检索结果-内蒙古大学图书馆

Label-Efficient Self-Supervised Federated Learning for Tackling Data Heterogeneity in Medical Imaging

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yan, Rui Qu, Liangqiong Wei, Qingyue Huang, Shih-Cheng Shen, Liyue Rubin, Daniel Xing, Lei Zhou, Yuyin The Institute for Computational and Mathematical Engineering Stanford University StanfordCA94305 United States The Department of Statistics and Actuarial Science The Institute of Data Science The University of Hong Kong 999077 Hong Kong The Department of Biomedical Data Science Stanford University StanfordCA94305 United States The Department of Electrical Engineering Stanford University StanfordCA94305 United States The Department of Biomedical Informatics Stanford University StanfordCA94305 United States The Department of Radiation Oncology Stanford University StanfordCA94305 United States The Department of Computer Science and Engineering University of California Santa CruzCA95064 United States

The collection and curation of large-scale medical datasets from multiple institutions is essential for training accurate deep learning models, but privacy concerns often hinder data sharing. Federated learning (FL) is a promising solution that enables privacy-preserving collaborative learning among different institutions, but it generally suffers from performance deterioration due to heterogeneous data distributions and a lack of quality labeled data. In this paper, we present a robust and label-efficient self-supervised FL framework for medical image analysis. Our method introduces a novel Transformer-based self-supervised pre-training paradigm that pre-trains models directly on decentralized target task datasets using masked image modeling, to facilitate more robust representation learning on heterogeneous data and effective knowledge transfer to downstream models. Extensive empirical results on simulated and real-world medical imaging non-IID federated datasets show that masked image modeling with Transformers significantly improves the robustness of models against various degrees of data heterogeneity. Notably, under severe data heterogeneity, our method, without relying on any additional pre-training data, achieves an improvement of 5.06%, 1.53% and 4.58% in test accuracy on retinal, dermatology and chest X-ray classification compared to the supervised baseline with ImageNet pre-training. In addition, we show that our federated self-supervised pre-training methods yield models that generalize better to out-of-distribution data and perform more effectively when fine-tuning with limited labeled data, compared to existing FL algorithms. The code is available at https://***/rui-yan/SSL-FL. © 2022, CC BY.

关键词： Deep learning

A Dataset and Benchmark for Hospital Course Summarization with Adapted Large Language Models

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Aali, Asad Van Veen, Dave Arefeen, Yamin Ishraq Hom, Jason Bluethgen, Christian Reis, Eduardo Pontes Gatidis, Sergios Clifford, Namuun Daws, Joseph Tehrani, Arash S. Kim, Jangwon Chaudhari, Akshay S. Department of Radiology Stanford University StanfordCA United States Department of Electrical and Computer Engineering The University of Texas at Austin AustinTX United States Center for Artificial Intelligence in Medicine and Imaging Stanford University Palo AltoCA United States Department of Electrical Engineering Stanford University StanfordCA United States Department of Medicine StanfordCA United States University Hospital Zurich Zurich Switzerland Albert Einstein Israelite Hospital São Paulo Brazil School of Nursing The University of Texas at Austin AustinTX United States One Medical San FranciscoCA United States Amazon SeattleWA United States Department of Biomedical Data Science Stanford University StanfordCA United States Stanford University 1701 Page Mill Rd Palo AltoCA94304 United States

Objective: Brief hospital course (BHC) summaries are clinical documents that summarize a patient’s hospital stay. While large language models (LLMs) depict remarkable capabilities in automating real-world tasks, their capabilities for healthcare applications such as synthesizing BHCs from clinical notes have not been shown. We introduce a novel pre-processed dataset, the MIMIC-IV-BHC, encapsulating clinical note and brief hospital course (BHC) pairs to adapt LLMs for BHC synthesis. Furthermore, we introduce a benchmark of the summarization performance of two general-purpose LLMs and three healthcare-adapted LLMs. Materials and Methods: Using clinical notes as input, we apply prompting-based (using in-context learning) and fine-tuning-based adaptation strategies to three open-source LLMs (Clinical-T5-Large, Llama2-13B, FLAN-UL2) and two proprietary LLMs (GPT-3.5, GPT-4). We evaluate these LLMs across multiple context-length inputs using natural language similarity metrics. We further conduct a clinical study with five clinicians, comparing clinician-written and LLM-generated BHCs across 30 samples, focusing on their potential to enhance clinical decision-making through improved summary quality. We compare reader preferences for the original and LLM-generated summary using Wilcoxon Signed-Rank tests. We further request optional qualitative feedback from clinicians to gain deeper insights into their preferences, and we present the frequency of common themes arising from these comments. Results: The Llama2-13B fine-tuned LLM outperforms other domain-adapted models given quantitative evaluation metrics of BLEU and BERT-Score. GPT-4 with in-context learning shows more robustness to increasing context lengths of clinical note inputs than fine-tuned Llama2-13B. Despite comparable quantitative metrics, the reader study depicts a significant preference for summaries generated by GPT-4 with in-context learning compared to both Llama2-13B fine-tuned summaries and the original s

关键词： Electronic health record

25 GHz, 1 mV input resolution auxiliary circuit assisted comparator in 65 nm CMOS process

学校读者我要写书评

暂无评论

IET Circuits, Devices and Systems 2020年第3期14卷 340-346页

作者： Kala, Siddharth Rajkumar Chandaka, Sushma Nithin Kumar, Y.B. Harishchandra, Vasantha Moodabettu Bonizzoni, Edoardo Department of Electronics and Communication Engineering National Institute of Technology Goa Goa India Department of Electrical Computer and Biomedical Engineering University of Pavia Pavia Italy

The need for the high-speed analogue-to-digital converters demands the use of regenerative comparators. The strong positive feedback present in the regenerative comparators helps the comparator to work efficiently at the high-speed operations. This work proposes a low power auxiliary circuit to improve the high-frequency performance of the comparator. The proposed architecture along with the conventional comparators is simulated in 65-nm complementary metal oxide semiconductor (CMOS) technology with a supply voltage of 0.9 V. The maximum operating frequency of the proposed comparator is 6.25 GHz for a differential input voltage of 1 mV. © The Institution of engineering and Technology 2020.

关键词： Comparators (optical)

Controlled synthetic chirality in macroscopic assemblies of carbon nanotubes

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Doumani, Jacques Lou, Minhan Dewey, Oliver Hong, Nina Fan, Jichao Baydin, Andrey Yomogida, Yohei Yanagi, Kazuhiro Pasquali, Matteo Saito, Riichiro Kono, Junichiro Gao, Weilu Department of Electrical and Computer Engineering Rice University HoustonTX United States Applied Physics Graduate Program Smalley–Curl Institute Rice University HoustonTX United States Department of Electrical and Computer Engineering The University of Utah Salt Lake CityUT United States Carbon Hub Rice University HoustonTX United States Department of Chemical and Biomolecular Engineering Rice University HoustonTX United States J.A. Woollam Co. Inc. LincolnNE United States Smalley–Curl Institute Rice University HoustonTX United States Department of Physics Tokyo Metropolitan University Tokyo Japan Department of Chemistry Rice University HoustonTX United States Department of Materials Science and NanoEngineering Rice University HoustonTX United States Department of Physics Tohoku University Sendai Japan Department Physics and Astronomy Rice University HoustonTX United States

There is an emerging recognition that successful utilization of chiral degrees of freedom can bring new scientific and technological opportunities to diverse research areas. Hence, methods are being sought for creating artificial matter with controllable chirality in an uncomplicated and reproducible manner. Here, we report the development of two straightforward methods for fabricating wafer-scale chiral architectures of ordered carbon nanotubes (CNTs) with tunable and giant circular dichroism (CD). Both methods employ simple approaches, (i) mechanical rotation and (ii) twist-stacking, based on controlled vacuum filtration and do not involve any sophisticated nanofabrication processes. We used a racemic mixture of CNTs as the starting material, so the intrinsic chirality of chiral CNTs is not responsible for the observed chirality. In particular, by controlling the stacking angle and handedness in (ii), we were able to maximize the CD response and achieve a record-high deep-ultraviolet ellipticity of 40 ± 1 mdeg/nm. Our theoretical simulations using the transfer matrix method reproduce the salient features of the experimentally observed CD spectra and further predict that a film of twist-stacked CNTs with an optimized thickness will exhibit an ellipticity as high as 150 mdeg/nm. The created wafer-scale objects represent a new class of synthetic chiral matter consisting of ordered quantum wires whose macroscopic properties are governed by nanoscopic electronic signatures such as van Hove singularities. These artificial structures with engineered chirality will not only provide playgrounds for uncovering new chiral phenomena but also open up new opportunities for developing high-performance chiral photonic and optoelectronic devices. © 2023, CC BY.

关键词： Carbon nanotubes

Multi-Speaker Localization by Central and Lateral Microphone Arrays Based on the Combination of 2D-SRP and Subband GEVD Algorithms

学校读者我要写书评

暂无评论

Multi-Speaker Localization by Central and Lateral Microphone...

International Conference on Signal Processing and Communication (ICSC)

作者： Ali Dehghan Firoozabadi Pablo Irarrázaval Pablo Adasme David Zabala-Blanco Pablo Palacios-Játiva Hugo Durney Miguel Sanhueza Olave Cesar Azurdia Department of Electricity Universidad Tecnológica Metropolitana Av. José Pedro Alessandri 1242 Santiago Chile Electrical Engineering Department Pontificia Universidad Católica de Chile Santiago Chile Biomedical Imaging Center Pontificia Universidad Católica de Chile Santiago Chile Institute for Biological and Medical Engineering Pontificia Universidad Católica de Chile Santiago Chile Electrical Engineering Department Universidad de Santiago de Chile Av. Victor Jara 3519 Santiago Chile Department of Computer Science and Industry Universidad Católica del Maule Talca Chile Department of Electrical Engineering Universidad de Chile Santiago Chile Escuela de Informática y Telecomunicaciones Universidad Diego Portales Santiago Chile

ISBN: (纸本)9781665454315

In this paper, a novel speech source localization (SSL) algorithm by the use of the central and lateral microphone arrays is proposed with low complexity and high precision in undesirable acoustical environments. Firstly, a proposed star-shaped microphone array (SSMA) at the room’s center is considered in combination with the improved 2D steered response power (2D-SRP) algorithm for determining the speakers’ directions, where the PENS algorithm is considered for speakers counting. Then, the two closest cross-shaped arrays (CSMA) to each speaker on the wall are selected, where these arrays prepare the proper information with less microphones. Each CSMA is considered in combination with subband generalized eigenvalue decomposition (SB-GEVD) method on various time frames for estimating the two others speaker’s direction. Therefore, the 3D speakers’ positions are calculated by the crossing of the estimated directions from SSMA and two CSMAs. The proposed lateral-central microphone array-multiple sound source localization method (LCMA-MSSL) is compared with other research, which shows the superiority of the proposed method in undesirable noisy and reverberant scenarios.

关键词： Location awareness Three-dimensional displays Direction-of-arrival estimation Signal processing algorithms Estimation Parallel processing Microphone arrays

Spectro-ViT: A Vision Transformer Model for GABA-edited MRS Reconstruction Using Spectrograms

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Dias, Gabriel Berto, Rodrigo Pommot Oliveira, Mateus Ueda, Lucas Dertkigil, Sergio Costa, Paula D.P. Shamaei, Amirmohammad Souza, Roberto Harris, Ashley Rittner, Leticia School of Electrical and Computer Engineering University of Campinas Campinas Brazil Department of Biomedical Engineering University of Calgary Calgary Canada Hotchkiss Brain Institute University of Calgary Calgary Canada Alberta Children’s Hospital Research Institute Calgary Canada Research and Development Center in Telecommunications CPQD Campinas Brazil School of Medical Sciences University of Campinas Campinas Brazil Artificial Intelligence Lab. Recod.ai University of Campinas Campinas Brazil Department of Electrical and Software Engineering University of Calgary Calgary Canada Department of Radiology University of Calgary Calgary Canada ICT 352C 2500 University Dr NW CalgaryABT2N 1N4 Canada

Purpose: To investigate the use of a Vision Transformer (ViT) to reconstruct/denoise GABA-edited magnetic resonance spectroscopy (MRS) from a quarter of the typically acquired number of transients using spectrograms. Theory and Methods: A quarter of the typically acquired number of transients collected in GABA-edited MRS scans are pre-processed and converted to a spectrogram image representation using the Short-Time Fourier Transform (STFT). The image representation of the data allows the adaptation of a pre-trained ViT for reconstructing GABA-edited MRS spectra (Spectro-ViT). The Spectro-ViT is fine-tuned and then tested using in vivo GABA-edited MRS data. The Spectro-ViT performance is compared against other models in the literature using spectral quality metrics and estimated metabolite concentration values. Results: The Spectro-ViT model significantly outperformed all other models in four out of five quantitative metrics (mean squared error, shape score, GABA+/water fit error, and full width at half maximum). The metabolite concentrations estimated (GABA+/water, GABA+/Cr, and Glx/water) were consistent with the metabolite concentrations estimated using typical GABA-edited MRS scans reconstructed with the full amount of typically collected transients. Conclusion: The proposed Spectro-ViT model achieved state-of-the-art results in reconstructing GABA-edited MRS, and the results indicate these scans could be up to four times faster. © 2023, CC BY.

关键词： Deep learning

Multimodal monitoring of human brain organoids implanted in mice using transparent microelectrodes

学校读者我要写书评

暂无评论

Multimodal monitoring of human brain organoids implanted in ...

Optics and the Brain, BRAIN 2021 - Part of Biophotonics Congress: Optics in the Life Sciences 2021

作者： Thunemann, Martin Wilson, Madison Puppo, Francesca Pizzo, Donald P. Rickner, Hannah Bortz, Emma Mansour, Abed AlFatah Wolozin, Benjamin Han, Xue Gage, Fred H. Muotri, Alysson R. Kuzum, Duygu Devor, Anna Department of Biomedical Engineering Boston University BostonMA02215 United States Department of Electrical and Computer Engineering University of California San Diego La Jolla CA92093 United States Department of Pediatrics University of California San Diego La Jolla CA92093 United States Department of Pathology University of California San Diego La Jolla CA92093 United States Department of Pharmacology Boston University BostonMA02215 United States The Salk Institute for Biological Studies La Jolla CA92093 United States Martinos Center for Biomedical Imaging MGH Harvard Medical School CharlestownMA02129 United States

We present an in vivo platform merging stem cell technology, two-photon imaging, and electrophysiology with transparent graphene microelectrodes for longitudinal interrogation of human brain organoids implanted in the... 详细信息

ISBN: (纸本)9781557528209

关键词： Microelectrodes

Deep learning in biomedical optics

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Tian, Lei Hunt, Brady Lediju Bell, Muyinatu A. Yi, Ji Smith, Jason T. Ochoa, Marien Intes, Xavier Durr, Nicholas J. Department of Electrical and Computer Engineering Boston University BostonMA United States Thayer School of Engineering Dartmouth College HanoverNH United States Department of Electrical and Computer Engineering Johns Hopkins University BaltimoreMD United States Department of Biomedical Engineering Johns Hopkins University BaltimoreMD United States Department of Computer Science Johns Hopkins University BaltimoreMD United States Department of Ophthalmology Johns Hopkins University BaltimoreMD United States Center for Modeling Simulation and Imaging in Medicine Rensselaer Polytechnic Institute

This article reviews deep learning applications in biomedical optics with a particular emphasis on image formation. The review is organized by imaging domains within biomedical optics and includes microscopy, fluorescence lifetime imaging, in vivo microscopy, widefield endoscopy, optical coherence tomography, photoacoustic imaging, diffuse tomography, and functional optical brain imaging. For each of these domains, we summarize how deep learning has been applied and highlight methods by which deep learning can enable new capabilities for optics in medicine. Challenges and opportunities to improve translation and adoption of deep learning in biomedical optics are also summarized. Copyright © 2021, The Authors. All rights reserved.

关键词： Optical tomography

Denoising Wearable Armband ECG Data Using the Variable Frequency Complex Demodulation Technique

学校读者我要写书评

暂无评论

Denoising Wearable Armband ECG Data Using the Variable Frequ...

Annual International Conference of the IEEE engineering in Medicine and Biology Society (EMBC)

作者： Md-Billal Hossain Jesús Lázaro Yeonsik Noh Ki H. Chon Department of Biomedical Engineering University of Connecticut Storrs CT USA Department of Electrical and Computer Engineering University of Massachusetts Amherst

ISBN: (数字)9781728119908

ISBN: (纸本)9781728119915

We propose a novel electrocardiogram (ECG) denoising technique using the variable frequency complex demodulation (VFCDM) algorithm. We used VFCDM to perform the sub-band decomposition of the noise-contaminated ECG to remove the noise components so that accurate QRS complexes could be identified. The ECG quality was further improved by removing baseline drift and smoothing via adaptive mean filtering. The proposed method was validated on the MIT-BIH arrhythmia database (MITDB) and wearable armband ECG data. For the former, we added Gaussian white noise to the ECG signals at different signal-to-noise ratios and the denoising performance of the proposed method was compared with other denoising techniques. The proposed approach showed superior denoising performance compared to the other methods. We compared the QRS complex detection performance of the noisy to the denoised armband ECG. The performance of the proposed denoising method on the armband ECG resulted in comparable QRS complex detection as that obtained when using Holter monitor ECG signals. This demonstrates that the proposed algorithm can significantly increase the amount of usable armband ECG data, which would otherwise have been discarded due to electromyogram contamination especially during arm movements. Hence, the proposed algorithm has the potential to enable long-term monitoring of atrial fibrillation using the armband without the discomfort of skin irritation often experienced with Holter *** Relevance- The proposed ECG denoising method can significantly increase the ECG quality of wearable ECG devices, which are more susceptible to muscle and motion artifacts.

关键词： Electrocardiography Noise reduction Noise measurement Monitoring biomedical monitoring Filtering Signal to noise ratio