plus Vitamin C mixture group were significantly better than group who only received AMSC-CM (p value < 0.05). No serious adverse effect was observed during the study.Conclusion: AMSC-CM plus Vitamin C mixture application after laser fractional CO2 is promising as rejuvenation treatment in Asian skin.
{"title":"Beneficial Effect of Amniotic Membrane Stem Cell and Vitamin C after Fractional Carbon-Dioxide Laser for Photoaging Treatment in Asian Skin","authors":"M. Umborowati, I. Citrashanty, I. Surono, Maylita Sari, A. Endaryanto, C. Prakoeswa","doi":"10.1109/ICHE55634.2022.10179886","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179886","url":null,"abstract":"Background: Asia is located around equator with all year sun exposure. That is why Asians are more susceptible to aging caused by ultraviolet, or called as photoaging. Dyspigmentation and wrinkle is the most visible clinical sign of skin photoaging, making them the main target to treat.Objective: To investigate the efficacy of fractional Carbon Dioxide (CO2) laser-assisted amniotic membrane stem cell conditioned medium (AMSC-CM) plus vitamin C in Asian skin photoaging.Patients and methods: This were a randomized comparative clinical study in photoaged patients. Experimental group received amniotic membrane stem cell conditioned medium (AMSC-CM) plus vitamin C mixture after fractional CO2 laser, while another group received AMSC-CM after fractional CO2 laser. The formulation was applied three times with 4 weeks interval. Fractional CO2 laser was used to assist epidermal penetration. Wrinkle, pore, spot, and skin tone were evaluated before treatment and 4 weeks after last session.Results: This study included 60 women with an average age of over 50 years. Wrinkle and pore improvement after therapy on AMSC-CM plus Vitamin C mixture group were significantly better than group who only received AMSC-CM (p value < 0.05). No serious adverse effect was observed during the study.Conclusion: AMSC-CM plus Vitamin C mixture application after laser fractional CO2 is promising as rejuvenation treatment in Asian skin.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120961352","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179887
Raghdah Saem Aldahr, Munid Alanazi, Mohammad Ilyas
Electroencephalogram (EEG) signals are a key source in epileptic seizure recognition. The patient-specific data scantiness in EEG signals, referring to the inter-patient variability of EEG, hinders the accurate recognition of epileptic seizure patterns. This work presents a multi-class epileptic seizure detection scheme with a two-step solution for addressing the data scantiness and inter-patient variability constraint. The proposed Diversity-enhanced data Augmentation and graph theory-assisted FEw-shot Learning for Multi-class seizure detection (DAFEM) approach incorporates diversified data augmentation, graph theory-based feature extraction, and few-shot learning-based multi-class classification. Initially, the data augmentation utilized a Generative Adversarial Network (GAN) with diversified EEG sample generation to conquer EEG data scarcity. Subsequently, it extracts the potential set of features from the augmented data using the graph theory method based on the analysis of inherent dynamic characteristics of EEG data. In particular, to recognize the marginal and the drastic temporal fluctuations in EEG data patterns, it performs the Temporal Weight Fluctuation (TWF) in addition to the feature extraction scores. The data scarcity in the epileptic seizure classes is handled by adopting the few-shot learning strategy. By modeling the Siamese neural network for the multi-class classification of epilepsy, it discriminates the normal, preictal, and ictal patient samples over the constraint of inter-patient variability of EEG data. Finally, the proposed work is tested with two authoritative EEG datasets. The experimental outcomes illustrate that the proposed DAFEM yields 2.73% and 4.5% higher recall on Bonn and CHB-MIT datasets, respectively.
{"title":"Addressing Inter-Patient Variability in EEG: Diversity-Enhanced Data Augmentation and Few-Shot Learning-based Epilepsy Detection","authors":"Raghdah Saem Aldahr, Munid Alanazi, Mohammad Ilyas","doi":"10.1109/ICHE55634.2022.10179887","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179887","url":null,"abstract":"Electroencephalogram (EEG) signals are a key source in epileptic seizure recognition. The patient-specific data scantiness in EEG signals, referring to the inter-patient variability of EEG, hinders the accurate recognition of epileptic seizure patterns. This work presents a multi-class epileptic seizure detection scheme with a two-step solution for addressing the data scantiness and inter-patient variability constraint. The proposed Diversity-enhanced data Augmentation and graph theory-assisted FEw-shot Learning for Multi-class seizure detection (DAFEM) approach incorporates diversified data augmentation, graph theory-based feature extraction, and few-shot learning-based multi-class classification. Initially, the data augmentation utilized a Generative Adversarial Network (GAN) with diversified EEG sample generation to conquer EEG data scarcity. Subsequently, it extracts the potential set of features from the augmented data using the graph theory method based on the analysis of inherent dynamic characteristics of EEG data. In particular, to recognize the marginal and the drastic temporal fluctuations in EEG data patterns, it performs the Temporal Weight Fluctuation (TWF) in addition to the feature extraction scores. The data scarcity in the epileptic seizure classes is handled by adopting the few-shot learning strategy. By modeling the Siamese neural network for the multi-class classification of epilepsy, it discriminates the normal, preictal, and ictal patient samples over the constraint of inter-patient variability of EEG data. Finally, the proposed work is tested with two authoritative EEG datasets. The experimental outcomes illustrate that the proposed DAFEM yields 2.73% and 4.5% higher recall on Bonn and CHB-MIT datasets, respectively.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"41 7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131054833","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179868
Mohamud Abdullahi Hassan, Abdelrahman Zaian, Nur Syafiqah A. Hassan, E. Supriyanto
Majority of strokes are brought on by an unanticipated obstruction of blood flow to the brain and heart. Stroke severity can be reduced by being aware of the various stroke warning signs in advance. Previous study on stroke prediction had an accuracy less than 90%. Sample size of 1000 – 2000 for that study was insufficient to justify the results obtained by the trained model. In this study, comparisons are made among different approaches to the stroke prediction model, include four different classification methods, which are logistic regression, Random Forest, Decision Tree and Support Vector Machine (SVM). The results obtained by the classifiers were trained with 2000 samples and 3109. All the classifiers were then tested individually. The accuracy for each model are, 91% for Decision Tree, 95% for Random Forest, 95% for Logistic Regression and 100% Support Vector Machine (SVM). As a conclusion, our study suggested that SVM approach is fit well for stroke prediction model as it achieved the highest accuracy compared to the others.
{"title":"Stroke Prediction Model Using Machine Learning Method","authors":"Mohamud Abdullahi Hassan, Abdelrahman Zaian, Nur Syafiqah A. Hassan, E. Supriyanto","doi":"10.1109/ICHE55634.2022.10179868","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179868","url":null,"abstract":"Majority of strokes are brought on by an unanticipated obstruction of blood flow to the brain and heart. Stroke severity can be reduced by being aware of the various stroke warning signs in advance. Previous study on stroke prediction had an accuracy less than 90%. Sample size of 1000 – 2000 for that study was insufficient to justify the results obtained by the trained model. In this study, comparisons are made among different approaches to the stroke prediction model, include four different classification methods, which are logistic regression, Random Forest, Decision Tree and Support Vector Machine (SVM). The results obtained by the classifiers were trained with 2000 samples and 3109. All the classifiers were then tested individually. The accuracy for each model are, 91% for Decision Tree, 95% for Random Forest, 95% for Logistic Regression and 100% Support Vector Machine (SVM). As a conclusion, our study suggested that SVM approach is fit well for stroke prediction model as it achieved the highest accuracy compared to the others.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134574906","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179864
Hana Ali Ibrahim, Mathiventtan N. Thamilvanan, Abdelrahman Zaian, E. Supriyanto
During an in vitro fertilization (IVF), an egg cell and sperm are combined outside of the body. The selection of embryos during IVF is very important. The quality of the embryo needs to be evaluated before it may be transferred. At this moment, the quality of embryos is evaluated visually. The morphological judgment is dependent on the expertise and experience of the attending physician or embryologist. The evaluation of embryo images can be done with the use of artificial intelligence (AI), which can be utilized to achieve unbiased automatic embryo segmentation. Both supervised and unsupervised methods can be used to complete the segmentation process. CNN is utilized in this study to perform the segmentation of embryo pictures. The model that performs the best in this research makes use of typical training data and divides it up into two classes. It has an accuracy of 93.8 percent, and by using it, the research can assess whether an embryo is usable.
{"title":"Fertility Assessment Model For Embryo Grading Using Convolutional Neural Network (CNN)","authors":"Hana Ali Ibrahim, Mathiventtan N. Thamilvanan, Abdelrahman Zaian, E. Supriyanto","doi":"10.1109/ICHE55634.2022.10179864","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179864","url":null,"abstract":"During an in vitro fertilization (IVF), an egg cell and sperm are combined outside of the body. The selection of embryos during IVF is very important. The quality of the embryo needs to be evaluated before it may be transferred. At this moment, the quality of embryos is evaluated visually. The morphological judgment is dependent on the expertise and experience of the attending physician or embryologist. The evaluation of embryo images can be done with the use of artificial intelligence (AI), which can be utilized to achieve unbiased automatic embryo segmentation. Both supervised and unsupervised methods can be used to complete the segmentation process. CNN is utilized in this study to perform the segmentation of embryo pictures. The model that performs the best in this research makes use of typical training data and divides it up into two classes. It has an accuracy of 93.8 percent, and by using it, the research can assess whether an embryo is usable.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"2015 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114444933","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179865
K. Lai, Pauline Shan Qing Yeoh, S. Goh, K. Hasikin, Xiang Wu
Automated knee segmentation plays an important role in knee osteoarthritis diagnosis as this disease exhibits different imaging biomarkers as it progresses. A good knee segmentation model that is practical and computationally efficient allows a more efficient clinical workflow. This paper presents a preliminary study on Depthwise Separable convolutional layers utilizing the end-to-end segmentation network, UNet architecture on knee segmentation. Results showed that DS2D-UNet and DS3D-UNet perform more efficiently with the adoption of Depthwise Separable convolutional layers with fewer cost of computations, without compromising the overall performance. The models produced strong results of Balanced Accuracy ranging between 90–93% and Dice Similarity Coefficient ranging between 91–93%. In conclusion, the potential of Depthwise Separable convolution should be further investigated to optimize the efficiency of 3D deep learning architectures, specifically on knee imaging volumes.
{"title":"Depthwise Separable Convolutional Neural Network for Knee Segmentation: Data from the Osteoarthritis Initiative","authors":"K. Lai, Pauline Shan Qing Yeoh, S. Goh, K. Hasikin, Xiang Wu","doi":"10.1109/ICHE55634.2022.10179865","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179865","url":null,"abstract":"Automated knee segmentation plays an important role in knee osteoarthritis diagnosis as this disease exhibits different imaging biomarkers as it progresses. A good knee segmentation model that is practical and computationally efficient allows a more efficient clinical workflow. This paper presents a preliminary study on Depthwise Separable convolutional layers utilizing the end-to-end segmentation network, UNet architecture on knee segmentation. Results showed that DS2D-UNet and DS3D-UNet perform more efficiently with the adoption of Depthwise Separable convolutional layers with fewer cost of computations, without compromising the overall performance. The models produced strong results of Balanced Accuracy ranging between 90–93% and Dice Similarity Coefficient ranging between 91–93%. In conclusion, the potential of Depthwise Separable convolution should be further investigated to optimize the efficiency of 3D deep learning architectures, specifically on knee imaging volumes.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125528875","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179871
Khairul Eahsun Fahim, Hayati Yassin, Md Hasnatul Amin, Priyanka Das Dewan, Aminul Islam
Among the significant causes of death worldwide, cardiovascular diseases (CVD) account for 80 percent of deaths in low- and middle-income countries such as Bangladesh, according to the World Health Organization (WHO). In Bangladesh, the prevalence of HIV/AIDS and the mortality linked with it have climbed considerably over the past few decades. The rising incidence of cardiovascular disease in Bangladesh needs a complete understanding of the epidemiology of CVD risk among the population. Clinical data analysis is a significant concern for someone dealing with cardiovascular illness. When it comes to generating decisions and making predictions from the vast volumes of data generated by the healthcare industry, machine learning (ML) is to be extremely useful. It is proposed in this research to apply a supervised machine learning algorithm to detect cardiovascular disease (CVD) in individuals early on, allowing them to become concerned about their medical status and avert significant illnesses. When it comes to detecting the disease, four different machine learning methods have been used. The dataset of patients was used, and various machine learning methods, including K-nearest neighbors, Random Forest, Decision trees, and XGBoost, were used to make predictions. As a consequence of the tests, the XGBoost method is superior to the other three tactics (73.72 percent). Moreover, for the modified dataset where smoking, alcohol intake, and physical activity are positive, the percentage is 81.14% to show the effect of smoking and alcohol consumption in a physically active person in terms of cardiovascular disease. Furthermore, these strategies have been evaluated regarding their ability to detect early-stage CVD inpatients. This paper examined the Kaggle dataset to observe the trait and suitability to implement the system for primary data collected from Bangladeshi patients.
{"title":"Detection of Cardiovascular Disease of Patients at an Early Stage Using Machine Learning Algorithms","authors":"Khairul Eahsun Fahim, Hayati Yassin, Md Hasnatul Amin, Priyanka Das Dewan, Aminul Islam","doi":"10.1109/ICHE55634.2022.10179871","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179871","url":null,"abstract":"Among the significant causes of death worldwide, cardiovascular diseases (CVD) account for 80 percent of deaths in low- and middle-income countries such as Bangladesh, according to the World Health Organization (WHO). In Bangladesh, the prevalence of HIV/AIDS and the mortality linked with it have climbed considerably over the past few decades. The rising incidence of cardiovascular disease in Bangladesh needs a complete understanding of the epidemiology of CVD risk among the population. Clinical data analysis is a significant concern for someone dealing with cardiovascular illness. When it comes to generating decisions and making predictions from the vast volumes of data generated by the healthcare industry, machine learning (ML) is to be extremely useful. It is proposed in this research to apply a supervised machine learning algorithm to detect cardiovascular disease (CVD) in individuals early on, allowing them to become concerned about their medical status and avert significant illnesses. When it comes to detecting the disease, four different machine learning methods have been used. The dataset of patients was used, and various machine learning methods, including K-nearest neighbors, Random Forest, Decision trees, and XGBoost, were used to make predictions. As a consequence of the tests, the XGBoost method is superior to the other three tactics (73.72 percent). Moreover, for the modified dataset where smoking, alcohol intake, and physical activity are positive, the percentage is 81.14% to show the effect of smoking and alcohol consumption in a physically active person in terms of cardiovascular disease. Furthermore, these strategies have been evaluated regarding their ability to detect early-stage CVD inpatients. This paper examined the Kaggle dataset to observe the trait and suitability to implement the system for primary data collected from Bangladeshi patients.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131761844","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179869
Mohd Amiruddin Abd Rahman, C. E. A. Bundak, Muhammad Khairul Anwar bin Mohd Yusof
Vaccination is essential to minimize the transmission of the Covid-19 virus and its possible influence on morbidity and mortality rates around the world. In this paper, we first performed exploratory data analysis (EDA) on Covid-19 deaths in Malaysia depending on vaccine dose and next we used this vaccine dataset to predict the death cases using a machine learning algorithm. In EDA, we evaluated the vaccination dose impact according to each type of vaccines on the deaths count in Malaysia. The analysed data is compared to the number of dosages, comorbidity status and age variation. Aside from that, we observed the number of deceased people who were tested positive for Covid-19 after vaccination and the death count days after getting vaccinated. Our finding shows that the highest deaths number is mostly occurred to the person who received first dose vaccine, have more than one disease and lastly having the age range of 50 to 60 years old. In the second part of the paper, we used the death cases, daily cases, and daily vaccination to predict the death cases in which both the daily cases and the daily vaccination is used as the input factor. PSO-SVR with three kernel function (linear, polynomial, and radial basis function) is used to predict 30 days of death cases. From the prediction, the input factor of daily vaccination (RMSE=107.98) gives twice better accuracy compared to using the daily cases (RMSE=48.71). However, when using both input factor, the error reduces to (RMSE=16.77). The best kernel function for prediction is RBF in which for both input factors, RBF gives results of (RMSE=16.77) compared to linear (RMSE=17.43) and polynomial (RMSE=17.24).
{"title":"Analytics of the COVID-19 Death According to the Vaccine Dose: Malaysia Case Study","authors":"Mohd Amiruddin Abd Rahman, C. E. A. Bundak, Muhammad Khairul Anwar bin Mohd Yusof","doi":"10.1109/ICHE55634.2022.10179869","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179869","url":null,"abstract":"Vaccination is essential to minimize the transmission of the Covid-19 virus and its possible influence on morbidity and mortality rates around the world. In this paper, we first performed exploratory data analysis (EDA) on Covid-19 deaths in Malaysia depending on vaccine dose and next we used this vaccine dataset to predict the death cases using a machine learning algorithm. In EDA, we evaluated the vaccination dose impact according to each type of vaccines on the deaths count in Malaysia. The analysed data is compared to the number of dosages, comorbidity status and age variation. Aside from that, we observed the number of deceased people who were tested positive for Covid-19 after vaccination and the death count days after getting vaccinated. Our finding shows that the highest deaths number is mostly occurred to the person who received first dose vaccine, have more than one disease and lastly having the age range of 50 to 60 years old. In the second part of the paper, we used the death cases, daily cases, and daily vaccination to predict the death cases in which both the daily cases and the daily vaccination is used as the input factor. PSO-SVR with three kernel function (linear, polynomial, and radial basis function) is used to predict 30 days of death cases. From the prediction, the input factor of daily vaccination (RMSE=107.98) gives twice better accuracy compared to using the daily cases (RMSE=48.71). However, when using both input factor, the error reduces to (RMSE=16.77). The best kernel function for prediction is RBF in which for both input factors, RBF gives results of (RMSE=16.77) compared to linear (RMSE=17.43) and polynomial (RMSE=17.24).","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"144 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122684671","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179889
I. Gunawan, Gita Rindang Lestari, R. N. Hidayati, E. Supriyanto, Vita Nurdinawati
We designed cranial electro stimulation with a low current intensity, which is applied to the head indirectly using direct current intensity. Unidirectional transcranial stimulation is a non-invasive brain stimulation method that has been shown to be effective in modulating cortical excitability and guiding human perception and behavior. The purpose of this research is to design a low-current-intensity cranial electro-stimulation therapy device that is affordable, dependable, and feasible. The CES has a frequency range of 10, 13, and 15 Hz and treatment times of 15, 30, and 45 minutes. The CES generates a current intensity of 0.25, 0.5, 0.75, and 1 mA. The design of CES prototypes was tested at Balai Pengamanan Fasilitas Kesehatan (BPFK) Jakarta, Indonesia, which includes electrical safety measurement, performance testing, and battery reliability. The BPFK Jakarta declares that the test meets the requirements of the testing method and that the tool has passed the test. In addition, the tool has been issued a certificate with no YK.01.03/XLVII.2/PK/2022.
{"title":"Conformity Evaluation of Cranial Electro-stimulation Prototype with Low Current Intensity and Smartphone Application","authors":"I. Gunawan, Gita Rindang Lestari, R. N. Hidayati, E. Supriyanto, Vita Nurdinawati","doi":"10.1109/ICHE55634.2022.10179889","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179889","url":null,"abstract":"We designed cranial electro stimulation with a low current intensity, which is applied to the head indirectly using direct current intensity. Unidirectional transcranial stimulation is a non-invasive brain stimulation method that has been shown to be effective in modulating cortical excitability and guiding human perception and behavior. The purpose of this research is to design a low-current-intensity cranial electro-stimulation therapy device that is affordable, dependable, and feasible. The CES has a frequency range of 10, 13, and 15 Hz and treatment times of 15, 30, and 45 minutes. The CES generates a current intensity of 0.25, 0.5, 0.75, and 1 mA. The design of CES prototypes was tested at Balai Pengamanan Fasilitas Kesehatan (BPFK) Jakarta, Indonesia, which includes electrical safety measurement, performance testing, and battery reliability. The BPFK Jakarta declares that the test meets the requirements of the testing method and that the tool has passed the test. In addition, the tool has been issued a certificate with no YK.01.03/XLVII.2/PK/2022.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129843869","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}
Pub Date : 2022-09-23DOI: 10.1109/ICHE55634.2022.10179875
Puja Saha, S. Chowdhury, Afsana Mehrab, Jahangir Alam
In recent years, the widespread dominance of convolutional neural networks (CNN) in numerous computer vision applications, particularly in medical imaging, has been compelling. However, their applications as classifiers are tedious since they need high volume (usually several hundred to several thousand) and thorough preparation of training samples to learn competently. Sometimes, it is nearly impossible to collect such a large number of unique images, especially for rare diseases (i.e., Multiple Sclerosis). Hence, we proposed a CNN that required only sixty unique and nearly unprocessed samples to learn to classify disparate samples of the same disorder with an accuracy of 85%, making it highly likely to overcome the aforementioned constraint. Although due to the paucity of patients with rare brain disorders, in this research we deployed the model to perform classifications of tumorous and hemorrhagic scans against normal ones, it could be generalized to images of other conditions, even rarer ones, since it does not require much to learn.
{"title":"Convolutional Neural Network to Classify Medical Images of Rare Brain Disorders","authors":"Puja Saha, S. Chowdhury, Afsana Mehrab, Jahangir Alam","doi":"10.1109/ICHE55634.2022.10179875","DOIUrl":"https://doi.org/10.1109/ICHE55634.2022.10179875","url":null,"abstract":"In recent years, the widespread dominance of convolutional neural networks (CNN) in numerous computer vision applications, particularly in medical imaging, has been compelling. However, their applications as classifiers are tedious since they need high volume (usually several hundred to several thousand) and thorough preparation of training samples to learn competently. Sometimes, it is nearly impossible to collect such a large number of unique images, especially for rare diseases (i.e., Multiple Sclerosis). Hence, we proposed a CNN that required only sixty unique and nearly unprocessed samples to learn to classify disparate samples of the same disorder with an accuracy of 85%, making it highly likely to overcome the aforementioned constraint. Although due to the paucity of patients with rare brain disorders, in this research we deployed the model to perform classifications of tumorous and hemorrhagic scans against normal ones, it could be generalized to images of other conditions, even rarer ones, since it does not require much to learn.","PeriodicalId":289905,"journal":{"name":"2022 International Conference on Healthcare Engineering (ICHE)","volume":"150 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132516949","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：