Pub Date : 2025-01-01DOI: 10.1016/j.bbe.2024.12.001
Min-Kyoung Kang , Keum-Shik Hong , Dalin Yang , Ho Kyung Kim
Mild cognitive impairment (MCI) is recognized as an early stage preceding Alzheimer’s disease. Functional near-infrared spectroscopy (fNIRS) has recently been used to differentiate MCI patients from healthy controls (HCs) by analyzing their hemodynamic responses. This paper proposes a new method that uses the entire time series data from all fNIRS channels, skipping the feature extraction step. It involves a multi-scale convolutional neural network (CNN) integrated with long short-term memory (LSTM) layers to extract spatial and temporal features simultaneously. The study involves 64 participants (37 MCI patients and 27 HCs) performing three mental tasks: N-back, Stroop, and verbal fluency tests (VFT). The algorithm’s performance was assessed using 10-fold cross-validation across oxyhemoglobin (HbO), deoxyhemoglobin (HbR), and total hemoglobin (HbT). The highest classification accuracies were achieved with HbT, reaching 93.22 % for the N-back task, 91.14 % for the Stroop task, and 89.58 % for the VFT. It was found that using all types of hemodynamic signals from all channels provides better results than analyzing the region of interest data, eliminating the need for data segmentation and feature extraction procedures. Additionally, HbR (or HbT) gives better classification accuracy than HbO. The developed method can be implemented online for clinical applications and real-time monitoring of cognitive disorders.
{"title":"Multi-scale neural networks classification of mild cognitive impairment using functional near-infrared spectroscopy","authors":"Min-Kyoung Kang , Keum-Shik Hong , Dalin Yang , Ho Kyung Kim","doi":"10.1016/j.bbe.2024.12.001","DOIUrl":"10.1016/j.bbe.2024.12.001","url":null,"abstract":"<div><div>Mild cognitive impairment (MCI) is recognized as an early stage preceding Alzheimer’s disease. Functional near-infrared spectroscopy (fNIRS) has recently been used to differentiate MCI patients from healthy controls (HCs) by analyzing their hemodynamic responses. This paper proposes a new method that uses the entire time series data from all fNIRS channels, skipping the feature extraction step. It involves a multi-scale convolutional neural network (CNN) integrated with long short-term memory (LSTM) layers to extract spatial and temporal features simultaneously. The study involves 64 participants (37 MCI patients and 27 HCs) performing three mental tasks: <em>N</em>-back, Stroop, and verbal fluency tests (VFT). The algorithm’s performance was assessed using 10-fold cross-validation across oxyhemoglobin (HbO), deoxyhemoglobin (HbR), and total hemoglobin (HbT). The highest classification accuracies were achieved with HbT, reaching 93.22 % for the <em>N</em>-back task, 91.14 % for the Stroop task, and 89.58 % for the VFT. It was found that using all types of hemodynamic signals from all channels provides better results than analyzing the region of interest data, eliminating the need for data segmentation and feature extraction procedures. Additionally, HbR (or HbT) gives better classification accuracy than HbO. The developed method can be implemented online for clinical applications and real-time monitoring of cognitive disorders.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"45 1","pages":"Pages 11-22"},"PeriodicalIF":5.3,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.bbe.2025.01.005
J.D. Chiza-Ocaña , G. Realpe , C.A. López-Albán , E. Rosero , J.M. Ramírez-Scarpetta
This paper presents a two state quasi-linear parameter varying dynamic model for gas exchange dynamics using the cycle-ergometer test. The obtained model, is based on the analysis of stationary and dynamic energy flow, and the method analysis, applies to both oxidative and glycolytic physical activities performed by an individual. The model parameters were identified by a power meter measuring the mechanical power at the pedal level on an ergometer bicycle (input signal), a commercial gas analyzer measuring the flow of oxygen uptake and the flow of carbon dioxide excreted (output signals), with data generated from two test protocols: a mixed protocol and an incremental cycling protocol. The model’s parameters are obtained in parts, from the measurements taken in the oxidative stage, the glycolytic stage, and the transition stage between the two, using the mixed protocol. The resulting model is validated using data from the incremental cycling protocol of nine individuals: six males and three females. The validated models obtained an accuracy of above 84.8% for the flow of oxygen and 89.1% for the flow of carbon dioxide. The dynamic model could be used to aid in creating personalized physical exercise programs for overweight individuals, simulating training plans within the operational thresholds of the human body or in structuring high performance training for athletes.
{"title":"Two state quasi-LPV dynamic model for gas exchange dynamics using the cycle-ergometer test","authors":"J.D. Chiza-Ocaña , G. Realpe , C.A. López-Albán , E. Rosero , J.M. Ramírez-Scarpetta","doi":"10.1016/j.bbe.2025.01.005","DOIUrl":"10.1016/j.bbe.2025.01.005","url":null,"abstract":"<div><div>This paper presents a two state quasi-linear parameter varying <span><math><mrow><mo>(</mo><mi>q</mi><mi>u</mi><mi>a</mi><mi>s</mi><mi>i</mi><mo>−</mo><mi>L</mi><mi>P</mi><mi>V</mi><mo>)</mo></mrow></math></span> dynamic model for gas exchange dynamics using the cycle-ergometer test. The obtained model, is based on the analysis of stationary and dynamic energy flow, and the <span><math><mrow><mi>V</mi><mo>−</mo><mi>s</mi><mi>l</mi><mi>o</mi><mi>p</mi><mi>e</mi></mrow></math></span> method analysis, applies to both oxidative and glycolytic physical activities performed by an individual. The model parameters were identified by a power meter measuring the mechanical power at the pedal level on an ergometer bicycle (input signal), a commercial gas analyzer measuring the flow of oxygen uptake and the flow of carbon dioxide excreted (output signals), with data generated from two test protocols: a mixed protocol and an incremental cycling protocol. The model’s parameters are obtained in parts, from the measurements taken in the oxidative stage, the glycolytic stage, and the transition stage between the two, using the mixed protocol. The resulting model is validated using data from the incremental cycling protocol of nine individuals: six males and three females. The validated models obtained an accuracy of above 84.8% for the flow of oxygen and 89.1% for the flow of carbon dioxide. The dynamic model could be used to aid in creating personalized physical exercise programs for overweight individuals, simulating training plans within the operational thresholds of the human body or in structuring high performance training for athletes.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"45 1","pages":"Pages 105-113"},"PeriodicalIF":5.3,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143422720","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.bbe.2025.01.002
Konrad Kwiecień , Karolina Knap , Rick Heida , Jonasz Czajkowski , Alan Gorter , Dorota Ochońska , Przemysław Mielczarek , Agata Dorosz , Daria Niewolik , Katarzyna Reczyńska-Kolman , Katarzyna Jaszcz , Monika Brzychczy-Włoch , Tomasz R. Sosnowski , Peter Olinga , Elżbieta Pamuła
By many chronic lung diseases, there is a problem of recurrent bacterial infections that require frequent usage of antibiotics. They can be more effective and cause fewer side effects when administrated directly via the pulmonary route. For such purposes, various types of inhalers are used of which dry powder inhalers (DPIs) are one of the most common. Formulations such as dry powders usually consist of an active pharmaceutical ingredient (API) and a carrier material that is supposed to provide adequate properties to deliver the bioactive molecules to the site of action, effectively. Copolymers of sebacic acid (SA) and poly(ethylene glycol) (PEG) have been regarded as suitable materials for such formulations. Here, we present a study about the manufacturing of microparticles from such materials dedicated to inhalation which have been loaded with azithromycin (AZM). The microparticles (MPs) were 0.5 to 5 µm in size, presenting either a spherical or elongated shape depending on the material type and composition. The encapsulation efficiency (EE) of the MPs were almost complete with the drug loading up to 23.1 %. The powders had fair or good flowability based on Carr’s index and Hausner ratio. Due to the presence of the drug, the tendency to agglomerate decreased. As a result, up to 90 % of the obtained powders showed diameters below 5 µm. Also, the fine particles fraction (FPF) of the chosen formulation reached 66.3 ± 4.5 % and the mass median aerodynamic diameter was 3.8 ± 0.4 µm. The microparticles degraded quickly in vitro losing up to 50 % of their mass within 24 h and up to 80 % within 96 h of their incubation in phosphate-buffered saline (PBS). They were also nontoxic up to 100 µg/ml when added to cultures of A549 and BEAS-2B lung epithelial cells as well as to rat lung tissue slices tested ex vivo. The microparticles showed bactericidal effects against various strains of Staphylococcus aureus in lower than cytotoxic concentrations.
{"title":"Novel copolymers of poly(sebacic anhydride) and poly(ethylene glycol) as azithromycin carriers to the lungs","authors":"Konrad Kwiecień , Karolina Knap , Rick Heida , Jonasz Czajkowski , Alan Gorter , Dorota Ochońska , Przemysław Mielczarek , Agata Dorosz , Daria Niewolik , Katarzyna Reczyńska-Kolman , Katarzyna Jaszcz , Monika Brzychczy-Włoch , Tomasz R. Sosnowski , Peter Olinga , Elżbieta Pamuła","doi":"10.1016/j.bbe.2025.01.002","DOIUrl":"10.1016/j.bbe.2025.01.002","url":null,"abstract":"<div><div>By many chronic lung diseases, there is a problem of recurrent bacterial infections that require frequent usage of antibiotics. They can be more effective and cause fewer side effects when administrated directly via the pulmonary route. For such purposes, various types of inhalers are used of which dry powder inhalers (DPIs) are one of the most common. Formulations such as dry powders usually consist of an active pharmaceutical ingredient (API) and a carrier material that is supposed to provide adequate properties to deliver the bioactive molecules to the site of action, effectively. Copolymers of sebacic acid (SA) and poly(ethylene glycol) (PEG) have been regarded as suitable materials for such formulations. Here, we present a study about the manufacturing of microparticles from such materials dedicated to inhalation which have been loaded with azithromycin (AZM). The microparticles (MPs) were 0.5 to 5 µm in size, presenting either a spherical or elongated shape depending on the material type and composition. The encapsulation efficiency (EE) of the MPs were almost complete with the drug loading up to 23.1 %. The powders had fair or good flowability based on Carr’s index and Hausner ratio. Due to the presence of the drug, the tendency to agglomerate decreased. As a result, up to 90 % of the obtained powders showed diameters below 5 µm. Also, the fine particles fraction (FPF) of the chosen formulation reached 66.3 ± 4.5 % and the mass median aerodynamic diameter was 3.8 ± 0.4 µm. The microparticles degraded quickly <em>in vitro</em> losing up to 50 % of their mass within 24 h and up to 80 % within 96 h of their incubation in phosphate-buffered saline (PBS). They were also nontoxic up to 100 µg/ml when added to cultures of A549 and BEAS-2B lung epithelial cells as well as to rat lung tissue slices tested <em>ex vivo</em>. The microparticles showed bactericidal effects against various strains of <em>Staphylococcus aureus</em> in lower than cytotoxic concentrations.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"45 1","pages":"Pages 114-136"},"PeriodicalIF":5.3,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143422108","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.bbe.2025.01.003
Fufei Li , Li Chen , Ge Song , Lianzheng Su , Shian Wang , Qiuyue Fu , Yongqi Nie , Peng Wang
Diagnosing renal and urinary system illnesses usually entails analysing the sediment found in urine. The components in microscopic urine images are diverse and show high similarity, with low contrast due to noise, impeding the progress of automated urine analysis. In order to tackle this difficulty, we propose a region-constrained consistency contrastive learning approach for automated urine analysis. In the first stage, we tackle the complex overlap phenomena in microscopic urine images by innovating the Urine Sediment Paste (US-Paste) positive sample construction method based on supervised contrastive learning. This method uses label information to apply regional constraints and improves the performance of out-of-distribution detection. We also rebuilt the Global Guidance Module (GG Module) and the Enhanced Supervision Module(ES Module). The former improves contrast in urine sediment images by restoring important image details guided by an encoder–decoder structure, while the latter achieves strong feature consistency by combining the most pertinent feature responses from four sets of attention feature maps, which are further mapped via a projection network. In the second phase, we enhance the representations acquired in the initial phase by incorporating a linear classification layer. Our region-constrained consistency contrastive learning algorithm attained an average classification accuracy of 98.30%, precision of 98.33%, recall of 98.30%, and F1-score of 98.30% on the private dataset. Furthermore, in the public urine sediment dataset, the approach achieved an average classification accuracy of 96.19%, precision of 95.79%, recall of 96.19%, and F1-score of 95.94%. The public chromosomal dataset yielded an average classification accuracy of 95.46%, precision of 94.84%, recall of 95.47%, and F1-score of 95.15%. Our methodology surpasses the most advanced methods and demonstrates exceptional performance in urine analysis. This showcases the efficiency of our label-based regional limitations, the outstanding out-of-distribution detection performance of US-Paste, and the robust feature consistency achieved by the Guided Supervision Encoder (GS Encoder). This substantially enhances diagnostic efficiency for clinicians and significantly advances the progress of automated urine sediment analysis.
{"title":"Regional constraint consistency contrastive learning for automatic detection of urinary sediment in microscopic images","authors":"Fufei Li , Li Chen , Ge Song , Lianzheng Su , Shian Wang , Qiuyue Fu , Yongqi Nie , Peng Wang","doi":"10.1016/j.bbe.2025.01.003","DOIUrl":"10.1016/j.bbe.2025.01.003","url":null,"abstract":"<div><div>Diagnosing renal and urinary system illnesses usually entails analysing the sediment found in urine. The components in microscopic urine images are diverse and show high similarity, with low contrast due to noise, impeding the progress of automated urine analysis. In order to tackle this difficulty, we propose a region-constrained consistency contrastive learning approach for automated urine analysis. In the first stage, we tackle the complex overlap phenomena in microscopic urine images by innovating the Urine Sediment Paste (US-Paste) positive sample construction method based on supervised contrastive learning. This method uses label information to apply regional constraints and improves the performance of out-of-distribution detection. We also rebuilt the Global Guidance Module (GG Module) and the Enhanced Supervision Module(ES Module). The former improves contrast in urine sediment images by restoring important image details guided by an encoder–decoder structure, while the latter achieves strong feature consistency by combining the most pertinent feature responses from four sets of attention feature maps, which are further mapped via a projection network. In the second phase, we enhance the representations acquired in the initial phase by incorporating a linear classification layer. Our region-constrained consistency contrastive learning algorithm attained an average classification accuracy of 98.30%, precision of 98.33%, recall of 98.30%, and F1-score of 98.30% on the private dataset. Furthermore, in the public urine sediment dataset, the approach achieved an average classification accuracy of 96.19%, precision of 95.79%, recall of 96.19%, and F1-score of 95.94%. The public chromosomal dataset yielded an average classification accuracy of 95.46%, precision of 94.84%, recall of 95.47%, and F1-score of 95.15%. Our methodology surpasses the most advanced methods and demonstrates exceptional performance in urine analysis. This showcases the efficiency of our label-based regional limitations, the outstanding out-of-distribution detection performance of US-Paste, and the robust feature consistency achieved by the Guided Supervision Encoder (GS Encoder). This substantially enhances diagnostic efficiency for clinicians and significantly advances the progress of automated urine sediment analysis.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"45 1","pages":"Pages 74-89"},"PeriodicalIF":5.3,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143092967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.bbe.2024.12.002
R. Sudha , K.M. Uma Maheswari
Lung malignant tumors are abnormal growths of cells in the lungs that have the potential to invade nearby tissues and spread to other parts of the body. Early detection of these malignant lung tumors is crucial to avoid complications and improve patient outcomes. However, manual processing consumes time and is a tedious process. This might result in poor estimation on cancer-prognosis, leading the patients into a higher risk of mortality. Many existing literatures have detected the malignant tumors, yet, found certain difficulties with the identification of size, appearance and spread of cancerous-cells in lung region to determine how far it has been occupied. Hence, the present study aims to overcome the existing complications through Deep Learning based Swarm Intelligence Algorithms. Implementation of the proposed work is involved with three stages such as pre-processing, segmentation and classification. Besides, CT scan possess the capability for giving a comprehensive view than X-rays. Data are collected from LIDC-IDRI (Lung Image Database Consortium-Image Database Resource Initiative) with lung CT-images and accomplishes pre-processing by removing noise efficiently using wiener filter. Further, changes in soft tissues of lungs are identified and segmented in the subsequent phase using U-Net and finally classification is performed using CFSO (Convolutional Neural Network Fish Swarm Optimization) to overcome the slight chance of misclassification error as proposed CFSO can lead to more efficient computational processes since FSO algorithms are designed to minimize computational costs while maximizing performance through their metaheuristic nature. This efficiency is particularly beneficial when dealing with large datasets typical in medical imaging, allowing faster processing times without sacrificing accuracy. Hence, amalgamation of CFSO can reduce the number of features, thus speeding up training and inference times. Through the performance assessment, IoU (Intersection over Union) value attained through the analysis is found to be 0.7822. Further, accuracy obtained by the proposed model is 97.80%, recall is 98.49%, precision is 96.8% and F1-score is 97.32%. Findings of the study exhibits the purposefulness of the study in clinical settings by potentially reducing false negatives in lung cancer screening, ultimately improving patient survival rates through earlier detection and treatment.
{"title":"Analysis on grading of lung nodule images with segmentation using u-net and classification with Convolutional Neural Network Fish Swarm Optimization","authors":"R. Sudha , K.M. Uma Maheswari","doi":"10.1016/j.bbe.2024.12.002","DOIUrl":"10.1016/j.bbe.2024.12.002","url":null,"abstract":"<div><div>Lung malignant tumors are abnormal growths of cells in the lungs that have the potential to invade nearby tissues and spread to other parts of the body. Early detection of these malignant lung tumors is crucial to avoid complications and improve patient outcomes. However, manual processing consumes time and is a tedious process. This might result in poor estimation on cancer-prognosis, leading the patients into a higher risk of mortality. Many existing literatures have detected the malignant tumors, yet, found certain difficulties with the identification of size, appearance and spread of cancerous-cells in lung region to determine how far it has been occupied. Hence, the present study aims to overcome the existing complications through Deep Learning based Swarm Intelligence Algorithms. Implementation of the proposed work is involved with three stages such as pre-processing, segmentation and classification. Besides, CT scan possess the capability for giving a comprehensive view than X-rays. Data are collected from LIDC-IDRI (Lung Image Database Consortium-Image Database Resource Initiative) with lung CT-images and accomplishes pre-processing by removing noise efficiently using wiener filter. Further, changes in soft tissues of lungs are identified and segmented in the subsequent phase using U-Net and finally classification is performed using CFSO (Convolutional Neural Network Fish Swarm Optimization) to overcome the slight chance of misclassification error as proposed CFSO can lead to more efficient computational processes since FSO algorithms are designed to minimize computational costs while maximizing performance through their metaheuristic nature. This efficiency is particularly beneficial when dealing with large datasets typical in medical imaging, allowing faster processing times without sacrificing accuracy. Hence, amalgamation of CFSO can reduce the number of features, thus speeding up training and inference times. Through the performance assessment, IoU (Intersection over Union) value attained through the analysis is found to be 0.7822. Further, accuracy obtained by the proposed model is 97.80%, recall is 98.49%, precision is 96.8% and F1-score is 97.32%. Findings of the study exhibits the purposefulness of the study in clinical settings by potentially reducing false negatives in lung cancer screening, ultimately improving patient survival rates through earlier detection and treatment.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"45 1","pages":"Pages 90-104"},"PeriodicalIF":5.3,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143127983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-01DOI: 10.1016/j.bbe.2024.09.003
Diego Castillo-Barnes , Nicolás J. Gallego-Molina , Marco A. Formoso , Andrés Ortiz , Patrícia Figueiredo , Juan L. Luque
This work explores the intricate neural dynamics associated with dyslexia through the lens of Cross-Frequency Coupling (CFC) analysis applied to electroencephalography (EEG) signals evaluated from 48 seven-year-old Spanish readers from the LEEDUCA research platform. The analysis focuses on CFS (Cross-Frequency phase Synchronization) maps, capturing the interaction between different frequency bands during low-level auditory processing stimuli. Then, making use of Gaussian Mixture Models (GMMs), CFS activations are quantified and classified, offering a compressed representation of EEG activation maps. The study unveils promising results specially at the Theta-Gamma coupling (Area Under the Curve = 0.821), demonstrating the method’s sensitivity to dyslexia-related neural patterns and highlighting potential applications in the early identification of dyslexic individuals.
{"title":"Probabilistic and explainable modeling of Phase–Phase Cross-Frequency Coupling patterns in EEG. Application to dyslexia diagnosis","authors":"Diego Castillo-Barnes , Nicolás J. Gallego-Molina , Marco A. Formoso , Andrés Ortiz , Patrícia Figueiredo , Juan L. Luque","doi":"10.1016/j.bbe.2024.09.003","DOIUrl":"10.1016/j.bbe.2024.09.003","url":null,"abstract":"<div><div>This work explores the intricate neural dynamics associated with dyslexia through the lens of Cross-Frequency Coupling (CFC) analysis applied to electroencephalography (EEG) signals evaluated from 48 seven-year-old Spanish readers from the LEEDUCA research platform. The analysis focuses on CFS (Cross-Frequency phase Synchronization) maps, capturing the interaction between different frequency bands during low-level auditory processing stimuli. Then, making use of Gaussian Mixture Models (GMMs), CFS activations are quantified and classified, offering a compressed representation of EEG activation maps. The study unveils promising results specially at the Theta-Gamma coupling (Area Under the Curve = 0.821), demonstrating the method’s sensitivity to dyslexia-related neural patterns and highlighting potential applications in the early identification of dyslexic individuals.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"44 4","pages":"Pages 814-823"},"PeriodicalIF":5.3,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142535970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-01DOI: 10.1016/j.bbe.2024.11.003
Hongyuan Zhang , Zijian Zhao , Chong Liu , Miao Duan , Zhiguo Lu , Hong Wang
A brain-computer interface (BCI) is a technology that creates a communication path between the brain and external devices. Raw EEG data in BCI contain a large amount of complex information, but only some of it needs to be focused on in research. So Feature extraction and classification play an important role in BCI by reducing the data dimensionality and improving the accuracy of subsequent classification. Wavelet scattering transform is an emerging feature extraction method that generates time-shift invariant representations of EEG signals. We applied the wavelet scattering transform to extract features from motor imagery EEG signals, and utilized these features for classification purposes. To achieve this, we proposed a new method that combines wavelet scattering transform with a bidirectional long short-term memory (BiLSTM) network in a fusion deep learning network. Wavelet scattering transform can deeply mine the feature information in EEG signals. In the classification stage, multiple time window features obtained in the scattering transform are sent to the BiLSTM network for classification. The final result will be determined by a vote. In addition, for the processing of raw EEG data, we proposed a time-step based time window strategy that can better utilize the small dataset. This operation can obtain EEG data of multiple time steps. The proposed method was validated using BCI competition II dataset III and BCI competition IV dataset 2b. The results show that the proposed method in this paper can effectively improve the accuracy of motor imagery EEG and provide a new idea for the feature extraction and classification research of motor imagery brain-computer interface.
{"title":"Classification of motor imagery EEG signals using wavelet scattering transform and Bi-directional long short-term memory networks","authors":"Hongyuan Zhang , Zijian Zhao , Chong Liu , Miao Duan , Zhiguo Lu , Hong Wang","doi":"10.1016/j.bbe.2024.11.003","DOIUrl":"10.1016/j.bbe.2024.11.003","url":null,"abstract":"<div><div>A brain-computer interface (BCI) is a technology that creates a communication path between the brain and external devices. Raw EEG data in BCI contain a large amount of complex information, but only some of it needs to be focused on in research. So Feature extraction and classification play an important role in BCI by reducing the data dimensionality and improving the accuracy of subsequent classification. Wavelet scattering transform is an emerging feature extraction method that generates time-shift invariant representations of EEG signals. We applied the wavelet scattering transform to extract features from motor imagery EEG signals, and utilized these features for classification purposes. To achieve this, we proposed a new method that combines wavelet scattering transform with a bidirectional long short-term memory (BiLSTM) network in a fusion deep learning network. Wavelet scattering transform can deeply mine the feature information in EEG signals. In the classification stage, multiple time window features obtained in the scattering transform are sent to the BiLSTM network for classification. The final result will be determined by a vote. In addition, for the processing of raw EEG data, we proposed a time-step based time window strategy that can better utilize the small dataset. This operation can obtain EEG data of multiple time steps. The proposed method was validated using BCI competition II dataset III and BCI competition IV dataset 2b. The results show that the proposed method in this paper can effectively improve the accuracy of motor imagery EEG and provide a new idea for the feature extraction and classification research of motor imagery brain-computer interface.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"44 4","pages":"Pages 874-884"},"PeriodicalIF":5.3,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143158887","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-01DOI: 10.1016/j.bbe.2024.09.005
Sabri Altunkaya
Hippocampal field potentials are widely used in research on neurodegenerative diseases, epilepsy, neuropharmacology, and particularly long- and short-term synaptic plasticity. To conduct these studies, it is necessary to identify specific components within hippocampal field potential signals. However, manually marking the relevant signal points for analysis is a time-consuming, error-prone, and subjective process. Currently, there is no specialized software dedicated to automating this task. In this study, three different recurrent neural network-based deep learning architectures were examined for the automatic segmentation of hippocampal field potential signals in two separate experimental studies. In the first experimental study, 10,836 epochs of field potential signals recorded from 54 rats were used, and in the second experimental study, field potential signals with noise added to the above data at different rates were used. The best model achieved an average f-score of 98.1% on noise-free data and 97.15% on data with noise, highlighting its robustness in real-world scenarios. Furthermore, we assessed system stability using the repeated holdout method, which randomly split the data into training and testing sets 100 times, and each time trained a new version of the system. As a result, the proposed system was proven to be reliable and generalizable by showing similar average scores and low variability across all 100 iterations of the test.
海马场电位被广泛应用于神经退行性疾病、癫痫、神经药理学，特别是长短期突触可塑性的研究。要进行这些研究，就必须识别海马场电位信号中的特定成分。然而，手动标记相关信号点进行分析是一个耗时、易出错且主观的过程。目前，还没有专门的软件来自动完成这项任务。本研究在两项独立的实验研究中，考察了三种不同的基于递归神经网络的深度学习架构，用于自动分割海马场电位信号。在第一项实验研究中，使用了 54 只大鼠记录的 10836 个历时场电位信号；在第二项实验研究中，使用了以不同速率向上述数据添加噪声的场电位信号。最佳模型在无噪声数据上的平均 f 分数为 98.1%，在有噪声数据上的平均 f 分数为 97.15%，这突出表明了该模型在实际应用中的鲁棒性。此外，我们还使用重复保持法评估了系统的稳定性，该方法将数据随机分为训练集和测试集 100 次，每次训练一个新版本的系统。结果表明，所提出的系统在所有 100 次迭代测试中显示出相似的平均得分和较低的变异性，从而证明了该系统的可靠性和通用性。
{"title":"Automating synaptic plasticity analysis: A deep learning approach to segmenting hippocampal field potential signal","authors":"Sabri Altunkaya","doi":"10.1016/j.bbe.2024.09.005","DOIUrl":"10.1016/j.bbe.2024.09.005","url":null,"abstract":"<div><div>Hippocampal field potentials are widely used in research on neurodegenerative diseases, epilepsy, neuropharmacology, and particularly long- and short-term synaptic plasticity. To conduct these studies, it is necessary to identify specific components within hippocampal field potential signals. However, manually marking the relevant signal points for analysis is a time-consuming, error-prone, and subjective process. Currently, there is no specialized software dedicated to automating this task. In this study, three different recurrent neural network-based deep learning architectures were examined for the automatic segmentation of hippocampal field potential signals in two separate experimental studies. In the first experimental study, 10,836 epochs of field potential signals recorded from 54 rats were used, and in the second experimental study, field potential signals with noise added to the above data at different rates were used. The best model achieved an average f-score of 98.1% on noise-free data and 97.15% on data with noise, highlighting its robustness in real-world scenarios. Furthermore, we assessed system stability using the repeated holdout method, which randomly split the data into training and testing sets 100 times, and each time trained a new version of the system. As a result, the proposed system was proven to be reliable and generalizable by showing similar average scores and low variability across all 100 iterations of the test.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"44 4","pages":"Pages 804-813"},"PeriodicalIF":5.3,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-01DOI: 10.1016/j.bbe.2024.10.002
Mauro Pietribiasi , John K. Leypoldt , Monika Wieliczko , Malgorzata Twardowska-Kawalec , Malgorzata Debowska , Jolanta Malyszko , Jacek Waniewski
Background
Delivery of bicarbonate during hemodialysis (HD) is aimed at correcting metabolic acidosis in end-stage renal disease patients. We tested modified prescriptions of bicarbonate concentration in dialysis fluid (CD,bic), aimed to achieve an optimal pre-dialytic bicarbonate plasma concentration (CP,bic).
Methods
We used a mathematical model to prescribe individualized HD treatments consisting of 1) adjustment of CD,bic to get the pre-dialytic CP,bic in a prescribed range, 2) increase of bicarbonate load before the long interdialytic break, and 3) a single step of increase in CD,bic after two hours. The outcomes were tested in 24 stable HD patients, monitored during a week of standard HD (Test Week) and a week of modified treatment (Intervention Week).
Results
The response to the model-based prescription was different whether the average CD,bic during the Intervention Week was higher or lower than the constant value used for the Test Week. For patients with lower average CD,bic during the Intervention Week, a significant fraction achieved the target (22 ≤ CP,bic ≤ 24 mEq/L). In the group with higher average CD,bic, the interventions were effective only in increasing post-dialytic CP,bic. The simple step-increase profile was effective in linearizing the intradialytic increase in bicarbonate and decreasing the amount of time spent by patients at high plasma CP,bic.
Conclusions
The interventions were effective mostly in patients who needed to lower their pre-dialytic CP,bic. The resistance of the system to increasing pre-dialytic CP,bic in other patients might be caused by modifications of breathing or in hydrogen generation that were not accounted for by our model.
{"title":"Profiled delivery of bicarbonate during weekly cycle of hemodialysis","authors":"Mauro Pietribiasi , John K. Leypoldt , Monika Wieliczko , Malgorzata Twardowska-Kawalec , Malgorzata Debowska , Jolanta Malyszko , Jacek Waniewski","doi":"10.1016/j.bbe.2024.10.002","DOIUrl":"10.1016/j.bbe.2024.10.002","url":null,"abstract":"<div><h3>Background</h3><div>Delivery of bicarbonate during hemodialysis (HD) is aimed at correcting metabolic acidosis in end-stage renal disease patients. We tested modified prescriptions of bicarbonate concentration in dialysis fluid (C<sub>D,bic</sub>), aimed to achieve an optimal pre-dialytic bicarbonate plasma concentration (C<sub>P,bic</sub>).</div></div><div><h3>Methods</h3><div>We used a mathematical model to prescribe individualized HD treatments consisting of 1) adjustment of C<sub>D,bic</sub> to get the pre-dialytic C<sub>P,bic</sub> in a prescribed range, 2) increase of bicarbonate load before the long interdialytic break, and 3) a single step of increase in C<sub>D,bic</sub> after two hours. The outcomes were tested in 24 stable HD patients, monitored during a week of standard HD (Test Week) and a week of modified treatment (Intervention Week).</div></div><div><h3>Results</h3><div>The response to the model-based prescription was different whether the average C<sub>D,bic</sub> during the Intervention Week was higher or lower than the constant value used for the Test Week. For patients with lower average C<sub>D,bic</sub> during the Intervention Week, a significant fraction achieved the target (22 ≤ C<sub>P,bic</sub> ≤ 24 mEq/L). In the group with higher average C<sub>D,bic</sub>, the interventions were effective only in increasing post-dialytic C<sub>P,bic</sub>. The simple step-increase profile was effective in linearizing the intradialytic increase in bicarbonate and decreasing the amount of time spent by patients at high plasma C<sub>P,bic</sub>.</div></div><div><h3>Conclusions</h3><div>The interventions were effective mostly in patients who needed to lower their pre-dialytic CP<sub>,bic</sub>. The resistance of the system to increasing pre-dialytic C<sub>P,bic</sub> in other patients might be caused by modifications of breathing or in hydrogen generation that were not accounted for by our model.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"44 4","pages":"Pages 836-843"},"PeriodicalIF":5.3,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142662769","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-10-01DOI: 10.1016/j.bbe.2024.10.001
Matheus B. Rocha , Flavio P. Loss , Pedro H. da Cunha , Madson Poltronieri Zanoni , Leandro M. de Lima , Isadora Tavares Nascimento , Isabella Rezende , Tania R.P. Canuto , Luciana de Paula Vieira , Renan Rossoni , Maria C.S. Santos , Patricia Lyra Frasson , Wanderson Romão , Paulo R. Filgueiras , Renato A. Krohling
Skin lesions are classified in benign or malignant. Among the malignant, melanoma is a very aggressive cancer and the major cause of deaths. So, early diagnosis of skin cancer is very desired. In the last few years, there is a growing interest in computer aided diagnostic (CAD) of skin lesions. Near-Infrared (NIR) spectroscopy may provide an alternative source of information to automated CAD of skin lesions to be used with the modern techniques of machine learning and deep learning (MDL). One of the main limitations to apply MDL to spectroscopy is the lack of public datasets. Since there is no public dataset of NIR spectral data to skin lesions, as far as we know, an effort has been made and a new dataset named NIR-SC-UFES, has been collected, annotated and analyzed generating the gold-standard for classification of NIR spectral data to skin cancer. Next, the machine learning algorithms XGBoost, CatBoost, LightGBM, 1D-convolutional neural network (1D-CNN) and standard algorithms as SVM and PLS-DA were investigated to classify cancer and non-cancer skin lesions. Experimental results indicate that the best performance was obtained by LightGBM with pre-processing using standard normal variate (SNV), feature extraction and data augmentation with Generative Adversarial Networks (GAN) providing values of 0.839 for balanced accuracy, 0.851 for recall, 0.852 for precision, and 0.850 for F-score. The obtained results indicate the first steps in CAD of skin lesions aiming the automated triage of patients with skin lesions in vivo using NIR spectral data.
{"title":"Skin cancer diagnosis using NIR spectroscopy data of skin lesions in vivo using machine learning algorithms","authors":"Matheus B. Rocha , Flavio P. Loss , Pedro H. da Cunha , Madson Poltronieri Zanoni , Leandro M. de Lima , Isadora Tavares Nascimento , Isabella Rezende , Tania R.P. Canuto , Luciana de Paula Vieira , Renan Rossoni , Maria C.S. Santos , Patricia Lyra Frasson , Wanderson Romão , Paulo R. Filgueiras , Renato A. Krohling","doi":"10.1016/j.bbe.2024.10.001","DOIUrl":"10.1016/j.bbe.2024.10.001","url":null,"abstract":"<div><div>Skin lesions are classified in benign or malignant. Among the malignant, melanoma is a very aggressive cancer and the major cause of deaths. So, early diagnosis of skin cancer is very desired. In the last few years, there is a growing interest in computer aided diagnostic (CAD) of skin lesions. Near-Infrared (NIR) spectroscopy may provide an alternative source of information to automated CAD of skin lesions to be used with the modern techniques of machine learning and deep learning (MDL). One of the main limitations to apply MDL to spectroscopy is the lack of public datasets. Since there is no public dataset of NIR spectral data to skin lesions, as far as we know, an effort has been made and a new dataset named NIR-SC-UFES, has been collected, annotated and analyzed generating the gold-standard for classification of NIR spectral data to skin cancer. Next, the machine learning algorithms XGBoost, CatBoost, LightGBM, 1D-convolutional neural network (1D-CNN) and standard algorithms as SVM and PLS-DA were investigated to classify cancer and non-cancer skin lesions. Experimental results indicate that the best performance was obtained by LightGBM with pre-processing using standard normal variate (SNV), feature extraction and data augmentation with Generative Adversarial Networks (GAN) providing values of 0.839 for balanced accuracy, 0.851 for recall, 0.852 for precision, and 0.850 for F-score. The obtained results indicate the first steps in CAD of skin lesions aiming the automated triage of patients with skin lesions <em>in vivo</em> using NIR spectral data.</div></div>","PeriodicalId":55381,"journal":{"name":"Biocybernetics and Biomedical Engineering","volume":"44 4","pages":"Pages 824-835"},"PeriodicalIF":5.3,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142535971","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号