检索结果-内蒙古大学图书馆

A Framework for Interpretability in Machine Learning for Medical Imaging

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Alan Q. Karaman, Batuhan K. Kim, Heejong Rosenthal, Jacob Saluja, Rachit Young, Sean I. Sabuncu, Mert R. School of Electrical and Computer Engineering Cornell University Cornell Tech New YorkNY10044 United States Department of Radiology Weill Cornell Medical School New YorkNY10065 United States Weill Cornell/Rockefeller/Sloan Kettering Tri-Institutional MD-PhD Program New YorkNY10065 United States Martinos Center for Biomedical Imaging Harvard Medical School BostonMA02129 United States Computer Science and Artificial Intelligence Laboratory MIT CambridgeMA02139 United States

Interpretability for machine learning models in medical imaging (MLMI) is an important direction of research. However, there is a general sense of murkiness in what interpretability means. Why does the need for interpretability in MLMI arise? What goals does one actually seek to address when interpretability is needed? To answer these questions, we identify a need to formalize the goals and elements of interpretability in MLMI. By reasoning about real-world tasks and goals common in both medical image analysis and its intersection with machine learning, we identify five core elements of interpretability: localization, visual recognizability, physical attribution, model transparency, and actionability. From this, we arrive at a framework for interpretability in MLMI, which serves as a step-by-step guide to approaching interpretability in this context. Overall, this paper formalizes interpretability needs in the context of medical imaging, and our applied perspective clarifies concrete MLMI-specific goals and considerations in order to guide method design and improve real-world usage. Our goal is to provide practical and didactic information for model designers and practitioners, inspire developers of models in the medical imaging field to reason more deeply about what interpretability is achieving, and suggest future directions of interpretability research. © 2023, CC0.

关键词： Machine learning

A Deep Learning Approach for Nerve Injury Classification in Brachial Plexopathies Using Magnetic Resonance Neurography with Modified Hiking Optimization Algorithm

学校读者我要写书评

暂无评论

Academic Radiology 2025年

作者： Dahou, Abdelghani Elaziz, Mohamed Abd Khattap, Mohamed G. Hassan, Hend Galal Eldeen Mohamed Ali School of Computer Science and Technology Zhejiang Normal University Jinhua 321004 China Mathematics and Computer Science department University of Ahmed DRAIA Adrar 01000 Algeria Department of Mathematics Faculty of Science Zagazig University Zagazig 44519 Egypt Faculty of Computer Science and Engineering Galala University Suez 435611 Egypt Artificial Intelligence Research Center (AIRC) Ajman University Ajman 346 United Arab Emirates Department of Diagnostic Interventional Radiology and Molecular Imaging Faculty of Medicine Ain Shams University Cairo 11591 Egypt Technology of Radiology and Medical Imaging Program Faculty of Applied Health Sciences Technology Galala University Suez 435611 Egypt

Rationale and Objectives: Brachial plexopathies (BPs) encompass a complex spectrum of nerve injuries affecting motor and sensory function in the upper extremities. Diagnosis is challenging due to the intricate anatomy and symptom overlap with other neuropathies. Magnetic Resonance Neurography (MRN) provides advanced imaging but requires specialized interpretation. This study proposes an AI-based framework that combines deep learning (DL) with the modified Hiking Optimization Algorithm (MHOA) enhanced by a Comprehensive Learning (CL) technique to improve the classification of nerve injuries (neuropraxia, axonotmesis, neurotmesis) using MRN data. Materials and Methods: The framework utilizes MobileNetV4 for feature extraction and MHOA for optimized feature selection across different MRI sequences (STIR, T2, T1, and DWI). A dataset of 39 patients diagnosed with BP was used. The framework classifies injuries based on Seddon's criteria, distinguishing between normal and abnormal conditions as well as injury severity. Results: The model achieved excellent performance, with 1.0000 accuracy in distinguishing normal from abnormal conditions using STIR and T2 sequences. For injury severity classification, accuracy was 0.9820 in STIR, outperforming the original HOA and other metaheuristic algorithms. Additionally, high classification accuracy (0.9667) was observed in DWI. The proposed framework outperformed traditional methods and demonstrated high sensitivity and specificity. Conclusion: The proposed AI-based framework significantly improves the diagnosis of BP by accurately classifying nerve injury types. By integrating DL and optimization techniques, it reduces diagnostic variability, making it a valuable tool for clinical settings with limited specialized neuroimaging expertise. This framework has the potential to enhance clinical decision-making and optimize patient outcomes through precise and timely diagnoses. © 2025 The Association of University Radiologists

关键词： Brachial Plexopathy (BP) Comprehensive Learning (CL) Deep Learning (DL) Feature Selection (FS) Hiking Optimization Algorithm (HOA) Magnetic Resonance Neurography (MRN) MobileNetV4

Lightweight equivariant model for efficient machine learning interatomic potentials

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yang, Ziduo Wang, Xian Li, Yifan Lv, Qiujie Chen, Calvin Yu-Chian Shen, Lei Department of Mechanical Engineering National University of Singapore Singapore117575 Singapore Artificial Intelligence Medical Research Center School of Intelligent Systems Engineering Shenzhen Campus of Sun Yat-sen University Shenzhen518107 China Department of Physics National University of Singapore Singapore117551 Singapore -Preferred Program School of Electronic and Computer Engineering Peking University Shenzhen Graduate School Shenzhen518055 China State Key Laboratory of Chemical Oncogenomics School of Chemical Biology and Biotechnology Peking University Shenzhen Graduate School Shenzhen518055 China Guangdong L-Med Biotechnology Co. Ltd Guangdong Meizhou514699 China Research Institute Chongqing401123 China

In modern computational materials science, deep learning has shown the capability to predict interatomic potentials, thereby supporting and accelerating conventional simulations. However, existing models typically sacrifice either accuracy or efficiency. Moreover, lightweight models are highly demanded for offering simulating systems on a considerably larger scale at reduced computational costs. Here, we introduce a lightweight equivariant interaction graph neural network (LEIGNN) that can enable accurate and efficient interatomic potential and force predictions for molecules and crystals. Rather than relying on higher-order representations, LEIGNN employs a scalar-vector dual representation to encode equivariant features. By learning geometric symmetry information, our model remains lightweight while ensuring prediction accuracy and robustness through the equivariance. Our results show that LEIGNN consistently outperforms the prediction performance of the representative baselines and achieves significant efficiency across diverse datasets, which include catalysts, molecules, and organic isomers. Furthermore, we conduct molecular dynamics (MD) simulations using the LEIGNN force field across solid, liquid, and gas systems. It is found that LEIGNN can achieve the accuracy of ab initio MD across all examined systems. Copyright © 2023, The Authors. All rights reserved.

关键词： Prediction models

Ant Colony Optimization based Medical Image Preservation and Segmentation

学校读者我要写书评

暂无评论

Ant Colony Optimization based Medical Image Preservation and...

Advanced Technologies in Intelligent Control, Environment, Computing & Communication engineering (ICATIECE), International Conference on

作者： Priya Nandihal Piyush Kumar Pareek Victor Hugo Costa De Albuquerque Madhumala R B Ashish Khanna V Sunil Kumar Dept of Computer Science and Design Dayananda Sagar Academy of Technology and Management Bangalore Graduate Program in Telecommunication Engineering Federal Institute of Education Science and Technology of Ceara Fortaleza Brazil Department of Teleinformatics Engineering Federal University of Ceara Brazil Department of ISE Dayananda Sagar Academy of Technology and Management Bangalore CSE Dept Maharaja Agrasen Institute of Technology GGSIPU Department of Artificial Intelligence & Machine Learning Nitte Meenakshi Institute of Technology Bengaluru

ISBN: (纸本)9781665493970

Image segmentation is an essential component in many different types of computer vision systems. Image segmentation is used in order to identify objects and boundaries within pictures. It is important to note that the success of recognition algorithms is heavily dependent on the quality of the picture segmentation, and there is yet no technique that can universally segment images. Boundary detection has a long history of being recognised as one of the most difficult challenges in the fields of image processing and pattern analysis, in particular, when it comes to applications in medical imaging. The clarity of the picture is being muddled by more information, which is the source of the problem's complexity. The creation of segmentation algorithms for medical pictures is an active field of study, and segmentation algorithms are meant to account for a variety of factors, including noise, intensity inhomogeneity, poor contrast between the lesion and the surrounding skin, and so on. There is currently no active contour model that combines the locally statistical property of being more noise-resistant, in which the intensities in the transformed domain have less overlapping in the statistics, and the locally image fitting property of extracting the local image information in order to be able to segment images with intensity inhomogeneity. This is because there is no such model. So the proposed work includes Ant colony optimization based double threshold segmentation in medical images. Performance of the proposed algorithm is more efficient, reliable and accurate than the existing algorithm. Performance parameters such as segmentation accuracy, error rate, and Dice co-efficient are used in order to conduct the evaluation of the proposed work.

关键词： Integrated optics Image segmentation Ant colony optimization Biomedical optical imaging Optical design Image edge detection Metaheuristics

Developing a medical artificial intelligence course for high school students

学校读者我要写书评

暂无评论

Developing a medical artificial intelligence course for high...

International Forum on Medical Imaging in Asia 2021, IFMIA 2021

作者： Huang, Chao-Jung Wu, Tinghui Lu, Jui-Ting Lin, Beatrice Chang, Dawei Wang, Pochuan Wang, Mei-Chi Lee, Peijung Wang, Weichung National Taiwan University-Stanford Joint Program Office of AI in Biotechnology Ministry of Science and Technology Joint Research Center for Artificial Intelligence Technology and All Vista Healthcare Taipei Taiwan Institute of Applied Mathematical Sciences National Taiwan University Taipei Taiwan Department of Mathematics Ecole Normale Supérieure de Lyon Lyon France Department of Computer Science and Information Engineering National Taiwan University Taipei Taiwan Department of Electronic and Computer Engineering National Taiwan University of Science and Technology Taipei Taiwan Department of Mathematics National Taiwan University Taipei Taiwan

ISBN: (数字)9781510644212

ISBN: (纸本)9781510644205

artificial intelligence (AI) training courses often require prerequisites such as calculus or statistics. It is hence challenging to design and develop an introductory AI course for students of secondary education. This research intends to develop a medical AI course, provide high school students with an overview of deep learning applications in medical image analysis, and inspire them to pursue careers in the field of medical AI. We designed a 20-hour course, including lectures and two hands-on projects based on medical image classification. The proposed courses provided medical AI disciplines and built up their knowledge from basic to advanced levels. During the ten-day online courses, all the students were fully engaged and gave us positive feedback. The students endeavored to complete the experimental study in training, testing, and hypothesis of medical images application in the course. Their performance exceeded all expectations, for they did further analysis by tuning different hyperparameters. We designed a course evaluation form, which suggested that the students found it essential and expected to interact with the instructors. The results indicate that combining lectures with hands-on sessions would lead to evidently better achievement in terms of high school students' medical AI knowledge and positive attitudes while addressing real-world problems in the projects. Through this innovative education model, high school students regained their enthusiasm and were encouraged to cultivate their medical AI skills through self-learning while finishing the project. We conclude that this course could be successfully applied to interdisciplinary education in high school. © 2021 SPIE.

关键词： Students

Conditional Generative Modeling for Amorphous Multi-Element Materials

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Li, Honglin Liu, Chuhao Guo, Yongfeng Luo, Xiaoshan Chen, Yijie Liu, Guangsheng Li, Yu Wang, Ruoyu Wang, Zhenyu Wu, Jianzhuo Ma, Cheng Xie, Zhuohang Lv, Jian Ding, Yufei Zhang, Huabin Luo, Jian Zhong, Zhicheng Li, Mufan Wang, Yanchao Li, Wan-Lu Key Laboratory of Material Simulation Methods and Software Ministry of Education College of Physics Jilin University Changchun China Aiiso Yufeng Li Family Department of Chemical and Nano Engineering University of California La Jolla San DiegoCA United States Institute of Molecular Engineering Plus College of Chemistry Fuzhou University Fuzhou China College of Chemistry and Molecular Engineering Peking University Beijing China School of Physics Nankai University Tianjin China Institute of Modern Physics Fudan University Shanghai China Program in Materials Science and Engineering University of California La Jolla San DiegoCA United States School of Artificial Intelligence and Data Science University of Science and Technology of China Hefei230026 China Suzhou Institute for Advanced Research University of Science and Technology of China Suzhou215123 China International Center of Future Science Jilin University Changchun China Department of Computer Science and Engineering University of California La Jolla San DiegoCA United States Center for Renewable Energy and Storage Technologies Physical Science and Engineering Division King Abdullah University of Science and Technology Thuwal Saudi Arabia Suzhou Lab Suzhou215123 China

Amorphous multi-element materials offer unprecedented tunability in composition and properties, yet their rational design remains challenging due to the lack of predictive structure-property relationships and the vast configurational space. Traditional modeling struggles to capture the intricate short-range order that dictates their stability and functionality. We here introduce ApolloX, a pioneering predictive framework for amorphous multi-element materials, establishing a new paradigm by integrating physics-informed generative modeling with particle swarm optimization, using chemical short-range order as an explicit constraint. By systematically navigating the disordered energy landscape, ApolloX enables the targeted design of thermodynamically stable amorphous configurations. It accurately predicts atomic-scale arrangements, including composition-driven metal clustering and amorphization trends, which are well-validated by experiments, while also guiding synthesis by leveraging sluggish diffusion to control elemental distribution and disorder. The resulting structural evolution, governed by composition, directly impacts catalytic performance, leading to improved activity and stability with increasing amorphization. This predictive-experimental synergy transforms the discovery of amorphous materials, unlocking new frontiers in catalysis, energy storage, and functional disordered systems. Copyright © 2025, The Authors. All rights reserved.

关键词： Amorphization

Spatiotemporal dilated convolution with uncertain matching for video-based crowd estimation

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Ma, Yu-Jen Shuai, Hong-Han Cheng, Wen-Huang Department of Electrical and Computer Engineering National Chiao Tung University Hsinchu Taiwan Institute of Electronics National Chiao Tung Univresity Hsinchu300 Taiwan Artificial Intelligence and Data Science Program National Chung Hsing University Taichung400 Taiwan

In this paper, we propose a novel SpatioTemporal convolutional Dense Network (STDNet) to address the video-based crowd counting problem, which contains the decomposition of 3D convolution and the 3D spatiotemporal dilated dense convolution to alleviate the rapid growth of the model size caused by the Conv3D layer. Moreover, since the dilated convolution extracts the multiscale features, we combine the dilated convolution with the channel attention block to enhance the feature representations. Due to the error that occurs from the difficulty of labeling crowds, especially for videos, imprecise or standard-inconsistent labels may lead to poor convergence for the model. To address this issue, we further propose a new patch-wise regression loss (PRL) to improve the original pixel-wise loss. Experimental results on three video-based benchmarks, i.e., the UCSD, Mall and WorldExpo’10 datasets, show that STDNet outperforms both image- and video-based state-of-the-art methods. The source codes are released at https://***/STDNet/STDNet. Copyright © 2021, The Authors. All rights reserved.

关键词： Convolution

TargetNet: Functional microRNA target prediction with deep neural networks

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Min, Seonwoo Lee, Byunghan Yoon, Sungroh Department of Electrical and Computer Engineering Seoul National University Seoul08826 Korea Republic of LG AI Research Seoul07796 Korea Republic of Department of Electronic and IT Media Engineering Seoul National University of Science and Technology Seoul01811 Korea Republic of Interdisciplinary Program in Artificial Intelligence and Interdisciplinary Program in Bioinformatics Seoul National University Seoul08826 Korea Republic of

Motivation: MicroRNAs (miRNAs) play pivotal roles in gene expression regulation by binding to target sites of messenger RNAs (mRNAs). While identifying functional targets of miRNAs is of utmost importance, their prediction remains a great challenge. Previous computational algorithms have major limitations. They use conservative candidate target site (CTS) selection criteria mainly focusing on canonical site types, rely on laborious and time-consuming manual feature extraction, and do not fully capitalize on the information underlying miRNA-CTS interactions. Results: In this paper, we introduce TargetNet, a novel deep learning-based algorithm for functional miRNA target prediction. To address the limitations of previous approaches, TargetNet has three key components: (1) relaxed CTS selection criteria accommodating irregularities in the seed region, (2) a novel miRNA-CTS sequence encoding scheme incorporating extended seed region alignments, and (3) a deep residual network-based prediction model. The proposed model was trained with miRNA-CTS pair datasets and evaluated with miRNA-mRNA pair datasets. TargetNet advances the previous state-of-theart algorithms used in functional miRNA target classification. Furthermore, it demonstrates great potential for distinguishing high-functional miRNA targets. © 2021, CC BY-NC-ND.

关键词： RNA

Sparse Structure Learning via Graph Neural Networks for Inductive Document Classification

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Piao, Yinhua Lee, Sangseon Lee, Dohoon Kim, Sun Department of Computer Science and Engineering Seoul National University Korea Republic of Institute of Computer Technology Seoul National University Korea Republic of Bioinformatics Institute Seoul National University Korea Republic of AIGENDRUG Co. Ltd Interdisciplinary Program in Artificial Intelligence Seoul National University Korea Republic of

Recently, graph neural networks (GNNs) have been widely used for document classification. However, most existing methods are based on static word co-occurrence graphs without sentence-level information, which poses three challenges:(1) word ambiguity, (2) word synonymity, and (3) dynamic contextual dependency. To address these challenges, we propose a novel GNN-based sparse structure learning model for inductive document classification. Specifically, a document-level graph is initially generated by a disjoint union of sentence-level word co-occurrence graphs. Our model collects a set of trainable edges connecting disjoint words between sentences, and employs structure learning to sparsely select edges with dynamic contextual dependencies. Graphs with sparse structures can jointly exploit local and global contextual information in documents through GNNs. For inductive learning, the refined document graph is further fed into a general readout function for graph-level classification and optimization in an end-to-end manner. Extensive experiments on several real-world datasets demonstrate that the proposed model outperforms most state-of-the-art results, and reveal the necessity to learn sparse structures for each document. © 2021, CC BY.

关键词： Graph neural networks