Deep learning models have the potential to improve the accuracy and speed of medical microwave imaging. However, their performance often suffers due to a lack of high-quality data. Generative models, especially Denoising Diffusion Probabilistic Models (DDPM), solve this problem by creating realistic data for training and validation. These models have been used in various fields like text-to-image generation, time series generation, and EEG signal synthesis. However, they are not yet used in microwave head imaging for signal generation. Generating meaningful signals for stroke detection in microwave head imaging is challenging because the signals must show both the type and location of strokes. In this paper, DDPM is introduced for conditional signal generation in microwave head imaging. Also, different ways to embed the relevant conditions are explored. The generated signals are evaluated using quantitative metrics and the distorted Born iterative method to check their physical plausibility. Our results show that DDPM, with specially designed condition embeddings and noise schedulers, generates realistic signals, offering a new approach to train and validate deep learning models for microwave head imaging.
{"title":"Conditional Synthetic Signal Generation for Microwave Head Imaging Using Diffusion Models","authors":"Wei-chung Lai;Alina Bialkowski;Lei Guo;Konstanty Bialkowski;Amin Abbosh","doi":"10.1109/JERM.2025.3581576","DOIUrl":"https://doi.org/10.1109/JERM.2025.3581576","url":null,"abstract":"Deep learning models have the potential to improve the accuracy and speed of medical microwave imaging. However, their performance often suffers due to a lack of high-quality data. Generative models, especially Denoising Diffusion Probabilistic Models (DDPM), solve this problem by creating realistic data for training and validation. These models have been used in various fields like text-to-image generation, time series generation, and EEG signal synthesis. However, they are not yet used in microwave head imaging for signal generation. Generating meaningful signals for stroke detection in microwave head imaging is challenging because the signals must show both the type and location of strokes. In this paper, DDPM is introduced for conditional signal generation in microwave head imaging. Also, different ways to embed the relevant conditions are explored. The generated signals are evaluated using quantitative metrics and the distorted Born iterative method to check their physical plausibility. Our results show that DDPM, with specially designed condition embeddings and noise schedulers, generates realistic signals, offering a new approach to train and validate deep learning models for microwave head imaging.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"466-477"},"PeriodicalIF":3.2,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560643","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-07-03DOI: 10.1109/JERM.2025.3578677
Zhimeng Xu;Yichun Chen;Dan Li;Liangqin Chen;Yueming Gao;Zhizhang David Chen
Monitoring vital signs is essential for assessing individuals' health status and supporting various medical interventions; however, conventional methods depend on expensive and invasive hospital-based or wearable devices. This article presents a novel approach to contactless heart rate monitoring that leverages an Antenna-on-Package Pulse Coherent Radar (AoP PCR) system. To address the inherently low sampling rates associated with the pulse repetition frequency of the PCR during remote monitoring, a signal enhancement algorithm is presented. This algorithm leverages the quasi-periodic nature of chest displacement signals, leading to significantly improved temporal resolution and enabling reliable heart rate monitoring using a cost-effective PCR system. Furthermore, extracting heartbeat signals faces a significant challenge in optimally tuning the parameters of Variational Modal Decomposition (VMD) due to variations in distance and angle. To tackle this, an enhanced method called VMD based on the Whale Optimization Algorithm with Quasi-Reflection Learning (QRWOA-VMD) has been devised to enhance the precision of parameter optimization in VMD, thereby improving the decomposition accuracy of heartbeat signals across diverse angles and distances, leading to more reliable and robust heartbeat signal extraction. Comprehensive evaluation demonstrates that the proposed method achieves over 97% accuracy in heart rate monitoring under standard conditions, with the radar facing the chest within a 1.5-meter range. Even in challenging scenarios, such as a ±30° azimuth angles and a 20° elevation angle relative to the chest, accuracy remains above 93%.
{"title":"QRWOA-VMD Enhanced Heart Rate Monitoring Using PCR Radar","authors":"Zhimeng Xu;Yichun Chen;Dan Li;Liangqin Chen;Yueming Gao;Zhizhang David Chen","doi":"10.1109/JERM.2025.3578677","DOIUrl":"https://doi.org/10.1109/JERM.2025.3578677","url":null,"abstract":"Monitoring vital signs is essential for assessing individuals' health status and supporting various medical interventions; however, conventional methods depend on expensive and invasive hospital-based or wearable devices. This article presents a novel approach to contactless heart rate monitoring that leverages an Antenna-on-Package Pulse Coherent Radar (AoP PCR) system. To address the inherently low sampling rates associated with the pulse repetition frequency of the PCR during remote monitoring, a signal enhancement algorithm is presented. This algorithm leverages the quasi-periodic nature of chest displacement signals, leading to significantly improved temporal resolution and enabling reliable heart rate monitoring using a cost-effective PCR system. Furthermore, extracting heartbeat signals faces a significant challenge in optimally tuning the parameters of Variational Modal Decomposition (VMD) due to variations in distance and angle. To tackle this, an enhanced method called VMD based on the Whale Optimization Algorithm with Quasi-Reflection Learning (QRWOA-VMD) has been devised to enhance the precision of parameter optimization in VMD, thereby improving the decomposition accuracy of heartbeat signals across diverse angles and distances, leading to more reliable and robust heartbeat signal extraction. Comprehensive evaluation demonstrates that the proposed method achieves over 97% accuracy in heart rate monitoring under standard conditions, with the radar facing the chest within a 1.5-meter range. Even in challenging scenarios, such as a ±30° azimuth angles and a 20° elevation angle relative to the chest, accuracy remains above 93%.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"455-465"},"PeriodicalIF":3.2,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560642","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-16DOI: 10.1109/JERM.2025.3577268
R. Crusi;N. Colistra;F. Camera;G. Monti;M. S. Zappatore;C. Merla;L. Tarricone
In this paper, a novel application of deep learning is proposed, to predict and optimize key parameters in cardiac Pulsed-Field Ablation (PFA) treatments. Building on our extensive experience and on a set of experimental data extracted from scientific literature, we leveraged artificial neuronal networks to accurately predict the ablated area, optimize electrode configurations, and tune various heterogeneous parameters, including electric signal characteristics. Tests performed on experimental data available in the literature demonstrate that deep learning algorithms can effectively predict PFA treatment parameters using both single-target and multi-target networks with comparable performance. The overall accuracy of the predictions confirms the potential of this approach for optimizing PFA treatments. The promising results underscore the power of deep learning in leveraging extensive PFA clinical data and guiding future applications. This approach indeed represents a significant advancement toward developing patient-specific PFA protocols.
{"title":"Cardiac Pulsed-Field Ablation: Deep Learning Solutions for Multi-Parameter Predictions","authors":"R. Crusi;N. Colistra;F. Camera;G. Monti;M. S. Zappatore;C. Merla;L. Tarricone","doi":"10.1109/JERM.2025.3577268","DOIUrl":"https://doi.org/10.1109/JERM.2025.3577268","url":null,"abstract":"In this paper, a novel application of deep learning is proposed, to predict and optimize key parameters in cardiac Pulsed-Field Ablation (PFA) treatments. Building on our extensive experience and on a set of experimental data extracted from scientific literature, we leveraged artificial neuronal networks to accurately predict the ablated area, optimize electrode configurations, and tune various heterogeneous parameters, including electric signal characteristics. Tests performed on experimental data available in the literature demonstrate that deep learning algorithms can effectively predict PFA treatment parameters using both single-target and multi-target networks with comparable performance. The overall accuracy of the predictions confirms the potential of this approach for optimizing PFA treatments. The promising results underscore the power of deep learning in leveraging extensive PFA clinical data and guiding future applications. This approach indeed represents a significant advancement toward developing patient-specific PFA protocols.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"445-454"},"PeriodicalIF":3.2,"publicationDate":"2025-06-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560641","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-13DOI: 10.1109/JERM.2025.3572116
Mauricio Hernández;Hamid Akbari-Chelaresi;Ghazaleh Tashtarian;Omar M. Ramahi
This work introduces an electromagnetic energy scanning technique with specific application to the detection of breast cancer. The technique is based on a metasurface field detector probe composed of an ensemble of electrically small elements resonating at 700 MHz, where the middle single element represents the field detector. The sensor scans the two-dimensional plane that contains the energy transmitted through the breast. Once the scan is completed, we generate a contrast image composed of N X N pixels that represent the different components of the breast tissue. We then compile a dataset of these contrast images to train a convolutional neural network (CNN) to differentiate between healthy and unhealthy breast tissue. Thie probe provides a resolution that cannot be matched by either electrically small probes or resonance-based probes that have dimensions comparable to the wavelength. The field emanating from a specific structure, such as a human female breast, can be scanned by the proposed probe to achieve a resolution in the millimeter range while operating in the low-microwave frequency spectrum. The probe was tested numerically, and a prototype was tested experimentally, demonstrating its effectiveness in providing a field resolution of approximately 5 mm.
本文介绍了一种特殊应用于乳腺癌检测的电磁能量扫描技术。该技术基于一个超表面场探测器探头，该探头由一组谐振频率为700 MHz的电子小元件组成，其中中间的单个元件代表场探测器。传感器扫描包含通过乳房传输能量的二维平面。一旦扫描完成，我们生成一个由N X N像素组成的对比图像，代表乳房组织的不同组成部分。然后，我们编译这些对比图像的数据集来训练卷积神经网络（CNN）来区分健康和不健康的乳房组织。该探头提供的分辨率是电小探头或尺寸与波长相当的基于共振的探头无法比拟的。从特定结构（如人类女性乳房）发出的场可以被提议的探针扫描，在低微波频谱中工作时达到毫米范围的分辨率。对探针进行了数值测试，并对原型进行了实验测试，证明了其在提供约5毫米的场分辨率方面的有效性。
{"title":"Breast Cancer Detection Using a Metasurface-Based Microwave Probe","authors":"Mauricio Hernández;Hamid Akbari-Chelaresi;Ghazaleh Tashtarian;Omar M. Ramahi","doi":"10.1109/JERM.2025.3572116","DOIUrl":"https://doi.org/10.1109/JERM.2025.3572116","url":null,"abstract":"This work introduces an electromagnetic energy scanning technique with specific application to the detection of breast cancer. The technique is based on a metasurface field detector probe composed of an ensemble of electrically small elements resonating at 700 MHz, where the middle single element represents the field detector. The sensor scans the two-dimensional plane that contains the energy transmitted through the breast. Once the scan is completed, we generate a contrast image composed of N X N pixels that represent the different components of the breast tissue. We then compile a dataset of these contrast images to train a convolutional neural network (CNN) to differentiate between healthy and unhealthy breast tissue. Thie probe provides a resolution that cannot be matched by either electrically small probes or resonance-based probes that have dimensions comparable to the wavelength. The field emanating from a specific structure, such as a human female breast, can be scanned by the proposed probe to achieve a resolution in the millimeter range while operating in the low-microwave frequency spectrum. The probe was tested numerically, and a prototype was tested experimentally, demonstrating its effectiveness in providing a field resolution of approximately 5 mm.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"436-444"},"PeriodicalIF":3.2,"publicationDate":"2025-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560761","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-06-12DOI: 10.1109/JERM.2025.3574086
Melany Gutierrez-Hernandez;Sally P. Duarte;Daniel Parrado Triana;Satheesh Bojja-Venkatakrishnan;Jorge Riera Diaz;John L. Volakis
Electrocorticogram (ECoG) is frequently used to identify the origin of seizures. ECoG recording is highly invasive and usually involves wires that remain protruding from the skull. Alternative implantable ECoG systems have been designed. The latter requires a power supply with high power consumption that can generate heat, possibly causing tissue damage. Previously, our group introduced and developed a battery-less, wireless neurosensing system (WiNS). Also, comparable battery-free sensors demonstrated a minimum detectable signal (MDS) of 60 μVpp to 200 μVpp in benchtop measurements using high impedance electrodes ($sim$ 33 k$Omega$). In this paper, we introduce a groundbreaking 3D-printed neural recorder with 50% reduced size over previous designs. Also, the matching circuit and antennas are designed for optimal communications and coupling, improving the MDS up to 10-fold. In addition, a full implantation in a large, regulated animal model (swine) was done for the first time, successfully recording evoked neural potentials. This will create the foundation for evaluating this system in chronic epileptic pigs and humans. Overall, the paper demonstrates the cutting-edge capabilities of our proposed recorder in the field of neurological activity monitoring.
肾上腺皮质电图（ECoG）常用于识别癫痫发作的起源。ECoG记录是高度侵入性的，通常涉及颅骨外伸出的金属丝。替代性植入式ECoG系统已经被设计出来。后者需要一个高功耗的电源，可以产生热量，可能导致组织损伤。此前，我们的团队推出并开发了一种无电池无线神经传感系统（WiNS）。此外，类似的无电池传感器在使用高阻抗电极（$sim$ 33 k $Omega$）的台式测量中显示了60 μVpp至200 μVpp的最小可检测信号（MDS）。在本文中，我们介绍了一种具有开创性的3d打印神经记录仪% reduced size over previous designs. Also, the matching circuit and antennas are designed for optimal communications and coupling, improving the MDS up to 10-fold. In addition, a full implantation in a large, regulated animal model (swine) was done for the first time, successfully recording evoked neural potentials. This will create the foundation for evaluating this system in chronic epileptic pigs and humans. Overall, the paper demonstrates the cutting-edge capabilities of our proposed recorder in the field of neurological activity monitoring.
{"title":"Lightweight, Battery-Less and Wireless Sensor for Monitoring Neuronal Activity in Swine","authors":"Melany Gutierrez-Hernandez;Sally P. Duarte;Daniel Parrado Triana;Satheesh Bojja-Venkatakrishnan;Jorge Riera Diaz;John L. Volakis","doi":"10.1109/JERM.2025.3574086","DOIUrl":"https://doi.org/10.1109/JERM.2025.3574086","url":null,"abstract":"Electrocorticogram (ECoG) is frequently used to identify the origin of seizures. ECoG recording is highly invasive and usually involves wires that remain protruding from the skull. Alternative implantable ECoG systems have been designed. The latter requires a power supply with high power consumption that can generate heat, possibly causing tissue damage. Previously, our group introduced and developed a battery-less, wireless neurosensing system (WiNS). Also, comparable battery-free sensors demonstrated a minimum detectable signal (MDS) of 60 μV<sub>pp</sub> to 200 μV<sub>pp</sub> in benchtop measurements using high impedance electrodes (<inline-formula><tex-math>$sim$</tex-math></inline-formula> 33 k<inline-formula><tex-math>$Omega$</tex-math></inline-formula>). In this paper, we introduce a groundbreaking 3D-printed neural recorder with 50% reduced size over previous designs. Also, the matching circuit and antennas are designed for optimal communications and coupling, improving the MDS up to 10-fold. In addition, a full implantation in a large, regulated animal model (swine) was done for the first time, successfully recording evoked neural potentials. This will create the foundation for evaluating this system in chronic epileptic pigs and humans. Overall, the paper demonstrates the cutting-edge capabilities of our proposed recorder in the field of neurological activity monitoring.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"427-435"},"PeriodicalIF":3.2,"publicationDate":"2025-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560760","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Transcranial focused magnetoacoustic electrical stimulation (TFMAES) is an emerging composite neurostimulation technique, which can achieve precise electrical modulation of brain neural tissues with the help of low-intensity focused ultrasound. The aim of this study was to investigate the ameliorative effect of TFMAES on Alzheimer's disease (AD) symptoms. Through finite element simulation calculations, we determined the parameters of the TFMAES experiment. And 6 transgenic AD rats were used to perform continuous TFMAES experiments for 7 days at 2 different symptom stages in the progression of AD symptoms to observe the changes of electroencephalography (EEG) signals before and after the stimulation. The results showed that in the first stage, the amyloid beta (Aβ) deposition period, TFMAES led to a decrease in the delta-band energy percentage and an increase in the gamma-band energy percentage of EEG in AD rats. In the second stage, the onset of dementia symptoms, the effect of TFMAES on the delta-band power percentage of AD was significant compared to the gamma-band power percentage, but the effect on the overall EEG activity remained significant. These preliminary results suggest that TFMAES has a significant effect on the delta-band of EEG in AD rats, which may improve the activity of the cerebral cortex and thus improve AD symptoms. Moreover, this ameliorative effect is more significant in the early AD stage, which provides a basis for subsequent research on the mechanism of TFMAES quantitative parameters in AD improvement.
{"title":"Effects of Transcranial Focused Magnetoacoustic Electrical Stimulation on the EEG Signal of Alzheimer's Disease Rats","authors":"Ruolan Yang;Manxi Xu;Jixin Luan;Aocai Yang;Chao Zhang;Kuan Lv;Yuanyuan Li;Wenwei Zhang;Guoqiang Liu;Guolin Ma;Hui Xia","doi":"10.1109/JERM.2025.3573039","DOIUrl":"https://doi.org/10.1109/JERM.2025.3573039","url":null,"abstract":"Transcranial focused magnetoacoustic electrical stimulation (TFMAES) is an emerging composite neurostimulation technique, which can achieve precise electrical modulation of brain neural tissues with the help of low-intensity focused ultrasound. The aim of this study was to investigate the ameliorative effect of TFMAES on Alzheimer's disease (AD) symptoms. Through finite element simulation calculations, we determined the parameters of the TFMAES experiment. And 6 transgenic AD rats were used to perform continuous TFMAES experiments for 7 days at 2 different symptom stages in the progression of AD symptoms to observe the changes of electroencephalography (EEG) signals before and after the stimulation. The results showed that in the first stage, the amyloid beta (Aβ) deposition period, TFMAES led to a decrease in the delta-band energy percentage and an increase in the gamma-band energy percentage of EEG in AD rats. In the second stage, the onset of dementia symptoms, the effect of TFMAES on the delta-band power percentage of AD was significant compared to the gamma-band power percentage, but the effect on the overall EEG activity remained significant. These preliminary results suggest that TFMAES has a significant effect on the delta-band of EEG in AD rats, which may improve the activity of the cerebral cortex and thus improve AD symptoms. Moreover, this ameliorative effect is more significant in the early AD stage, which provides a basis for subsequent research on the mechanism of TFMAES quantitative parameters in AD improvement.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"417-426"},"PeriodicalIF":3.2,"publicationDate":"2025-06-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560759","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-04-30DOI: 10.1109/JERM.2025.3562571
Ayumi Ueda;Shouhei Kidera
This paper presents a cancer recognition scheme based on backscattered raw data using effective surface clutter rejection in microwave breast cancer diagnosis. Microwave breast cancer diagnostics enables a non-ionizing, non-compressive, low-cost examination, which can enhance the examination rate and frequency. A traditional radar image based cancer diagnosis faces a critical difficulty in discriminating cancer in highly dense breasts due to low contrast from fibro-glandular tissues. Therefore, this study introduces a direct recognition scheme from a complex-valued scattered signal, without using any imaging process, in which an efficient skin surface reflection (SSR) approach is introduced. Clinical data from over 100 Japanese subjects show that our SSR approach can enhance the recognition rate of cancerous tissues via a support vector machine (SVM) based learning approach.
{"title":"Clinical Validations on Effective Skin Clutter Rejection for Microwave Breast Cancer Diagnosis","authors":"Ayumi Ueda;Shouhei Kidera","doi":"10.1109/JERM.2025.3562571","DOIUrl":"https://doi.org/10.1109/JERM.2025.3562571","url":null,"abstract":"This paper presents a cancer recognition scheme based on backscattered raw data using effective surface clutter rejection in microwave breast cancer diagnosis. Microwave breast cancer diagnostics enables a non-ionizing, non-compressive, low-cost examination, which can enhance the examination rate and frequency. A traditional radar image based cancer diagnosis faces a critical difficulty in discriminating cancer in highly dense breasts due to low contrast from fibro-glandular tissues. Therefore, this study introduces a direct recognition scheme from a complex-valued scattered signal, without using any imaging process, in which an efficient skin surface reflection (SSR) approach is introduced. Clinical data from over 100 Japanese subjects show that our SSR approach can enhance the recognition rate of cancerous tissues via a support vector machine (SVM) based learning approach.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"400-407"},"PeriodicalIF":3.2,"publicationDate":"2025-04-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10980412","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560744","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-04-21DOI: 10.1109/JERM.2025.3559051
Carolina Moncion;Lakshmini Balachandar;Melany Gutierrez-Hernandez;John L. Volakis;Jorge Riera Diaz
To address the limitations of existing wired and wireless battery-powered designs for monitoring neuronal activity, our team has developed a novel implantable, battery-free, Wireless Neurosensing System (WiNS). Here, we aim to develop low-impedance microelectrodes to improve the minimum detectable signal for recording intracranial activity. For the first time, WiNS is evaluated in vivo to capture multiunit neuronal spiking, after adding a passive impedance matching (PIM) network. We explored different techniques for fabricating microelectrodes and demonstrated the resulting reduction in our system signal attenuation. Specifically, we were able to recover signals of 20 µVpp, a 10-fold improvement over previous designs. Furthermore, we fabricated microdrives to facilitate neuronal recording activity with WiNS and a wired system as a “gold” standard comparison. With the necessary components for our experiment, rat somatosensory cortex evoked neural activity was recorded. Broadband electrophysiological recordings were analyzed on multiple scales by evaluating the low-frequency component for elements corresponding to local field potentials (LFP) and the high-frequency component for multiunit neuronal spiking activity. LFP analysis involved the extraction of somatosensory evoked potentials to evaluate the recordings of each system. Concurrently, the neuronal spiking contributions were analyzed for spike occurrence and characteristics. Results at both scales indicate that recordings performed with WiNS are comparable with those of a wired system (p-value <0.050, Mann-Whitney Test). Therefore, the addition of PIM and the careful design of microelectrodes, led to the validation of WiNS using commercially available wired system. Furthermore, WiNS enables countless applications in neuroscience for continuous monitoring of neuronal activity.
{"title":"Efficient Recording of Rodent Neuronal Activity Using Microelectrodes With a Battery Free Wireless Neurosensing System","authors":"Carolina Moncion;Lakshmini Balachandar;Melany Gutierrez-Hernandez;John L. Volakis;Jorge Riera Diaz","doi":"10.1109/JERM.2025.3559051","DOIUrl":"https://doi.org/10.1109/JERM.2025.3559051","url":null,"abstract":"To address the limitations of existing wired and wireless battery-powered designs for monitoring neuronal activity, our team has developed a novel implantable, battery-free, Wireless Neurosensing System (WiNS). Here, we aim to develop low-impedance microelectrodes to improve the minimum detectable signal for recording intracranial activity. For the first time, WiNS is evaluated in vivo to capture multiunit neuronal spiking, after adding a passive impedance matching (PIM) network. We explored different techniques for fabricating microelectrodes and demonstrated the resulting reduction in our system signal attenuation. Specifically, we were able to recover signals of 20 µVpp, a 10-fold improvement over previous designs. Furthermore, we fabricated microdrives to facilitate neuronal recording activity with WiNS and a wired system as a “gold” standard comparison. With the necessary components for our experiment, rat somatosensory cortex evoked neural activity was recorded. Broadband electrophysiological recordings were analyzed on multiple scales by evaluating the low-frequency component for elements corresponding to local field potentials (LFP) and the high-frequency component for multiunit neuronal spiking activity. LFP analysis involved the extraction of somatosensory evoked potentials to evaluate the recordings of each system. Concurrently, the neuronal spiking contributions were analyzed for spike occurrence and characteristics. Results at both scales indicate that recordings performed with WiNS are comparable with those of a wired system (p-value <0.050, Mann-Whitney Test). Therefore, the addition of PIM and the careful design of microelectrodes, led to the validation of WiNS using commercially available wired system. Furthermore, WiNS enables countless applications in neuroscience for continuous monitoring of neuronal activity.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"393-399"},"PeriodicalIF":3.2,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560757","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-04-09DOI: 10.1109/JERM.2025.3555198
Shangyang Shang;Milad Mokhtari;Milica Popović
Microwave technology has recently been explored as a non-invasive method for skin cancer diagnosis. While most research has focused on reflection signal analysis, transmission-based approaches remain under-explored. In this work, surface wave transmissions generated by two antennas are employed to characterize skin cancer in a non-invasive way. We refined the theoretical model proposed in our previous work, showing that the transmission coefficient ($S_{21}$) amplitude and phase shift can effectively indicate both the presence and the size of skin tumors. Controlled experiments were conducted to validate the theory, using oil-gelatin-based phantoms to mimic both healthy skin and malignant tissue with varying tumor sizes. The experimental results strongly support the theoretical predictions and align with the simulation outcomes from our previous study.
{"title":"Employing Surface Waves for Characterizing Skin: Experimental Validation","authors":"Shangyang Shang;Milad Mokhtari;Milica Popović","doi":"10.1109/JERM.2025.3555198","DOIUrl":"https://doi.org/10.1109/JERM.2025.3555198","url":null,"abstract":"Microwave technology has recently been explored as a non-invasive method for skin cancer diagnosis. While most research has focused on reflection signal analysis, transmission-based approaches remain under-explored. In this work, surface wave transmissions generated by two antennas are employed to characterize skin cancer in a non-invasive way. We refined the theoretical model proposed in our previous work, showing that the transmission coefficient (<inline-formula><tex-math>$S_{21}$</tex-math></inline-formula>) amplitude and phase shift can effectively indicate both the presence and the size of skin tumors. Controlled experiments were conducted to validate the theory, using oil-gelatin-based phantoms to mimic both healthy skin and malignant tissue with varying tumor sizes. The experimental results strongly support the theoretical predictions and align with the simulation outcomes from our previous study.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 2","pages":"110-116"},"PeriodicalIF":3.0,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144117313","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-04-07DOI: 10.1109/JERM.2025.3555236
Xi Wang;Xiaofan Jia;Shao Ying Huang;Abdulkadir C. Yucel
Objective: As magnetic resonance imaging (MRI) technologies advance, predicting local Specific Absorption Rate (SAR) distributions becomes more challenging. This difficulty arises from the unique anatomical structures and dielectric properties of individual subjects, coupled with the inherent non-uniformity of energy deposition within tissues during scanning. To rapidly estimate SAR values induced by ultra-high-field (UHF) MRI birdcage RF coil in near real-time, this paper proposes a deep learning-based framework. Methods: The proposed framework consists of two stages. During the dataset generation stage, high-dimensional model representation, a polynomial-based surrogate modeling technique, is used to generate a large and diverse dataset, thereby reducing the reliance on resource-intensive deterministic simulations performed by physics-based simulators. During the inference stage, the framework employs 3D Attention U-Net, processing relative permittivity and conductivity maps of head models along with incident electric fields to predict SAR distributions. Results: The 3D Attention U-Net outperforms all other 3D U-Net variants and demonstrates remarkable accuracy, with mean relative errors of 7.57% for voxel SAR, 5.63% for 10g-averaged SAR, and 2.60% for peak spatial SAR. Each prediction can be performed in less than half a second, outperforming traditional physics-based simulators by at least three orders of magnitude. Conclusion: The framework provides a significant computational advantage over traditional physics-based simulators while maintaining satisfactory accuracy. Significance: The computational framework, available on GitHub, enables real-time SAR predictions on permittivity and conductivity distributions on any unseen MRI head models. The framework will allow ultra-fast optimization and uncertainty quantification studies to be performed while designing new UHF MRI coils.
{"title":"Deep Learning-Based Prediction of Specific Absorption Rate Induced by Ultra-High-Field MRI RF Head Coil","authors":"Xi Wang;Xiaofan Jia;Shao Ying Huang;Abdulkadir C. Yucel","doi":"10.1109/JERM.2025.3555236","DOIUrl":"https://doi.org/10.1109/JERM.2025.3555236","url":null,"abstract":"<italic>Objective:</i> As magnetic resonance imaging (MRI) technologies advance, predicting local Specific Absorption Rate (SAR) distributions becomes more challenging. This difficulty arises from the unique anatomical structures and dielectric properties of individual subjects, coupled with the inherent non-uniformity of energy deposition within tissues during scanning. To rapidly estimate SAR values induced by ultra-high-field (UHF) MRI birdcage RF coil in near real-time, this paper proposes a deep learning-based framework. <italic>Methods:</i> The proposed framework consists of two stages. During the dataset generation stage, high-dimensional model representation, a polynomial-based surrogate modeling technique, is used to generate a large and diverse dataset, thereby reducing the reliance on resource-intensive deterministic simulations performed by physics-based simulators. During the inference stage, the framework employs 3D Attention U-Net, processing relative permittivity and conductivity maps of head models along with incident electric fields to predict SAR distributions. <italic>Results:</i> The 3D Attention U-Net outperforms all other 3D U-Net variants and demonstrates remarkable accuracy, with mean relative errors of 7.57% for voxel SAR, 5.63% for 10g-averaged SAR, and 2.60% for peak spatial SAR. Each prediction can be performed in less than half a second, outperforming traditional physics-based simulators by at least three orders of magnitude. <italic>Conclusion:</i> The framework provides a significant computational advantage over traditional physics-based simulators while maintaining satisfactory accuracy. <italic>Significance:</i> The computational framework, available on GitHub, enables real-time SAR predictions on permittivity and conductivity distributions on any unseen MRI head models. The framework will allow ultra-fast optimization and uncertainty quantification studies to be performed while designing new UHF MRI coils.","PeriodicalId":29955,"journal":{"name":"IEEE Journal of Electromagnetics RF and Microwaves in Medicine and Biology","volume":"9 4","pages":"379-392"},"PeriodicalIF":3.2,"publicationDate":"2025-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145560660","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：