arXiv - PHYS - Medical Physics最新文献

{"title":"Characterizing the circularly-oriented macular pigment using spatiotemporal sensitivity to structured light entoptic phenomena","authors":"Dmitry A. Pushin, Davis V. Garrad, Connor Kapahi, Andrew E. Silva, Pinki Chahal, David G. Cory, Mukhit Kulmaganbetov, Iman Salehi, Melanie Mungalsingh, Taranjit Singh, Benjamin Thompson, Dusan Sarenac","doi":"arxiv-2409.04416","DOIUrl":"https://doi.org/arxiv-2409.04416","url":null,"abstract":"The macular pigment (MP) in the radially-oriented Henle fibers that surround\u0000the foveola enables the ability to perceive the orientation of polarized blue\u0000light through an entoptic phenomena known as the Haidinger's brush. The MP has\u0000been linked to eye diseases and central field dysfunctions, most notably\u0000age-related macular degeneration (AMD), a globally leading cause of\u0000irreversible blindness. Recent integration of structured light techniques into\u0000vision science has allowed for the development of more selective and versatile\u0000entoptic probes of eye health that provide interpretable thresholds. For\u0000example, it enabled the use of variable spatial frequencies and arbitrary\u0000obstructions in the presented stimuli. Additionally, it expanded the 2{deg}\u0000retinal eccentricity extent of the Haidinger's brush to 5{deg} for a similar\u0000class of fringe-based stimuli. In this work, we develop a spatiotemporal\u0000sensitivity model that maps perceptual thresholds of entoptic phenomenon to the\u0000underlying MP structure that supports its perception. We therefore selectively\u0000characterize the circularly-oriented macular pigment optical density (coMPOD)\u0000rather than total MPOD as typically measured, providing an additional\u0000quantification of macular health. A study was performed where the retinal\u0000eccentricity thresholds were measured for five structured light stimuli with\u0000unique spatiotemporal frequencies. The results from fifteen healthy young\u0000participants indicate that the coMPOD is inversely proportional to retinal\u0000eccentricity in the range of 1.5{deg} to 5.5{deg}. Good agreement between the\u0000model and the collected data is found with a Pearson $chi^2$ fit statistic of\u00000.06. The presented techniques can be applied in novel early diagnostic tests\u0000for a variety of diseases related to macular degeneration such as AMD, macular\u0000telangiectasia, and pathological myopia.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"27 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176635","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}

{"title":"TG-LMM: Enhancing Medical Image Segmentation Accuracy through Text-Guided Large Multi-Modal Model","authors":"Yihao Zhao, Enhao Zhong, Cuiyun Yuan, Yang Li, Man Zhao, Chunxia Li, Jun Hu, Chenbin Liu","doi":"arxiv-2409.03412","DOIUrl":"https://doi.org/arxiv-2409.03412","url":null,"abstract":"We propose TG-LMM (Text-Guided Large Multi-Modal Model), a novel approach\u0000that leverages textual descriptions of organs to enhance segmentation accuracy\u0000in medical images. Existing medical image segmentation methods face several\u0000challenges: current medical automatic segmentation models do not effectively\u0000utilize prior knowledge, such as descriptions of organ locations; previous\u0000text-visual models focus on identifying the target rather than improving the\u0000segmentation accuracy; prior models attempt to use prior knowledge to enhance\u0000accuracy but do not incorporate pre-trained models. To address these issues,\u0000TG-LMM integrates prior knowledge, specifically expert descriptions of the\u0000spatial locations of organs, into the segmentation process. Our model utilizes\u0000pre-trained image and text encoders to reduce the number of training parameters\u0000and accelerate the training process. Additionally, we designed a comprehensive\u0000image-text information fusion structure to ensure thorough integration of the\u0000two modalities of data. We evaluated TG-LMM on three authoritative medical\u0000image datasets, encompassing the segmentation of various parts of the human\u0000body. Our method demonstrated superior performance compared to existing\u0000approaches, such as MedSAM, SAM and nnUnet.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"76 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176638","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}

{"title":"Deep Brain Ultrasound Ablation Thermal Dose Modeling with in Vivo Experimental Validation","authors":"Zhanyue Zhao, Benjamin Szewczyk, Matthew Tarasek, Charles Bales, Yang Wang, Ming Liu, Yiwei Jiang, Chitresh Bhushan, Eric Fiveland, Zahabiya Campwala, Rachel Trowbridge, Phillip M. Johansen, Zachary Olmsted, Goutam Ghoshal, Tamas Heffter, Katie Gandomi, Farid Tavakkolmoghaddam, Christopher Nycz, Erin Jeannotte, Shweta Mane, Julia Nalwalk, E. Clif Burdette, Jiang Qian, Desmond Yeo, Julie Pilitsis, Gregory S. Fischer","doi":"arxiv-2409.02395","DOIUrl":"https://doi.org/arxiv-2409.02395","url":null,"abstract":"Intracorporeal needle-based therapeutic ultrasound (NBTU) is a minimally\u0000invasive option for intervening in malignant brain tumors, commonly used in\u0000thermal ablation procedures. This technique is suitable for both primary and\u0000metastatic cancers, utilizing a high-frequency alternating electric field (up\u0000to 10 MHz) to excite a piezoelectric transducer. The resulting rapid\u0000deformation of the transducer produces an acoustic wave that propagates through\u0000tissue, leading to localized high-temperature heating at the target tumor site\u0000and inducing rapid cell death. To optimize the design of NBTU transducers for\u0000thermal dose delivery during treatment, numerical modeling of the acoustic\u0000pressure field generated by the deforming piezoelectric transducer is\u0000frequently employed. The bioheat transfer process generated by the input\u0000pressure field is used to track the thermal propagation of the applicator over\u0000time. Magnetic resonance thermal imaging (MRTI) can be used to experimentally\u0000validate these models. Validation results using MRTI demonstrated the\u0000feasibility of this model, showing a consistent thermal propagation pattern.\u0000However, a thermal damage isodose map is more advantageous for evaluating\u0000therapeutic efficacy. To achieve a more accurate simulation based on the actual\u0000brain tissue environment, a new finite element method (FEM) simulation with\u0000enhanced damage evaluation capabilities was conducted. The results showed that\u0000the highest temperature and ablated volume differed between experimental and\u0000simulation results by 2.1884{deg}C (3.71%) and 0.0631 cm$^3$ (5.74%),\u0000respectively. The lowest Pearson correlation coefficient (PCC) for peak\u0000temperature was 0.7117, and the lowest Dice coefficient for the ablated area\u0000was 0.7021, indicating a good agreement in accuracy between simulation and\u0000experiment.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"31 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176639","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}

{"title":"Fast and Accurate Collimator-Detector Response Compensation in High-Energy SPECT Imaging with 1D Convolutions and Rotations","authors":"Lucas Polson, Pedro Esquinas, Sara Kurkowska, Chenguang Li, Peyman Sheikhzadeh, Mehrshad Abbassi, Saeed Farzanehfar, Seyyede Mirabedian, Carlos Uribe, Arman Rahmim","doi":"arxiv-2409.03100","DOIUrl":"https://doi.org/arxiv-2409.03100","url":null,"abstract":"Modeling of the collimator-detector response (CDR) in SPECT reconstruction\u0000enables improved resolution and more accurate quantitation, especially for\u0000higher energy imaging (e.g.Lu-177 and Ac-225). Such modeling, however, can pose\u0000a significant computational bottleneck when there are substantial components of\u0000septal penetration and scatter in the acquired data, since a direct\u0000convolution-based approach requires large 2D kernels. The present work presents\u0000an alternative method for fast and accurate CDR compensation using a linear\u0000operator built from 1D convolutions and rotations (1D-R). To enable open-source\u0000development and use of these models in image reconstruction, we release a\u0000SPECTPSFToolbox repository for the PyTomography project on GitHub.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"46 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176637","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}

{"title":"Investigation of the spatial resolution of PET imaging system measuring polarization-correlated Compton events","authors":"Ana Marija Kožuljević, Tomislav Bokulić, Darko Grošev, Zdenka Kuncic, Siddharth Parashari, Luka Pavelić, Mihael Makek","doi":"arxiv-2409.01238","DOIUrl":"https://doi.org/arxiv-2409.01238","url":null,"abstract":"Recent studies of positron emission tomography (PET) devices have shown that\u0000the detection of polarization-correlated annihilation quanta can potentially\u0000reduce the background and creation of false lines of response (LORs) leading to\u0000improved image quality. We developed a novel PET demonstrator system, capable\u0000of measuring correlated gamma photons with single-layer Compton polarimeters to\u0000explore the potential of the method. We tested the system using sources with\u0000clinically relevant activities at the University Hospital Centre Zagreb. Here\u0000we present, for the first time, the images of two Ge-68 line sources,\u0000reconstructed solely from the correlated annihilation events. The spatial\u0000resolution at two different diameters is determined and compared to the one\u0000obtained from events with photoelectric interaction.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"87 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176641","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}

{"title":"Objective Features Extracted from Motor Activity Time Series for Food Addiction Analysis Using Machine Learning","authors":"Mikhail Borisenkov, Andrei Velichko, Maksim Belyaev, Dmitry Korzun, Tatyana Tserne, Larisa Bakutova, Denis Gubin","doi":"arxiv-2409.00310","DOIUrl":"https://doi.org/arxiv-2409.00310","url":null,"abstract":"This study investigates machine learning algorithms to identify objective\u0000features for diagnosing food addiction (FA) and assessing confirmed symptoms\u0000(SC). Data were collected from 81 participants (mean age: 21.5 years, range:\u000018-61 years, women: 77.8%) whose FA and SC were measured using the Yale Food\u0000Addiction Scale (YFAS). Participants provided demographic and anthropometric\u0000data, completed the YFAS, the Zung Self-Rating Depression Scale, and the Dutch\u0000Eating Behavior Questionnaire, and wore an actimeter on the non-dominant wrist\u0000for a week to record motor activity. Analysis of the actimetric data identified\u0000significant statistical and entropy-based features that accurately predicted FA\u0000and SC using ML. The Matthews correlation coefficient (MCC) was the primary\u0000metric. Activity-related features were more effective for FA prediction\u0000(MCC=0.88) than rest-related features (MCC=0.68). For SC, activity segments\u0000yielded MCC=0.47, rest segments MCC=0.38, and their combination MCC=0.51.\u0000Significant correlations were also found between actimetric features related to\u0000FA, emotional, and restrained eating behaviors, supporting the model's\u0000validity. Our results support the concept of a human bionic suite composed of\u0000IoT devices and ML sensors, which implements health digital assistance with\u0000real-time monitoring and analysis of physiological indicators related to FA and\u0000SC.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"19 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176640","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}

{"title":"Prediction of excitable wave dynamics using machine learning","authors":"Mahesh Kumar Mulimani, Sebastian Echeverria-Alar, Michael Reiss, Wouter-Jan Rappel","doi":"arxiv-2409.00278","DOIUrl":"https://doi.org/arxiv-2409.00278","url":null,"abstract":"Excitable systems, including cardiac tissue, can exhibit a variety of\u0000dynamics with different complexity, ranging from a single, stable spiral to\u0000spiral defect chaos (SDC), during which spiral waves are continuously formed\u0000and destroyed. Cardiac models typically involve a large number of variables and\u0000can be time-consuming to simulate. Here we trained a deep-learning (DL) model\u0000using snapshots from a single variable, obtained by simulating a single\u0000quasi-periodic spiral wave and spiral defect chaos (SDC) using a generic\u0000cardiac model. Using the trained DL model, we predicted the dynamics in both\u0000cases, using timesteps that are much larger than required for the simulations\u0000of the underlying equations. We show that the DL model is able to predict the\u0000trajectory of a quasi-periodic spiral wave and that the SDC activaton patterns\u0000can be predicted for approximately one Lyapunov time. Furthermore, we show that\u0000the DL model accurately captures the statistics of termination events in SDC,\u0000including the mean termination time. Finally, we show that a DL model trained\u0000using a specific domain size is able to replicate termination statistics on\u0000larger domains, resulting in significant computational savings.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"54 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176642","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}

{"title":"A theoretical framework for the assessment of water fraction-dependent longitudinal decay rates and magnetisation transfer in membrane lipid phantoms","authors":"Heiko Neeb, Felix Schyboll, Rona Shaharabani, Aviv A. Mezer, Oshrat Shtangel","doi":"arxiv-2408.17085","DOIUrl":"https://doi.org/arxiv-2408.17085","url":null,"abstract":"Phantom systems consisting of liposome suspensions are widely employed to\u0000investigate quantitative MRI parameters mimicking cellular membranes. The\u0000proper physical understanding of the measurement results, however, requires\u0000proper models for liposomes and their interaction with the surrounding water\u0000molecules. Here, we present an MD-based approach for the theoretical prediction\u0000of R1=1/T1, the dependence of R1 on water concentration and the magnetization\u0000exchange between lipids and interacting water layer in lipids and lipid\u0000mixtures. Moreover, a new parameter is introduced which quantitatively measures\u0000the amount of hydration water (hydration water fraction, f_HW) based on\u0000conventional spoiled gradient echo MR acquisitions. Both f_HW and the\u0000magnetisation exchange rate between lipids and hydration water were determined\u0000quantitatively from spoiled gradient echo data. We observed that liposome\u0000systems behaved similarly, apart from PLPC which showed both lower hydration\u0000water fraction and lower exchange rate. The extracted parameters accurately\u0000predicted the measured water fraction-dependent R1 rates and allowed for a\u0000theoretical understanding of MR parameters in liposomes of different\u0000composition.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"76 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176643","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}

{"title":"Motion-Driven Neural Optimizer for Prophylactic Braces Made by Distributed Microstructures","authors":"Xingjian Han, Yu Jiang, Weiming Wang, Guoxin Fang, Simeon Gill, Zhiqiang Zhang, Shengfa Wang, Jun Saito, Deepak Kumar, Zhongxuan Luo, Emily Whiting, Charlie C. L. Wang","doi":"arxiv-2408.16659","DOIUrl":"https://doi.org/arxiv-2408.16659","url":null,"abstract":"Joint injuries, and their long-term consequences, present a substantial\u0000global health burden. Wearable prophylactic braces are an attractive potential\u0000solution to reduce the incidence of joint injuries by limiting joint movements\u0000that are related to injury risk. Given human motion and ground reaction forces,\u0000we present a computational framework that enables the design of personalized\u0000braces by optimizing the distribution of microstructures and elasticity. As\u0000varied brace designs yield different reaction forces that influence kinematics\u0000and kinetics analysis outcomes, the optimization process is formulated as a\u0000differentiable end-to-end pipeline in which the design domain of microstructure\u0000distribution is parameterized onto a neural network. The optimized distribution\u0000of microstructures is obtained via a self-learning process to determine the\u0000network coefficients according to a carefully designed set of losses and the\u0000integrated biomechanical and physical analyses. Since knees and ankles are the\u0000most commonly injured joints, we demonstrate the effectiveness of our pipeline\u0000by designing, fabricating, and testing prophylactic braces for the knee and\u0000ankle to prevent potentially harmful joint movements.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"3 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176644","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}

{"title":"Electrical Impedance Tomography meets Reduced Order Modelling: a framework for faster and more reliable electrical conductivity estimations","authors":"Matthew R. Walker, Mariano Fernández-Corazza, Sergei Turovets, Leandro Beltrachini","doi":"arxiv-2408.15673","DOIUrl":"https://doi.org/arxiv-2408.15673","url":null,"abstract":"Objective: Inclusion of individualised electrical conductivities of head\u0000tissues is crucial for the accuracy of electrical source imaging techniques\u0000based on electro/magnetoencephalography and the efficacy of transcranial\u0000electrical stimulation. Parametric electrical impedance tomography (pEIT) is a\u0000method to cheaply and non-invasively estimate them using electrode arrays on\u0000the scalp to apply currents and measure the resulting potential distribution.\u0000Conductivities are then estimated by iteratively fitting a forward model to the\u0000measurements, incurring a prohibitive computational cost that is generally\u0000lowered at the expense of accuracy. Reducing the computational cost associated\u0000with the forward solutions would improve the accessibility of this method and\u0000unlock new capabilities. Approach: We introduce reduced order modelling (ROM)\u0000to massively speed up the calculations of these solutions for arbitrary\u0000conductivity values. Main results: We demonstrate this new ROM-pEIT framework\u0000using a realistic head model with six tissue compartments, with minimal errors\u0000in both the approximated numerical solutions and conductivity estimations. We\u0000show that the computational complexity required to reach a multi-parameter\u0000estimation with a negligible relative error is reduced by more than an order of\u0000magnitude when using this framework. Furthermore, we illustrate the benefits of\u0000this new framework in a number of practical cases, including its application to\u0000real pEIT data from three subjects. Significance: Results suggest that this\u0000framework can transform the use of pEIT for seeking personalised head\u0000conductivities, making it a valuable tool for researchers and clinicians.","PeriodicalId":501378,"journal":{"name":"arXiv - PHYS - Medical Physics","volume":"385 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142176650","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}