Pub Date : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856586
Rakesh Ranjan, Neeti, B. Sahana
Electroencephalography (EEG) is an indispensable non-invasive analytical method in the diagnosis and characterization of mental health. However, the conventional EEG interpretation process is quite subjective, time-consuming, and susceptible to error. The clinicians usually observe abnormalities in amplitude or frequency to markup the EEG signal as unhealthy, which is based on visual scrutiny of EEG data. In case of high-volume long-duration EEG recordings, it will be a grueling task for experts and may cause inaccurate classification of EEGs. In this work, a computer-aided automatic decision-making model has been designed to identify mental health status using only alpha band (8–12 Hz) of EEG signal to conquer the aforementioned difficulties. The demonstration of this study is carried out on the two publicly available EEG datasets of epileptical seizure and schizophrenia. The proposed simulation model followed the process flow of signal denoising, decomposition of EEG signal into various bands, feature extractions from alpha band of EEG data, and classification of mental health of human as healthy or unhealthy. The performance of chosen features is evaluated through popular classifiers. The ensemble bagged tree classifier outperforms the other methods on epileptical seizure and schizophrenia datasets with a classification accuracy of 99.5% and 98.68% respectively. Hence, this proposed method can be an alternative for the automatic classification of mental health status at the early stage of EEG analysis.
{"title":"Automatic Detection of Mental Health Status using Alpha Subband of EEG Data","authors":"Rakesh Ranjan, Neeti, B. Sahana","doi":"10.1109/MeMeA54994.2022.9856586","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856586","url":null,"abstract":"Electroencephalography (EEG) is an indispensable non-invasive analytical method in the diagnosis and characterization of mental health. However, the conventional EEG interpretation process is quite subjective, time-consuming, and susceptible to error. The clinicians usually observe abnormalities in amplitude or frequency to markup the EEG signal as unhealthy, which is based on visual scrutiny of EEG data. In case of high-volume long-duration EEG recordings, it will be a grueling task for experts and may cause inaccurate classification of EEGs. In this work, a computer-aided automatic decision-making model has been designed to identify mental health status using only alpha band (8–12 Hz) of EEG signal to conquer the aforementioned difficulties. The demonstration of this study is carried out on the two publicly available EEG datasets of epileptical seizure and schizophrenia. The proposed simulation model followed the process flow of signal denoising, decomposition of EEG signal into various bands, feature extractions from alpha band of EEG data, and classification of mental health of human as healthy or unhealthy. The performance of chosen features is evaluated through popular classifiers. The ensemble bagged tree classifier outperforms the other methods on epileptical seizure and schizophrenia datasets with a classification accuracy of 99.5% and 98.68% respectively. Hence, this proposed method can be an alternative for the automatic classification of mental health status at the early stage of EEG analysis.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117030688","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856482
A. Pennisi, D. Bloisi, D. Nardi, S. Varricchio, F. M. Donini
Oral tumors are responsible for about 170,000 deaths every year in the World. In this paper, we focus on oral squamous cell carcinoma (OSCC), which represents up to 80–90 % of all malignant neoplasms of the oral cavity. We present a novel deep learning-based method for segmenting whole slide image (WSI) samples at the pixel level. The proposed method is a modification of the well-known U-Net architecture through a multi-encoder structure. In particular, our network, called Multi-encoder U-Net, is a multi-encoder single decoder network that takes as input an image and splits it in tiles. For each tile, there is an encoder responsible for encoding it in the latent space, then a convolutional layer is responsible for merging the tiles into a single layer. Each layer of the decoder takes as input the previous up-sampled layer and concatenate it with the layer made by merging the corresponding layers of the multiple encoders. Experiments have been carried out on the publicly available ORal Cancer Annotated (ORCA) dataset, which contains annotated data from the TCGA repository. Quantitative experimental results, obtained using three different quality metrics, demonstrate the effectiveness of the proposed approach, which achieves 82% Pixel-wise Accuracy, 0.82 Dice similarity score, and 0.72 Mean Intersection Over Union.
{"title":"Multi-encoder U-Net for Oral Squamous Cell Carcinoma Image Segmentation","authors":"A. Pennisi, D. Bloisi, D. Nardi, S. Varricchio, F. M. Donini","doi":"10.1109/MeMeA54994.2022.9856482","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856482","url":null,"abstract":"Oral tumors are responsible for about 170,000 deaths every year in the World. In this paper, we focus on oral squamous cell carcinoma (OSCC), which represents up to 80–90 % of all malignant neoplasms of the oral cavity. We present a novel deep learning-based method for segmenting whole slide image (WSI) samples at the pixel level. The proposed method is a modification of the well-known U-Net architecture through a multi-encoder structure. In particular, our network, called Multi-encoder U-Net, is a multi-encoder single decoder network that takes as input an image and splits it in tiles. For each tile, there is an encoder responsible for encoding it in the latent space, then a convolutional layer is responsible for merging the tiles into a single layer. Each layer of the decoder takes as input the previous up-sampled layer and concatenate it with the layer made by merging the corresponding layers of the multiple encoders. Experiments have been carried out on the publicly available ORal Cancer Annotated (ORCA) dataset, which contains annotated data from the TCGA repository. Quantitative experimental results, obtained using three different quality metrics, demonstrate the effectiveness of the proposed approach, which achieves 82% Pixel-wise Accuracy, 0.82 Dice similarity score, and 0.72 Mean Intersection Over Union.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121898051","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856516
Silvia Chiera, A. Cristoforetti, L. Benedetti, Luca Borro, L. Mazzei, G. Nollo, F. Tessarolo
Face masks are used worldwide to reduce COVID-19 transmission in indoor environments. Differently from face respirators, there are no standards methods for measuring the fraction of air leaking at the face seal of loose-fitting masks such as medical and community masks. This study applies a recently developed method to quantify air leakage at the face seal to 14 medical and community mask models with the aim to understand the role of mask design and filter properties in air leakage. An instrumented head-form equipped with sensors for measuring volumetric airflow and differential pressure was used to simulate the air exhalation from the mouth of a person wearing a face mask. Results showed that the fraction of leaking air at the face seal is not negligible and can range from 10% to 95% according to mask model. The higher the exhaled airflow rate and the lower the amount of leaking fraction. A strong correlation was found between leaking fraction and filter breathability, indicating that a better breathability can lower air leakage. Highly breathable filtering materials should be employed in the production of medical and community face masks to maximize user comfort and minimize the fraction of exhaled air leaking unfiltered at the face seal.
{"title":"The role of filter breathability in reducing the fraction of exhaled air leaking from surgical and community face masks","authors":"Silvia Chiera, A. Cristoforetti, L. Benedetti, Luca Borro, L. Mazzei, G. Nollo, F. Tessarolo","doi":"10.1109/MeMeA54994.2022.9856516","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856516","url":null,"abstract":"Face masks are used worldwide to reduce COVID-19 transmission in indoor environments. Differently from face respirators, there are no standards methods for measuring the fraction of air leaking at the face seal of loose-fitting masks such as medical and community masks. This study applies a recently developed method to quantify air leakage at the face seal to 14 medical and community mask models with the aim to understand the role of mask design and filter properties in air leakage. An instrumented head-form equipped with sensors for measuring volumetric airflow and differential pressure was used to simulate the air exhalation from the mouth of a person wearing a face mask. Results showed that the fraction of leaking air at the face seal is not negligible and can range from 10% to 95% according to mask model. The higher the exhaled airflow rate and the lower the amount of leaking fraction. A strong correlation was found between leaking fraction and filter breathability, indicating that a better breathability can lower air leakage. Highly breathable filtering materials should be employed in the production of medical and community face masks to maximize user comfort and minimize the fraction of exhaled air leaking unfiltered at the face seal.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122147542","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856415
Sarah Adamo, C. Ricciardi, P. Ambrosino, M. Maniscalco, A. Biancardi, G. Cesarelli, L. Donisi, G. D'Addio
After the acute disease, post-COVID-19 patients may present several and persistent symptoms, known as the new paradigm of “post-acute COVID-19 syndrome”. This necessitates a multidisciplinary rehabilitation that has been proposed but whose effectiveness is still to be assessed. In this study, convalescent COVID-19 patients undergoing pulmonary rehabilitation (PR) after reporting long-term symptoms were consecutively enrolled. Then, they were grouped by laboratory parameters at admission through an unsupervised Machine Learning (ML) approach. We aimed to identify potential indicators that could discriminate several phenotypes leading to a different responsiveness to the rehabilitation program. A k-means clustering method was performed; then, statistical analysis was employed to compare clinical and hematochemical parameters of the obtained clusters. The dataset consisted of 78 patients (84.8% males, mean age 60.72 years). The optimal number for clustering was $boldsymbol{mathrm{k}=2}$ with a silhouette coefficient of 0.85, and D-Dimer resulted the most discriminating parameter, thus confirming its role as a marker of inflammation. The phenotypes exhibited statistically significant differences in terms of age $boldsymbol{(mathrm{p}=0.007)}$, packs of cigarettes per year $boldsymbol{(mathrm{p}=0.003)}$, uricemia $boldsymbol{(mathrm{p}=0.010)}$, PCR $boldsymbol{(mathrm{p}=0.026)}$, D-Dimer $boldsymbol{(mathrm{p} < 0.001)}$, red blood cells $boldsymbol{(mathrm{p}=0.005)}$, hemoglobin $boldsymbol{(mathrm{p}=0.039)}$, hematocrit $boldsymbol{(mathrm{p}=0.026), text{PaO}_{2} (mathrm{p}=0.006)},boldsymbol{text{SpO}_{2} (mathrm{p}=0.011)}$. Overall, our findings suggest the effectiveness of ML in identifying personalized prevention, interventional and rehabilitation strategies.
{"title":"Unsupervised Machine Learning to Identify Convalescent COVID-19 Phenotypes","authors":"Sarah Adamo, C. Ricciardi, P. Ambrosino, M. Maniscalco, A. Biancardi, G. Cesarelli, L. Donisi, G. D'Addio","doi":"10.1109/MeMeA54994.2022.9856415","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856415","url":null,"abstract":"After the acute disease, post-COVID-19 patients may present several and persistent symptoms, known as the new paradigm of “post-acute COVID-19 syndrome”. This necessitates a multidisciplinary rehabilitation that has been proposed but whose effectiveness is still to be assessed. In this study, convalescent COVID-19 patients undergoing pulmonary rehabilitation (PR) after reporting long-term symptoms were consecutively enrolled. Then, they were grouped by laboratory parameters at admission through an unsupervised Machine Learning (ML) approach. We aimed to identify potential indicators that could discriminate several phenotypes leading to a different responsiveness to the rehabilitation program. A k-means clustering method was performed; then, statistical analysis was employed to compare clinical and hematochemical parameters of the obtained clusters. The dataset consisted of 78 patients (84.8% males, mean age 60.72 years). The optimal number for clustering was $boldsymbol{mathrm{k}=2}$ with a silhouette coefficient of 0.85, and D-Dimer resulted the most discriminating parameter, thus confirming its role as a marker of inflammation. The phenotypes exhibited statistically significant differences in terms of age $boldsymbol{(mathrm{p}=0.007)}$, packs of cigarettes per year $boldsymbol{(mathrm{p}=0.003)}$, uricemia $boldsymbol{(mathrm{p}=0.010)}$, PCR $boldsymbol{(mathrm{p}=0.026)}$, D-Dimer $boldsymbol{(mathrm{p} < 0.001)}$, red blood cells $boldsymbol{(mathrm{p}=0.005)}$, hemoglobin $boldsymbol{(mathrm{p}=0.039)}$, hematocrit $boldsymbol{(mathrm{p}=0.026), text{PaO}_{2} (mathrm{p}=0.006)},boldsymbol{text{SpO}_{2} (mathrm{p}=0.011)}$. Overall, our findings suggest the effectiveness of ML in identifying personalized prevention, interventional and rehabilitation strategies.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"22 20","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120818552","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856555
A. Comba, A. Baldi, M. Alovisi, D. Pasqualini, E. Berutti, L. Breschi, A. Mazzoni, N. Scotti
The aim of this in vitro study was to investigate the effect of different etching times and ethanol pre-treatment on metalloproteinasis (MMPs) gelatinolytic activity on root dentin. Twelve single root teeth, extracted for periodontal reasons, were selected and an endodontic treatment was performed. After seven days, an 8-mm post space was prepared with dedicated drills. Specimens were randomly divided into four groups according to different adhesive protocols, based on different etching time in phosphoric acid and pre-treatment application. Cementation of the fiber post was performed with a dual-curing cement (DC Core, Kuraray) polymerized for 40s. In situ zymographic analyses was performed to investigate endogenous MMPs activity within the dentin hybrid layer. Quantification analyses of the MMPs activity revealed that all tested groups activated enzymes. However, the ethanol wet-bonding pretreatment was able to reduce the MMPs activity, above all when radicular dentin was extensively etched. In conclusion, MMPs gelatinolytic activity was detected in all groups and in-situ zymography was able to measure it. Further investigations are needed to clinically validate the data obtained of the present study.
{"title":"In-situ zymography to assess the MMPs activity with different etching time and ethanol wet-bonding on radicular dentin","authors":"A. Comba, A. Baldi, M. Alovisi, D. Pasqualini, E. Berutti, L. Breschi, A. Mazzoni, N. Scotti","doi":"10.1109/MeMeA54994.2022.9856555","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856555","url":null,"abstract":"The aim of this in vitro study was to investigate the effect of different etching times and ethanol pre-treatment on metalloproteinasis (MMPs) gelatinolytic activity on root dentin. Twelve single root teeth, extracted for periodontal reasons, were selected and an endodontic treatment was performed. After seven days, an 8-mm post space was prepared with dedicated drills. Specimens were randomly divided into four groups according to different adhesive protocols, based on different etching time in phosphoric acid and pre-treatment application. Cementation of the fiber post was performed with a dual-curing cement (DC Core, Kuraray) polymerized for 40s. In situ zymographic analyses was performed to investigate endogenous MMPs activity within the dentin hybrid layer. Quantification analyses of the MMPs activity revealed that all tested groups activated enzymes. However, the ethanol wet-bonding pretreatment was able to reduce the MMPs activity, above all when radicular dentin was extensively etched. In conclusion, MMPs gelatinolytic activity was detected in all groups and in-situ zymography was able to measure it. Further investigations are needed to clinically validate the data obtained of the present study.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"2018 30","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120848981","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856552
Michela Franzo', Simona Pascucci, M. Serrao, F. Marinozzi, F. Bini
Background: The purpose of a rehabilitation device is to generate tasks among cognitive difficulty stimulating the area of the cerebellum with no-automatics tasks when the situation is new to subject. An emerging technology, which overcomes these shortcomings, is the Mixed Reality. The aim of the study is to evaluate a possible updating of the prototype for rehabilitation of ataxic patients implementing exergame in Mixed Reality. Material and Methods: The version of the prototype based on Microsoft Kinect device and Arduino board with accelerometer/gyroscope sensor, presented in the previous congress, is compared with a reproduction of the same exergame in Mixed Reality environment with the HoloLens™ 2. The exergame consists in a pointing and reaching exercise to improve the control of upper limb during daily-life actions. Two subject performed the same exercise on the two different systems to investigate the differences between the systems. Results and Conclusion: The evaluation between the two systems was set up by analysing the differences between the subject's performances with the Kinect-based prototype and the HoloLens application: 3D trajectories and kinematics quantities. Despite of the restricted area of work, the high sample rate of HoloLens permits to follow much-unexpected patient's movements. The application of Mixed Reality for specific rehabilitation allows considering the requests of the therapists and the need of the patient to be always connected with the real world around him instead that in a total virtual space without real reference.
{"title":"Exergaming in mixed reality for the rehabilitation of ataxic patients","authors":"Michela Franzo', Simona Pascucci, M. Serrao, F. Marinozzi, F. Bini","doi":"10.1109/MeMeA54994.2022.9856552","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856552","url":null,"abstract":"Background: The purpose of a rehabilitation device is to generate tasks among cognitive difficulty stimulating the area of the cerebellum with no-automatics tasks when the situation is new to subject. An emerging technology, which overcomes these shortcomings, is the Mixed Reality. The aim of the study is to evaluate a possible updating of the prototype for rehabilitation of ataxic patients implementing exergame in Mixed Reality. Material and Methods: The version of the prototype based on Microsoft Kinect device and Arduino board with accelerometer/gyroscope sensor, presented in the previous congress, is compared with a reproduction of the same exergame in Mixed Reality environment with the HoloLens™ 2. The exergame consists in a pointing and reaching exercise to improve the control of upper limb during daily-life actions. Two subject performed the same exercise on the two different systems to investigate the differences between the systems. Results and Conclusion: The evaluation between the two systems was set up by analysing the differences between the subject's performances with the Kinect-based prototype and the HoloLens application: 3D trajectories and kinematics quantities. Despite of the restricted area of work, the high sample rate of HoloLens permits to follow much-unexpected patient's movements. The application of Mixed Reality for specific rehabilitation allows considering the requests of the therapists and the need of the patient to be always connected with the real world around him instead that in a total virtual space without real reference.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125991933","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856462
Luca Molinaro, Luca Russo, Francesco Cubelli, Juri Taborri, S. Rossi
Rasterstereography represents a viable alternative as screening tool for the analysis of the spinal abnormalities due to the advantages in comparison with the invasiveness of the radiology. In the last decade, several technologies have been proposed to accomplish with this aim. However, the reliability of such approach is still questioned. Tests were conducted on nine male healthy subjects, asking them to maintain four different upright positions while data was acquired by SPINE3D. Tests were repeated three times for each session, and three sessions were performed one week apart. The technologies allowed to compute indices related to transversal, sagittal and frontal plane associated with the spine posture of the subject. Reliability of the computed indices was performed only for the natural static position by using the inter-class correlation coefficient for both the intra and inter-day reliability. The results showed excellent intra-day and inter-day reliability in almost all analyzed parameters. Lower values emerged for pelvic torsion and trunk imbalance; whereas Trunk length proved to be the most reliable. Differences between NP and other positions were observed in some indices, such as pelvic and shoulder inclination, trunk length and kyphotic angle. These findings can open the possibility to use the SPINE3D as a clinical tool, also for follow-up.
{"title":"Reliability analysis of an innovative technology for the assessment of spinal abnormalities","authors":"Luca Molinaro, Luca Russo, Francesco Cubelli, Juri Taborri, S. Rossi","doi":"10.1109/MeMeA54994.2022.9856462","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856462","url":null,"abstract":"Rasterstereography represents a viable alternative as screening tool for the analysis of the spinal abnormalities due to the advantages in comparison with the invasiveness of the radiology. In the last decade, several technologies have been proposed to accomplish with this aim. However, the reliability of such approach is still questioned. Tests were conducted on nine male healthy subjects, asking them to maintain four different upright positions while data was acquired by SPINE3D. Tests were repeated three times for each session, and three sessions were performed one week apart. The technologies allowed to compute indices related to transversal, sagittal and frontal plane associated with the spine posture of the subject. Reliability of the computed indices was performed only for the natural static position by using the inter-class correlation coefficient for both the intra and inter-day reliability. The results showed excellent intra-day and inter-day reliability in almost all analyzed parameters. Lower values emerged for pelvic torsion and trunk imbalance; whereas Trunk length proved to be the most reliable. Differences between NP and other positions were observed in some indices, such as pelvic and shoulder inclination, trunk length and kyphotic angle. These findings can open the possibility to use the SPINE3D as a clinical tool, also for follow-up.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"124 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129569043","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856418
Ilaria Marcantoni, Giorgia Barchiesi, Sofia Barchiesi, Caterina Belbusti, Chiara Leoni, Sofia Romagnoli, A. Sbrollini, M. Morettini, L. Burattini
The development of on-board car electronics for automatic stress level detection is becoming an area of great interest. The literature showed that biomedical signal acquisition could provide significant information. Skin conductance (SC) and electrocardiogram (ECG) have demonstrated to provide the most significant stress-related features. Thus, the aim of this study is the classification of three-level and binary stress, using a minimal combination of SC and ECG features. The “Stress Recognition in Automobile Drivers” database was used to test a procedure based on linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA). The database protocol includes three driving periods, corresponding to different levels of stress (low-medium-high). After data preprocessing, LDA and QDA three-level classifications were applied on all the extracted SC and ECG features to determine the best classification approach. Boruta algorithm allowed to select the most significant features for the classification. Then, the best classification approach was applied on this restricted set of features, performing both three-level (low vs medium vs high) and binary (high+medium vs low) stress classification. QDA was the most accurate classification method (accuracy: 96.0% for QDA vs 85.3% for LDA, considering all the features). QDA accuracy, considering only the selected features, was 86.7% for the three-level classification and 94.7% for the binary classification. This result represents an acceptable trade-off between classification accuracy and computational cost, associated to the number of considered features. In conclusion, ECG together with SC are suitable for the objective and automatic identification and classification of driving-related stress with a good accuracy.
{"title":"Identification and Classification of Driving-Related Stress Using Electrocardiogram and Skin Conductance Signals","authors":"Ilaria Marcantoni, Giorgia Barchiesi, Sofia Barchiesi, Caterina Belbusti, Chiara Leoni, Sofia Romagnoli, A. Sbrollini, M. Morettini, L. Burattini","doi":"10.1109/MeMeA54994.2022.9856418","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856418","url":null,"abstract":"The development of on-board car electronics for automatic stress level detection is becoming an area of great interest. The literature showed that biomedical signal acquisition could provide significant information. Skin conductance (SC) and electrocardiogram (ECG) have demonstrated to provide the most significant stress-related features. Thus, the aim of this study is the classification of three-level and binary stress, using a minimal combination of SC and ECG features. The “Stress Recognition in Automobile Drivers” database was used to test a procedure based on linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA). The database protocol includes three driving periods, corresponding to different levels of stress (low-medium-high). After data preprocessing, LDA and QDA three-level classifications were applied on all the extracted SC and ECG features to determine the best classification approach. Boruta algorithm allowed to select the most significant features for the classification. Then, the best classification approach was applied on this restricted set of features, performing both three-level (low vs medium vs high) and binary (high+medium vs low) stress classification. QDA was the most accurate classification method (accuracy: 96.0% for QDA vs 85.3% for LDA, considering all the features). QDA accuracy, considering only the selected features, was 86.7% for the three-level classification and 94.7% for the binary classification. This result represents an acceptable trade-off between classification accuracy and computational cost, associated to the number of considered features. In conclusion, ECG together with SC are suitable for the objective and automatic identification and classification of driving-related stress with a good accuracy.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129067172","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856563
Ritika Jain, R. Ganesan
This study explores automated sleep-wake classification using Poincaré plots derived from a single EEG channel. In order to quantify the Poincaré plots and utilize them for the distinction of sleep and wake states of the healthy individuals and patients with sleep disorders, various descriptors are computed. The most commonly used standard descriptors are SD1 and SD2, which determine the width and length of Poincaré plot. Along with SD1 and SD2, the ratio of SD1 to SD2, area of the Poincaré plots, energy of the slopes, and offsets obtained by linear fits to Poincaré plots with distinct lags, standard deviation, and complex correlation measure are also computed. Random undersampling with boosting technique (RUSBoost) is adopted to deal with the class imbalance problem. The performance of the method is evaluated on three different publicly available datasets by using 50%-holdout and 10-fold crossvalidation techniques. We achieved crossvalidation accuracies of 98.2%, 96.0%, and 94.4% for Sleep-EDF, DREAMS-Subjects and DREAMS-Patients datasets, respectively, by utilizing only eight features, and a single EEG channel. Furthermore, for the patient population with various sleep disorders such as mixed apnea, periodic leg movement syndrome, sleep apnea-hypopnea syndrome, and dyssomnia, we obtained average sensitivity of 96.8%, precision of 95.6%, and F1-score of 96.2%, for the sleep state; and 88.3%, 91.3%, and 89.8%, respectively for the wake state. Our results are comparable to or better than the existing studies in the literature. Further, the classification accuracies for the patients with a model trained only on the healthy population are quite impressive. Thus, the model is effective and generalizes well for the patient population.
{"title":"Poincaré plot analysis for sleep-wake classification of unseen patients using a single EEG channel","authors":"Ritika Jain, R. Ganesan","doi":"10.1109/MeMeA54994.2022.9856563","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856563","url":null,"abstract":"This study explores automated sleep-wake classification using Poincaré plots derived from a single EEG channel. In order to quantify the Poincaré plots and utilize them for the distinction of sleep and wake states of the healthy individuals and patients with sleep disorders, various descriptors are computed. The most commonly used standard descriptors are SD1 and SD2, which determine the width and length of Poincaré plot. Along with SD1 and SD2, the ratio of SD1 to SD2, area of the Poincaré plots, energy of the slopes, and offsets obtained by linear fits to Poincaré plots with distinct lags, standard deviation, and complex correlation measure are also computed. Random undersampling with boosting technique (RUSBoost) is adopted to deal with the class imbalance problem. The performance of the method is evaluated on three different publicly available datasets by using 50%-holdout and 10-fold crossvalidation techniques. We achieved crossvalidation accuracies of 98.2%, 96.0%, and 94.4% for Sleep-EDF, DREAMS-Subjects and DREAMS-Patients datasets, respectively, by utilizing only eight features, and a single EEG channel. Furthermore, for the patient population with various sleep disorders such as mixed apnea, periodic leg movement syndrome, sleep apnea-hypopnea syndrome, and dyssomnia, we obtained average sensitivity of 96.8%, precision of 95.6%, and F1-score of 96.2%, for the sleep state; and 88.3%, 91.3%, and 89.8%, respectively for the wake state. Our results are comparable to or better than the existing studies in the literature. Further, the classification accuracies for the patients with a model trained only on the healthy population are quite impressive. Thus, the model is effective and generalizes well for the patient population.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127975892","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 : 2022-06-22DOI: 10.1109/MeMeA54994.2022.9856459
P. Casti, A. Mencattini, Sara Cardarelli, G. Antonelli, J. Filippi, M. D’Orazio, E. Martinelli
The advances in the deep learning field have paved the way to novel strategies to represent digital image data in the form of synthetic descriptors. Variational Auto-Encoders (VAE) architectures are generative powerful tools not only to reconstruct input images but also to extract meaningful information for the task of pattern classification. The first part of the VAE network, called encoder, aims to condense the image information into a reduced set of low-level descriptors, called latent variables. The second part, called decoder, aims to use the latent variable in a reverse process that reconstructs the original image in output. In this work, we exploited the VAE-based latent representation of colour normalized dermoscopic images for the discrimination of malignant and benign skin lesions. In particular, we investigated the sensitivity to the effect of skin colour variations over the final reconstruction error and on the discrimination capability of the VAE latent variables in terms of individual Area Under the roC curve (AUC). By exploiting and adapting state-of-the art skin colour variation models we obtained a performance worsening of about 10% either in the reconstruction error and in the discrimination capability of the latent variables. The achieved preliminary results demonstrate that, with suitable VAE adaptation, latent descriptors could be used in automatic skin lesions classification frameworks.
{"title":"Sensitivity analysis of latent variables in Variational Autoencoders for Dermoscopic Image Analysis","authors":"P. Casti, A. Mencattini, Sara Cardarelli, G. Antonelli, J. Filippi, M. D’Orazio, E. Martinelli","doi":"10.1109/MeMeA54994.2022.9856459","DOIUrl":"https://doi.org/10.1109/MeMeA54994.2022.9856459","url":null,"abstract":"The advances in the deep learning field have paved the way to novel strategies to represent digital image data in the form of synthetic descriptors. Variational Auto-Encoders (VAE) architectures are generative powerful tools not only to reconstruct input images but also to extract meaningful information for the task of pattern classification. The first part of the VAE network, called encoder, aims to condense the image information into a reduced set of low-level descriptors, called latent variables. The second part, called decoder, aims to use the latent variable in a reverse process that reconstructs the original image in output. In this work, we exploited the VAE-based latent representation of colour normalized dermoscopic images for the discrimination of malignant and benign skin lesions. In particular, we investigated the sensitivity to the effect of skin colour variations over the final reconstruction error and on the discrimination capability of the VAE latent variables in terms of individual Area Under the roC curve (AUC). By exploiting and adapting state-of-the art skin colour variation models we obtained a performance worsening of about 10% either in the reconstruction error and in the discrimination capability of the latent variables. The achieved preliminary results demonstrate that, with suitable VAE adaptation, latent descriptors could be used in automatic skin lesions classification frameworks.","PeriodicalId":106228,"journal":{"name":"2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131278758","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：