Transcranial focused ultrasound (tFUS) is an emerging non-invasive therapeutic technology that offers new brain stimulation modality. Precise localization of the acoustic focus to the desired brain target throughout the procedure is needed to ensure the safety and effectiveness of the treatment, but acoustic distortion caused by the skull poses a challenge. Although computational methods can provide the estimated location and shape of the focus, the computation has not reached sufficient speed for real-time inference, which is demanded in real-world clinical situations. Leveraging the advantages of deep learning, we propose multi-modal networks capable of generating intracranial pressure map in real-time.
Methods:
The dataset consisted of free-field pressure maps, intracranial pressure maps, medical images, and transducer placements was obtained from 11 human subjects. The free-field and intracranial pressure maps were computed using the k-space method. We developed network models based on convolutional neural networks and the Swin Transformer, featuring a multi-modal encoder and a decoder.
Results:
Evaluations on foreseen data achieved high focal volume conformity of approximately 93% for both computed tomography (CT) and magnetic resonance (MR) data. For unforeseen data, the networks achieved the focal volume conformity of 88% for CT and 82% for MR. The inference time of the proposed networks was under 0.02 s, indicating the feasibility for real-time simulation.
Conclusions:
The results indicate that our networks can effectively and precisely perform real-time simulation of the intracranial pressure map during tFUS applications. Our work will enhance the safety and accuracy of treatments, representing significant progress for low-intensity focused ultrasound (LIFU) therapies.
{"title":"Multi-modal networks for real-time monitoring of intracranial acoustic field during transcranial focused ultrasound therapy","authors":"Minjee Seo , Minwoo Shin , Gunwoo Noh , Seung-Schik Yoo , Kyungho Yoon","doi":"10.1016/j.cmpb.2024.108458","DOIUrl":"10.1016/j.cmpb.2024.108458","url":null,"abstract":"<div><h3>Background and objective:</h3><div>Transcranial focused ultrasound (tFUS) is an emerging non-invasive therapeutic technology that offers new brain stimulation modality. Precise localization of the acoustic focus to the desired brain target throughout the procedure is needed to ensure the safety and effectiveness of the treatment, but acoustic distortion caused by the skull poses a challenge. Although computational methods can provide the estimated location and shape of the focus, the computation has not reached sufficient speed for real-time inference, which is demanded in real-world clinical situations. Leveraging the advantages of deep learning, we propose multi-modal networks capable of generating intracranial pressure map in real-time.</div></div><div><h3>Methods:</h3><div>The dataset consisted of free-field pressure maps, intracranial pressure maps, medical images, and transducer placements was obtained from 11 human subjects. The free-field and intracranial pressure maps were computed using the <em>k</em>-space method. We developed network models based on convolutional neural networks and the Swin Transformer, featuring a multi-modal encoder and a decoder.</div></div><div><h3>Results:</h3><div>Evaluations on foreseen data achieved high focal volume conformity of approximately 93% for both computed tomography (CT) and magnetic resonance (MR) data. For unforeseen data, the networks achieved the focal volume conformity of 88% for CT and 82% for MR. The inference time of the proposed networks was under 0.02 s, indicating the feasibility for real-time simulation.</div></div><div><h3>Conclusions:</h3><div>The results indicate that our networks can effectively and precisely perform real-time simulation of the intracranial pressure map during tFUS applications. Our work will enhance the safety and accuracy of treatments, representing significant progress for low-intensity focused ultrasound (LIFU) therapies.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108458"},"PeriodicalIF":4.9,"publicationDate":"2024-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142496560","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 : 2024-10-11DOI: 10.1016/j.cmpb.2024.108453
C. Rousseau , Q.L. Vuong , Y. Gossuin , B. Maes , G. Rosolen
Background and Objective
: Theranostics is the combination of the diagnostic and therapeutic phases. Here we focus on simultaneous use of photothermal therapy and magnetic resonance imaging, employing a contrast-photothermal agent that converts incident light into heat and affects the transverse relaxation time, a key magnetic resonance imaging parameter. Our work considers a gold–magnetite nanoshell platform to gauge the feasibility of magnetic resonance imaging monitoring of the heating associated with phototherapy, by studying the modification of the transverse relaxation rate induced by laser illumination of a solution containing these hybrid nanoparticles.
Methods:
We simulate a system composed of an aqueous solution with hybrid nanoshells under continuous laser irradiation, enabling the evaluation of spatial variations of the transverse relaxation rate within the sample. We work with the hybrid nanoshell platform comprising a metal/gold shell for thermoplasmonic effects and a magnetite core for magnetic resonance imaging contrast enhancement. The optical properties of the nanoshells are first obtained through simulations using the finite element method. Next, the heating generated by the laser illumination is calculated by numerical integration. Finally, the transverse relaxation rate is obtained through the application of an analytical model. Additionally, we conduct an optimization of the nanoshell geometry to fulfill requirements of both magnetic resonance imaging and phototherapy techniques.
Results:
Our findings demonstrate a narrow range of nanoshell sizes exhibiting both a plasmonic absorption peak in the human biological window and a high response to laser illumination of the transverse relaxation rate. Furthermore, the illumination can induce up to a 30% modification in transverse relaxation rate compared to the non-illuminated scenario in this range of nanoshell sizes.
Conclusions:
In this work we establish the numerical understanding of the interplay between phototherapy and nuclear magnetic resonance imaging when employed concurrently. This allows magnetic resonance imaging monitoring of the heating associated with phototherapy.
{"title":"Concurrent photothermal therapy and nuclear magnetic resonance imaging with plasmonic–magnetic nanoparticles: A numerical study","authors":"C. Rousseau , Q.L. Vuong , Y. Gossuin , B. Maes , G. Rosolen","doi":"10.1016/j.cmpb.2024.108453","DOIUrl":"10.1016/j.cmpb.2024.108453","url":null,"abstract":"<div><h3>Background and Objective</h3><div>: Theranostics is the combination of the diagnostic and therapeutic phases. Here we focus on simultaneous use of photothermal therapy and magnetic resonance imaging, employing a contrast-photothermal agent that converts incident light into heat and affects the transverse relaxation time, a key magnetic resonance imaging parameter. Our work considers a gold–magnetite nanoshell platform to gauge the feasibility of magnetic resonance imaging monitoring of the heating associated with phototherapy, by studying the modification of the transverse relaxation rate induced by laser illumination of a solution containing these hybrid nanoparticles.</div></div><div><h3>Methods:</h3><div>We simulate a system composed of an aqueous solution with hybrid nanoshells under continuous laser irradiation, enabling the evaluation of spatial variations of the transverse relaxation rate within the sample. We work with the hybrid nanoshell platform comprising a metal/gold shell for thermoplasmonic effects and a magnetite core for magnetic resonance imaging contrast enhancement. The optical properties of the nanoshells are first obtained through simulations using the finite element method. Next, the heating generated by the laser illumination is calculated by numerical integration. Finally, the transverse relaxation rate is obtained through the application of an analytical model. Additionally, we conduct an optimization of the nanoshell geometry to fulfill requirements of both magnetic resonance imaging and phototherapy techniques.</div></div><div><h3>Results:</h3><div>Our findings demonstrate a narrow range of nanoshell sizes exhibiting both a plasmonic absorption peak in the human biological window and a high response to laser illumination of the transverse relaxation rate. Furthermore, the illumination can induce up to a 30% modification in transverse relaxation rate compared to the non-illuminated scenario in this range of nanoshell sizes.</div></div><div><h3>Conclusions:</h3><div>In this work we establish the numerical understanding of the interplay between phototherapy and nuclear magnetic resonance imaging when employed concurrently. This allows magnetic resonance imaging monitoring of the heating associated with phototherapy.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108453"},"PeriodicalIF":4.9,"publicationDate":"2024-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142459887","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 : 2024-10-11DOI: 10.1016/j.cmpb.2024.108455
Hardik Telangore , Victor Azad , Manish Sharma , Ankit Bhurane , Ru San Tan , U. Rajendra Acharya
Background and Objective:
Sudden cardiac death (SCD) is a critical health issue characterized by the sudden failure of heart function, often caused by ventricular fibrillation (VF). Early prediction of SCD is crucial to enable timely interventions. However, current methods predict SCD only a few minutes before its onset, limiting intervention time. This study aims to develop a deep learning-based model for the early prediction of SCD using electrocardiography (ECG) signals.
Methods:
A multimodal explainable deep learning-based model is developed to analyze ECG signals at discrete intervals ranging from 5 to 30 min before SCD onset. The raw ECG signals, 2D scalograms generated through wavelet transform and 2D Hilbert spectrum generated through Hilbert–Huang transform (HHT) of ECG signals were applied to multiple deep learning algorithms. For raw ECG, a combination of 1D-convolutional neural networks (1D-CNN) and long short-term memory networks were employed for feature extraction and temporal pattern recognition. Besides, to extract and analyze features from scalograms and Hilbert spectra, Vision Transformer (ViT) and 2D-CNN have been used.
Results:
The developed model achieved high performance, with accuracy, precision, recall and F1-score of 98.81%, 98.83%, 98.81%, and 98.81% respectively to predict SCD onset 30 min in advance. Further, the proposed model can accurately classify SCD patients and normal controls with 100% accuracy. Thus, the proposed method outperforms the existing state-of-the-art methods.
Conclusions:
The developed model is capable of capturing diverse patterns on ECG signals recorded at multiple discrete time intervals (at 5-minute increments from 5 min to 30 min) prior to SCD onset that could discriminate for SCD. The proposed model significantly improves early SCD prediction, providing a valuable tool for continuous ECG monitoring in high-risk patients.
{"title":"Early prediction of sudden cardiac death using multimodal fusion of ECG Features extracted from Hilbert–Huang and wavelet transforms with explainable vision transformer and CNN models","authors":"Hardik Telangore , Victor Azad , Manish Sharma , Ankit Bhurane , Ru San Tan , U. Rajendra Acharya","doi":"10.1016/j.cmpb.2024.108455","DOIUrl":"10.1016/j.cmpb.2024.108455","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>Sudden cardiac death (SCD) is a critical health issue characterized by the sudden failure of heart function, often caused by ventricular fibrillation (VF). Early prediction of SCD is crucial to enable timely interventions. However, current methods predict SCD only a few minutes before its onset, limiting intervention time. This study aims to develop a deep learning-based model for the early prediction of SCD using electrocardiography (ECG) signals.</div></div><div><h3>Methods:</h3><div>A multimodal explainable deep learning-based model is developed to analyze ECG signals at discrete intervals ranging from 5 to 30 min before SCD onset. The raw ECG signals, 2D scalograms generated through wavelet transform and 2D Hilbert spectrum generated through Hilbert–Huang transform (HHT) of ECG signals were applied to multiple deep learning algorithms. For raw ECG, a combination of 1D-convolutional neural networks (1D-CNN) and long short-term memory networks were employed for feature extraction and temporal pattern recognition. Besides, to extract and analyze features from scalograms and Hilbert spectra, Vision Transformer (ViT) and 2D-CNN have been used.</div></div><div><h3>Results:</h3><div>The developed model achieved high performance, with accuracy, precision, recall and F1-score of 98.81%, 98.83%, 98.81%, and 98.81% respectively to predict SCD onset 30 min in advance. Further, the proposed model can accurately classify SCD patients and normal controls with 100% accuracy. Thus, the proposed method outperforms the existing state-of-the-art methods.</div></div><div><h3>Conclusions:</h3><div>The developed model is capable of capturing diverse patterns on ECG signals recorded at multiple discrete time intervals (at 5-minute increments from 5 min to 30 min) prior to SCD onset that could discriminate for SCD. The proposed model significantly improves early SCD prediction, providing a valuable tool for continuous ECG monitoring in high-risk patients.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108455"},"PeriodicalIF":4.9,"publicationDate":"2024-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142496558","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}
Scaffolds designed for tissue engineering must consider multiple parameters, namely the permeability of the design and the wall shear stress experienced by the cells on the scaffold surface. However, these parameters are not independent from each other, with changes that improve wall shear stress, negatively impacting permeability and vice versa. This study introduces a novel multi-objective optimization framework using Direct MultiSearch (DMS) to design triply periodic minimal surface (TPMS) scaffolds for bone tissue engineering.
Method
The optimization algorithm focused on maximizing the permeability of the scaffolds and obtaining a desired value of average wall shear stress (which ranges between the values that promote osteogenic differentiation of 0.1 mPa and 10 mPa). Multiple fluid inlet velocities and target wall shear stress were analyzed. The DMS method successfully generated Pareto fronts for each configuration, enabling the selection of optimized scaffolds based on specific structural requirements.
Results
The findings reveal that increasing the target wall shear stress results in a greater number of non-dominated points on the Pareto front, highlighting a more robust optimization process. Additionally, it was also demonstrated that the tested Schwartz diamond scaffolds had a better permeability-wall shear stress relation when compared to Schoen gyroid geometries.
Conclusions
Direct MultiSearch was proven as an effective tool to aid in the design of tissue engineering scaffolds. This adaptable optimization framework has potential applications beyond bone tissue engineering, including cartilage tissue differentiation.
{"title":"Direct MultiSearch optimization of TPMS scaffolds for bone tissue engineering","authors":"T.H.V. Pires , J.F.A. Madeira , A.P.G. Castro , P.R. Fernandes","doi":"10.1016/j.cmpb.2024.108461","DOIUrl":"10.1016/j.cmpb.2024.108461","url":null,"abstract":"<div><h3>Background</h3><div>Scaffolds designed for tissue engineering must consider multiple parameters, namely the permeability of the design and the wall shear stress experienced by the cells on the scaffold surface. However, these parameters are not independent from each other, with changes that improve wall shear stress, negatively impacting permeability and vice versa. This study introduces a novel multi-objective optimization framework using Direct MultiSearch (DMS) to design triply periodic minimal surface (TPMS) scaffolds for bone tissue engineering.</div></div><div><h3>Method</h3><div>The optimization algorithm focused on maximizing the permeability of the scaffolds and obtaining a desired value of average wall shear stress (which ranges between the values that promote osteogenic differentiation of 0.1 mPa and 10 mPa). Multiple fluid inlet velocities and target wall shear stress were analyzed. The DMS method successfully generated Pareto fronts for each configuration, enabling the selection of optimized scaffolds based on specific structural requirements.</div></div><div><h3>Results</h3><div>The findings reveal that increasing the target wall shear stress results in a greater number of non-dominated points on the Pareto front, highlighting a more robust optimization process. Additionally, it was also demonstrated that the tested Schwartz diamond scaffolds had a better permeability-wall shear stress relation when compared to Schoen gyroid geometries.</div></div><div><h3>Conclusions</h3><div>Direct MultiSearch was proven as an effective tool to aid in the design of tissue engineering scaffolds. This adaptable optimization framework has potential applications beyond bone tissue engineering, including cartilage tissue differentiation.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108461"},"PeriodicalIF":4.9,"publicationDate":"2024-10-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142441512","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 : 2024-10-10DOI: 10.1016/j.cmpb.2024.108460
Georgia D. Liapi , Christos P. Loizou , Constantinos S. Pattichis , Marios S. Pattichis , Andrew N. Nicolaides , Maura Griffin , Efthyvoulos Kyriacou
<div><h3>Background and objective</h3><div>Carotid B-mode ultrasound (CBUS) imaging is often used to detect and assess atherosclerotic plaques. Doctors often need to segment plaques in the CBUS images to further examine them. Multiple studies have proposed two-dimensional CBUS plaque segmentation deep learning (DL)-based solutions, achieving promising results. In most of these studies, image standardization is not reported, while not all plaque types are represented. However, prior multiple studies have highlighted the importance of data standardization in computerized CBUS plaque classification or segmentation solutions. In this study, we propose and separately evaluate three progressive preprocessing schemes, to discover the most optimal to standardize CBUS images for DL-based carotid plaque segmentation, while we also assess the effect of each preprocessing in the segmentation performance per echodensity-based plaque type (I, II, III, IV and V).</div></div><div><h3>Methods</h3><div>We included three CBUS image datasets (276 CBUS images, from three medical centres), with which we produced 3 data folds (with the best possible equal inclusion of images from all centers per fold), to perform 3-fold cross validation-based training and evaluation of the pre-released Channel-wise Feature Pyramid Network for Medicine (CFPNet-M) model, in carotid plaque type segmentation. We included the three data folds in their original version (O), generating also three preprocessed versions of them, namely, the resolution-normalized (R), the resolution- and intensity-normalized (RN), and the resolution- and intensity-normalized combined with despeckling (RND) versions. The samples were cropped to the plaque level, and the intersection over union (IoU) and the Dice Similarity Coefficient (DSC), along with other metrics, were used to measure the model's performance. In each training round, 12 % of the images in the 2 training folds was used for internal validation (last fold was used in evaluation). Two experienced ultrasonographers manually delineated plaques in the dataset, to provide us with ground truths, while the plaque types (I to V) were extracted according to the Gray-Weale and Geroulakos classification system. We measured the mean±standard deviation of DSC within and across the three evaluated folds, per preprocessing scheme and per plaque type.</div></div><div><h3>Results</h3><div>CFPNet-M segmented the plaques in the CBUS images in all the data preprocessing versions, yielding progressively improved performances (mean DSC at 81.9 ± 9.1 %, 83.6 ± 9.0 %, 84.1 ± 8.3 %, and 84.4 ± 8.1 % for the O, R, RN and RND 3-fold cross validation processes, respectively), irrespective of the plaque type. Interestingly, CFPNet_M yielded improved performances, for all plaque types (I, II, III, IV and V), when trained and tested with the RND data versus the O version, achieving an 80.6 ± 11 % versus 77.6 ± 17 % DSC for type I, an 84.3 ± 8 % versus 81.2 ± 9 % DSC for type II
{"title":"Assessing the impact of ultrasound image standardization in deep learning-based segmentation of carotid plaque types","authors":"Georgia D. Liapi , Christos P. Loizou , Constantinos S. Pattichis , Marios S. Pattichis , Andrew N. Nicolaides , Maura Griffin , Efthyvoulos Kyriacou","doi":"10.1016/j.cmpb.2024.108460","DOIUrl":"10.1016/j.cmpb.2024.108460","url":null,"abstract":"<div><h3>Background and objective</h3><div>Carotid B-mode ultrasound (CBUS) imaging is often used to detect and assess atherosclerotic plaques. Doctors often need to segment plaques in the CBUS images to further examine them. Multiple studies have proposed two-dimensional CBUS plaque segmentation deep learning (DL)-based solutions, achieving promising results. In most of these studies, image standardization is not reported, while not all plaque types are represented. However, prior multiple studies have highlighted the importance of data standardization in computerized CBUS plaque classification or segmentation solutions. In this study, we propose and separately evaluate three progressive preprocessing schemes, to discover the most optimal to standardize CBUS images for DL-based carotid plaque segmentation, while we also assess the effect of each preprocessing in the segmentation performance per echodensity-based plaque type (I, II, III, IV and V).</div></div><div><h3>Methods</h3><div>We included three CBUS image datasets (276 CBUS images, from three medical centres), with which we produced 3 data folds (with the best possible equal inclusion of images from all centers per fold), to perform 3-fold cross validation-based training and evaluation of the pre-released Channel-wise Feature Pyramid Network for Medicine (CFPNet-M) model, in carotid plaque type segmentation. We included the three data folds in their original version (O), generating also three preprocessed versions of them, namely, the resolution-normalized (R), the resolution- and intensity-normalized (RN), and the resolution- and intensity-normalized combined with despeckling (RND) versions. The samples were cropped to the plaque level, and the intersection over union (IoU) and the Dice Similarity Coefficient (DSC), along with other metrics, were used to measure the model's performance. In each training round, 12 % of the images in the 2 training folds was used for internal validation (last fold was used in evaluation). Two experienced ultrasonographers manually delineated plaques in the dataset, to provide us with ground truths, while the plaque types (I to V) were extracted according to the Gray-Weale and Geroulakos classification system. We measured the mean±standard deviation of DSC within and across the three evaluated folds, per preprocessing scheme and per plaque type.</div></div><div><h3>Results</h3><div>CFPNet-M segmented the plaques in the CBUS images in all the data preprocessing versions, yielding progressively improved performances (mean DSC at 81.9 ± 9.1 %, 83.6 ± 9.0 %, 84.1 ± 8.3 %, and 84.4 ± 8.1 % for the O, R, RN and RND 3-fold cross validation processes, respectively), irrespective of the plaque type. Interestingly, CFPNet_M yielded improved performances, for all plaque types (I, II, III, IV and V), when trained and tested with the RND data versus the O version, achieving an 80.6 ± 11 % versus 77.6 ± 17 % DSC for type I, an 84.3 ± 8 % versus 81.2 ± 9 % DSC for type II","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108460"},"PeriodicalIF":4.9,"publicationDate":"2024-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142459886","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 : 2024-10-10DOI: 10.1016/j.cmpb.2024.108457
Peishuo Wu, Chi Zhu
Background and Objective:
Incorporating tissue support in fluid–structure interaction analysis of cardiovascular flows is crucial for accurately representing physiological constraints, achieving realistic vessel wall motion, and minimizing artificial oscillations. The generalized Robin boundary condition, which models tissue support with a spring-damper-type force, uses elastic and damping parameters to represent the viscoelastic behavior of perivascular tissues. Using spatially distributed parameters for tissue support, rather than uniform ones, is more realistic and aligns with the varying properties of vessel walls. However, considering the spatial distribution of both can increase the complexity of preprocessing and numerical implementation. In this work, we develop an effective property method for efficient modeling of non-uniform tissue support and quantifying the contribution of tissue support to the mechanical behaviors of vessel walls.
Methods:
Based on the theory of linear viscoelasticity, we derive the mathematical formulas for the effective property method, integrating the parameters of generalized Robin boundary condition into vessel wall properties. The pulse wave velocity incorporating the influence of tissue support is also analyzed. Furthermore, we modify the coupled momentum method, originally formulated for elastic problems, to account for the viscoelastic properties of the vessel wall.
Results:
The method is verified with three-dimensional fluid–structure interaction simulations, achieving a maximum relative error of less than 2.2% for flow rate and less than 0.7% for pressure. This method shows that tissue support parameters can be integrated into vessel wall properties, resulting in increased apparent wall stiffness and viscosity, and further changing pressure, flow rate, and wave propagation.
Conclusion:
In this study, we develop an effective property method for quantitatively assessing the impact of tissue support and for efficiently modeling non-uniform tissue support. Moreover, this method offers further insights into clinically measured pulse wave velocity, demonstrating that it reflects the combined influence of both vessel wall properties and tissue support.
{"title":"Effective property method for efficient modeling of non-uniform tissue support in fluid–structure interaction simulation of blood flows","authors":"Peishuo Wu, Chi Zhu","doi":"10.1016/j.cmpb.2024.108457","DOIUrl":"10.1016/j.cmpb.2024.108457","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>Incorporating tissue support in fluid–structure interaction analysis of cardiovascular flows is crucial for accurately representing physiological constraints, achieving realistic vessel wall motion, and minimizing artificial oscillations. The generalized Robin boundary condition, which models tissue support with a spring-damper-type force, uses elastic and damping parameters to represent the viscoelastic behavior of perivascular tissues. Using spatially distributed parameters for tissue support, rather than uniform ones, is more realistic and aligns with the varying properties of vessel walls. However, considering the spatial distribution of both can increase the complexity of preprocessing and numerical implementation. In this work, we develop an effective property method for efficient modeling of non-uniform tissue support and quantifying the contribution of tissue support to the mechanical behaviors of vessel walls.</div></div><div><h3>Methods:</h3><div>Based on the theory of linear viscoelasticity, we derive the mathematical formulas for the effective property method, integrating the parameters of generalized Robin boundary condition into vessel wall properties. The pulse wave velocity incorporating the influence of tissue support is also analyzed. Furthermore, we modify the coupled momentum method, originally formulated for elastic problems, to account for the viscoelastic properties of the vessel wall.</div></div><div><h3>Results:</h3><div>The method is verified with three-dimensional fluid–structure interaction simulations, achieving a maximum relative error of less than 2.2% for flow rate and less than 0.7% for pressure. This method shows that tissue support parameters can be integrated into vessel wall properties, resulting in increased apparent wall stiffness and viscosity, and further changing pressure, flow rate, and wave propagation.</div></div><div><h3>Conclusion:</h3><div>In this study, we develop an effective property method for quantitatively assessing the impact of tissue support and for efficiently modeling non-uniform tissue support. Moreover, this method offers further insights into clinically measured pulse wave velocity, demonstrating that it reflects the combined influence of both vessel wall properties and tissue support.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108457"},"PeriodicalIF":4.9,"publicationDate":"2024-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142434144","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 : 2024-10-10DOI: 10.1016/j.cmpb.2024.108459
Ankush Manocha , Munish Bhatia , Gulshan Kumar
Background and Objective:
In the realm of smart healthcare, precise monitoring and prediction services are crucial for mitigating the impact of infectious diseases. This study introduces an innovative digital twin technology-inspired monitoring architecture that employs a similarity-based hybrid modeling scheme to significantly enhance accuracy in the smart healthcare domain. The research also delves into the potential of IoT technology in delivering advanced technological healthcare solutions, with a specific focus on the rapid expansion of dengue fever.
Methods:
The proposed digital twin-inspired healthcare system is designed to proactively combat the spread of dengue virus by enabling ubiquitous monitoring and forecasting of individuals’ susceptibility to dengue infection. The system utilizes digital twin technology to observe the status of healthcare and generate likely predictions about the vulnerability to the virus by employing k-means Clustering and Artificial Neural Networks.
Results:
The proposed system has been validated and its effectiveness has been demonstrated through experimental evaluation using carefully defined methods. The results of the experimental assessment confirm that the system performs optimally in terms of Temporal Delay (14.15 s), Classification Accuracy (92.86%), Sensitivity (92.43%), Specificity (91.52%), F-measure (90.86%), and Prediction Effectiveness. Moreover, by integrating a hybrid model that corrects errors in physics-based predictions employing a model for error correction driven by data, this approach has demonstrated a noteworthy 48% reduction in prediction errors, particularly in health monitoring scenarios.
Conclusions:
The digital twin-inspired healthcare system proposed in this study can assist healthcare providers in assessing the health vulnerability of the dengue virus, thereby reducing the likelihood of long-term or catastrophic health consequences. The integration of a hybrid modeling approach and the utilization of IoT technology has shown promising results in enhancing the accuracy and effectiveness of smart health monitoring and prediction services.
{"title":"Smart monitoring solution for dengue infection control: A digital twin-inspired approach","authors":"Ankush Manocha , Munish Bhatia , Gulshan Kumar","doi":"10.1016/j.cmpb.2024.108459","DOIUrl":"10.1016/j.cmpb.2024.108459","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>In the realm of smart healthcare, precise monitoring and prediction services are crucial for mitigating the impact of infectious diseases. This study introduces an innovative digital twin technology-inspired monitoring architecture that employs a similarity-based hybrid modeling scheme to significantly enhance accuracy in the smart healthcare domain. The research also delves into the potential of IoT technology in delivering advanced technological healthcare solutions, with a specific focus on the rapid expansion of dengue fever.</div></div><div><h3>Methods:</h3><div>The proposed digital twin-inspired healthcare system is designed to proactively combat the spread of dengue virus by enabling ubiquitous monitoring and forecasting of individuals’ susceptibility to dengue infection. The system utilizes digital twin technology to observe the status of healthcare and generate likely predictions about the vulnerability to the virus by employing k-means Clustering and Artificial Neural Networks.</div></div><div><h3>Results:</h3><div>The proposed system has been validated and its effectiveness has been demonstrated through experimental evaluation using carefully defined methods. The results of the experimental assessment confirm that the system performs optimally in terms of Temporal Delay (14.15 s), Classification Accuracy (92.86%), Sensitivity (92.43%), Specificity (91.52%), F-measure (90.86%), and Prediction Effectiveness. Moreover, by integrating a hybrid model that corrects errors in physics-based predictions employing a model for error correction driven by data, this approach has demonstrated a noteworthy 48% reduction in prediction errors, particularly in health monitoring scenarios.</div></div><div><h3>Conclusions:</h3><div>The digital twin-inspired healthcare system proposed in this study can assist healthcare providers in assessing the health vulnerability of the dengue virus, thereby reducing the likelihood of long-term or catastrophic health consequences. The integration of a hybrid modeling approach and the utilization of IoT technology has shown promising results in enhancing the accuracy and effectiveness of smart health monitoring and prediction services.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108459"},"PeriodicalIF":4.9,"publicationDate":"2024-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142459888","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 : 2024-10-09DOI: 10.1016/j.cmpb.2024.108444
Demin Yang, Haochen Shi, Bolun Zeng, Xiaojun Chen
Background and Objectives:
Image-based 2D/3D registration is a crucial technology for fluoroscopy-guided surgical interventions. However, traditional registration methods relying on a single X-ray image into surgical navigation systems. This study proposes a novel 2D/3D registration approach utilizing biplanar X-ray images combined with computed tomography (CT) to significantly reduce registration and navigation errors. The method is successfully implemented in a surgical navigation system, enhancing its precision and reliability.
Methods:
First, we simultaneously register the frontal and lateral X-ray images with the CT image, enabling mutual complementation and more precise localization. Additionally, we introduce a novel similarity measure for image comparison, providing a more robust cost function for the optimization algorithm. Furthermore, a multi-resolution strategy is employed to enhance registration efficiency. Lastly, we propose a more accurate coordinate transformation method, based on projection and 3D reconstruction, to improve the precision of surgical navigation systems.Results: We conducted registration and navigation experiments using pelvic, spinal, and femur phantoms. The navigation results demonstrated that the feature registration errors (FREs) in the three experiments were 0.505±0.063 mm, 0.515±0.055 mm, and 0.577±0.056 mm, respectively. Compared to the point-to-point (PTP) registration method based on anatomical landmarks, our method reduced registration errors by 31.3%, 23.9%, and 26.3%, respectively.
Conclusion:
The results demonstrate that our method significantly reduces registration and navigation errors, highlighting its potential for application across various anatomical sites. Our code is available at: https://github.com/SJTUdemon/2D-3D-Registration
背景和目的:基于图像的二维/三维配准是荧光屏引导手术干预的关键技术。然而,传统的配准方法依赖于手术导航系统中的单一 X 射线图像。本研究提出了一种新型 2D/3D 配准方法,利用双平面 X 光图像与计算机断层扫描(CT)相结合,显著减少配准和导航误差。方法:首先,我们将正面和侧面的 X 光图像与 CT 图像同时配准,从而实现互补和更精确的定位。此外,我们还引入了一种用于图像对比的新型相似度测量方法,为优化算法提供了更稳健的成本函数。此外,我们还采用了多分辨率策略来提高配准效率。最后,我们提出了一种基于投影和三维重建的更精确坐标转换方法,以提高手术导航系统的精度:我们使用骨盆、脊柱和股骨模型进行了配准和导航实验。导航结果表明,三次实验中的特征配准误差(FREs)分别为 0.505±0.063 mm、0.515±0.055 mm 和 0.577±0.056 mm。与基于解剖地标的点对点(PTP)配准方法相比,我们的方法分别减少了 31.3%、23.9% 和 26.3% 的配准误差。我们的代码可在以下网址获取: https://github.com/SJTUdemon/2D-3D-Registration
{"title":"2D/3D registration based on biplanar X-ray and CT images for surgical navigation","authors":"Demin Yang, Haochen Shi, Bolun Zeng, Xiaojun Chen","doi":"10.1016/j.cmpb.2024.108444","DOIUrl":"10.1016/j.cmpb.2024.108444","url":null,"abstract":"<div><h3>Background and Objectives:</h3><div>Image-based 2D/3D registration is a crucial technology for fluoroscopy-guided surgical interventions. However, traditional registration methods relying on a single X-ray image into surgical navigation systems. This study proposes a novel 2D/3D registration approach utilizing biplanar X-ray images combined with computed tomography (CT) to significantly reduce registration and navigation errors. The method is successfully implemented in a surgical navigation system, enhancing its precision and reliability.</div></div><div><h3>Methods:</h3><div>First, we simultaneously register the frontal and lateral X-ray images with the CT image, enabling mutual complementation and more precise localization. Additionally, we introduce a novel similarity measure for image comparison, providing a more robust cost function for the optimization algorithm. Furthermore, a multi-resolution strategy is employed to enhance registration efficiency. Lastly, we propose a more accurate coordinate transformation method, based on projection and 3D reconstruction, to improve the precision of surgical navigation systems.<em>Results:</em> We conducted registration and navigation experiments using pelvic, spinal, and femur phantoms. The navigation results demonstrated that the feature registration errors (FREs) in the three experiments were 0.505±0.063 mm, 0.515±0.055 mm, and 0.577±0.056 mm, respectively. Compared to the point-to-point (PTP) registration method based on anatomical landmarks, our method reduced registration errors by 31.3%, 23.9%, and 26.3%, respectively.</div></div><div><h3>Conclusion:</h3><div>The results demonstrate that our method significantly reduces registration and navigation errors, highlighting its potential for application across various anatomical sites. Our code is available at: <span><span>https://github.com/SJTUdemon/2D-3D-Registration</span><svg><path></path></svg></span></div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108444"},"PeriodicalIF":4.9,"publicationDate":"2024-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142433480","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 : 2024-10-09DOI: 10.1016/j.cmpb.2024.108452
Wenhan Liu , Shurong Pan , Zhoutong Li , Sheng Chang , Qijun Huang , Nan Jiang
Background and Objective:
Electrocardiogram (ECG) is one of the most important diagnostic tools for cardiovascular diseases (CVDs). Recent studies show that deep learning models can be trained using labeled ECGs to achieve automatic detection of CVDs, assisting cardiologists in diagnosis. However, the deep learning models heavily rely on labels in training, while manual labeling is costly and time-consuming. This paper proposes a new self-supervised learning (SSL) method for multilead ECGs: bootstrap each lead’s latent (BELL) to reduce the reliance and boost model performance in various tasks, especially when training data are insufficient.
Method:
BELL is a variant of the well-known bootstrap your own latent (BYOL). The BELL aims to learn prior knowledge from unlabeled ECGs by pretraining, benefitting downstream tasks. It leverages the characteristics of multilead ECGs. First, BELL uses the multiple-branch skeleton, which is more effective in processing multilead ECGs. Moreover, it proposes intra-lead and inter-lead mean square error (MSE) to guide pretraining, and their fusion can result in better performances. Additionally, BELL inherits the main advantage of the BYOL: No negative pair is used in pretraining, making it more efficient.
Results:
In most cases, BELL surpasses previous works in the experiments. More importantly, the pretraining improves model performances by 0.69% 8.89% in downstream tasks when only 10% of training data are available. Furthermore, BELL shows excellent adaptability to uncurated ECG data from a real-world hospital. Only slight performance degradation occurs (1% in most cases) when using these data.
Conclusion:
The results suggest that the BELL can alleviate the reliance on manual ECG labels from cardiologists, a critical bottleneck of the current deep learning-based models. In this way, the BELL can also help deep learning extend its application on automatic ECG analysis, reducing the cardiologists’ burden in real-world diagnosis.
{"title":"Bootstrap each lead’s latent: A novel method for self-supervised learning of multilead electrocardiograms","authors":"Wenhan Liu , Shurong Pan , Zhoutong Li , Sheng Chang , Qijun Huang , Nan Jiang","doi":"10.1016/j.cmpb.2024.108452","DOIUrl":"10.1016/j.cmpb.2024.108452","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>Electrocardiogram (ECG) is one of the most important diagnostic tools for cardiovascular diseases (CVDs). Recent studies show that deep learning models can be trained using labeled ECGs to achieve automatic detection of CVDs, assisting cardiologists in diagnosis. However, the deep learning models heavily rely on labels in training, while manual labeling is costly and time-consuming. This paper proposes a new self-supervised learning (SSL) method for multilead ECGs: bootstrap each lead’s latent (BELL) to reduce the reliance and boost model performance in various tasks, especially when training data are insufficient.</div></div><div><h3>Method:</h3><div>BELL is a variant of the well-known bootstrap your own latent (BYOL). The BELL aims to learn prior knowledge from unlabeled ECGs by pretraining, benefitting downstream tasks. It leverages the characteristics of multilead ECGs. First, BELL uses the multiple-branch skeleton, which is more effective in processing multilead ECGs. Moreover, it proposes intra-lead and inter-lead mean square error (MSE) to guide pretraining, and their fusion can result in better performances. Additionally, BELL inherits the main advantage of the BYOL: No negative pair is used in pretraining, making it more efficient.</div></div><div><h3>Results:</h3><div>In most cases, BELL surpasses previous works in the experiments. More importantly, the pretraining improves model performances by 0.69% <span><math><mo>∼</mo></math></span> 8.89% in downstream tasks when only 10% of training data are available. Furthermore, BELL shows excellent adaptability to uncurated ECG data from a real-world hospital. Only slight performance degradation occurs (<span><math><mo><</mo></math></span>1% in most cases) when using these data.</div></div><div><h3>Conclusion:</h3><div>The results suggest that the BELL can alleviate the reliance on manual ECG labels from cardiologists, a critical bottleneck of the current deep learning-based models. In this way, the BELL can also help deep learning extend its application on automatic ECG analysis, reducing the cardiologists’ burden in real-world diagnosis.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108452"},"PeriodicalIF":4.9,"publicationDate":"2024-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142406219","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 : 2024-10-05DOI: 10.1016/j.cmpb.2024.108451
Keying Qi , Chenchen Yan , Donghao Niu , Bing Zhang , Dong Liang , Xiaojing Long
Background and Objective:
Fetal brain tissue segmentation provides foundational support for comprehensively understanding the neurodevelopment of normal and congenital disease-affected fetuses. Manual labeling is very time-consuming, and automated segmentation methods can greatly improve the efficiency of doctors. At the same time, fetal brain tissue undergoes various changes throughout the pregnancy, leading to a continuous change in tissue contrast, which greatly increases the difficulty of training segmentation methods. This study aims to develop an automated segmentation model that can efficiently and accurately segment fetal brain tissue, improving the workflow for medical professionals.
Methods:
We propose a novel deep learning-based segmentation model that incorporates three innovative components: Firstly, a new Dual Dilated Attention Block (DDAB) is proposed in the encoder part to enhance the feature extraction of local spatial and structural contextual information. Secondly, a Multi-scale Deformable Transformer (MSDT) is integrated into the bottleneck to improve the feature extraction of global information on local spatial and structural contextual information. Thirdly, we use a novel block based on Graph Convolution Attention (GCAB) in the decoder, which effectively enhances the features at the decoder.The code is available at https://github.com/unicoco7/MG-Net/.
Results:
We trained and tested on the FeTA 2021 and FeTA 2022 datasets, and evaluated using seven popular metrics, including Dice, IoU, MAE, BoundaryF, PRE, SEN, and SPE. Compared to the current state-of-the-art 3D segmentation models such as nnFormer, SwinUNETR, and 3DUX-net, our proposed method has surpassed all of them in metrics like Dice, IoU, and MAE. Specifically, on the FeTA 2021 dataset, our model achieved a Dice of 0.8666, an IoU of 0.7646, and an MAE of 0.0027; on the FeTA 2022 dataset, it achieved a Dice of 0.8552, an IoU of 0.7470, and an MAE of 0.0005.
Conclusion:
In this paper, we propose a model for three-dimensional fetal brain tissue segmentation based on multi-scale feature fusion and graph convolution attention mechanism, and conduct experimental evaluation on the FeTA 2021 and FeTA 2022 datasets. Understanding the boundaries of fetal brain tissue is crucial for doctors’ diagnosis, so the proposed model is expected to improve the speed and accuracy of doctors’ diagnoses.
{"title":"MG-Net: A fetal brain tissue segmentation method based on multiscale feature fusion and graph convolution attention mechanisms","authors":"Keying Qi , Chenchen Yan , Donghao Niu , Bing Zhang , Dong Liang , Xiaojing Long","doi":"10.1016/j.cmpb.2024.108451","DOIUrl":"10.1016/j.cmpb.2024.108451","url":null,"abstract":"<div><h3>Background and Objective:</h3><div>Fetal brain tissue segmentation provides foundational support for comprehensively understanding the neurodevelopment of normal and congenital disease-affected fetuses. Manual labeling is very time-consuming, and automated segmentation methods can greatly improve the efficiency of doctors. At the same time, fetal brain tissue undergoes various changes throughout the pregnancy, leading to a continuous change in tissue contrast, which greatly increases the difficulty of training segmentation methods. This study aims to develop an automated segmentation model that can efficiently and accurately segment fetal brain tissue, improving the workflow for medical professionals.</div></div><div><h3>Methods:</h3><div>We propose a novel deep learning-based segmentation model that incorporates three innovative components: Firstly, a new Dual Dilated Attention Block (DDAB) is proposed in the encoder part to enhance the feature extraction of local spatial and structural contextual information. Secondly, a Multi-scale Deformable Transformer (MSDT) is integrated into the bottleneck to improve the feature extraction of global information on local spatial and structural contextual information. Thirdly, we use a novel block based on Graph Convolution Attention (GCAB) in the decoder, which effectively enhances the features at the decoder.The code is available at <span><span>https://github.com/unicoco7/MG-Net/</span><svg><path></path></svg></span>.</div></div><div><h3>Results:</h3><div>We trained and tested on the FeTA 2021 and FeTA 2022 datasets, and evaluated using seven popular metrics, including Dice, IoU, MAE, BoundaryF, PRE, SEN, and SPE. Compared to the current state-of-the-art 3D segmentation models such as nnFormer, SwinUNETR, and 3DUX-net, our proposed method has surpassed all of them in metrics like Dice, IoU, and MAE. Specifically, on the FeTA 2021 dataset, our model achieved a Dice of 0.8666, an IoU of 0.7646, and an MAE of 0.0027; on the FeTA 2022 dataset, it achieved a Dice of 0.8552, an IoU of 0.7470, and an MAE of 0.0005.</div></div><div><h3>Conclusion:</h3><div>In this paper, we propose a model for three-dimensional fetal brain tissue segmentation based on multi-scale feature fusion and graph convolution attention mechanism, and conduct experimental evaluation on the FeTA 2021 and FeTA 2022 datasets. Understanding the boundaries of fetal brain tissue is crucial for doctors’ diagnosis, so the proposed model is expected to improve the speed and accuracy of doctors’ diagnoses.</div></div>","PeriodicalId":10624,"journal":{"name":"Computer methods and programs in biomedicine","volume":"257 ","pages":"Article 108451"},"PeriodicalIF":4.9,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142421047","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}