Published: April 23, 2025
Language: Английский
Medical Image Analysis, Journal Year: 2024, Volume and Issue: 95, P. 103207 - 103207
Published: May 15, 2024
Language: Английский
Citations
38
La radiologia medica, Journal Year: 2023, Volume and Issue: 128(12), P. 1483 - 1496
Published: Sept. 25, 2023
Abstract Objective To investigate the value of Computed Tomography (CT) radiomics derived from different peritumoral volumes interest (VOIs) in predicting epidermal growth factor receptor (EGFR) mutation status lung adenocarcinoma patients. Materials and methods A retrospective cohort 779 patients who had pathologically confirmed were enrolled. 640 randomly divided into a training set, validation an internal testing set (3:1:1), remaining 139 defined as external set. The intratumoral VOI (VOI_I) was manually delineated on thin-slice CT images, seven VOIs (VOI_P) automatically generated with 1, 2, 3, 4, 5, 10, 15 mm expansion along VOI_I. 1454 radiomic features extracted each VOI. t -test, least absolute shrinkage selection operator (LASSO), minimum redundancy maximum relevance (mRMR) algorithm used for feature selection, followed by construction models (VOI_I model, VOI_P model combined model). performance evaluated area under curve (AUC). Results 399 classified EGFR mutant (EGFR+), while 380 wild-type (EGFR−). In sets, VOI4 (intratumoral 4 mm) achieved best predictive performance, AUCs 0.877, 0.727, 0.701, respectively, outperforming VOI_I (AUCs 0.728, 0.698, 0.653, respectively). Conclusions Radiomics region can add extra patients, optimal range mm.
Language: Английский
Citations
29
Computers in Biology and Medicine, Journal Year: 2023, Volume and Issue: 168, P. 107758 - 107758
Published: Nov. 29, 2023
Language: Английский
Citations
28
Deleted Journal, Journal Year: 2025, Volume and Issue: 4(1)
Published: Jan. 1, 2025
Mental health disorders (MHDs) have significant medical and financial impacts on patients society. Despite the potential opportunities for artificial intelligence (AI) in mental field, there are no noticeable roles of these systems real environments. The main reason limitations is lack trust by domain experts decisions AI-based systems. Recently, trustworthy AI (TAI) guidelines been proposed to support building responsible (RAI) that robust, fair, transparent. This review aims investigate literature TAI machine learning (ML) deep (DL) architectures MHD domain. To best our knowledge, this first study analyzes trustworthiness ML DL models identifies advances RAI investigates how related current applicability We discover has severe compared other domains regarding standards implementations. discuss suggest possible future research directions could handle challenges.
Language: Английский
Citations
1
Computerized Medical Imaging and Graphics, Journal Year: 2025, Volume and Issue: 120, P. 102489 - 102489
Published: Jan. 5, 2025
Language: Английский
Citations
1
arXiv (Cornell University), Journal Year: 2022, Volume and Issue: unknown
Published: Jan. 1, 2022
The lack of annotated datasets is a major bottleneck for training new task-specific supervised machine learning models, considering that manual annotation extremely expensive and time-consuming. To address this problem, we present MONAI Label, free open-source framework facilitates the development applications based on artificial intelligence (AI) models aim at reducing time required to annotate radiology datasets. Through researchers can develop AI focusing their domain expertise. It allows readily deploy apps as services, which be made available clinicians via preferred user interface. Currently, Label supports locally installed (3D Slicer) web-based (OHIF) frontends offers two active strategies facilitate speed up segmentation algorithms. make incremental improvements AI-based application by making them other alike. Additionally, provides sample interactive non-interactive labeling applications, used directly off shelf, plug-and-play any given dataset. Significant reduced times using model observed public
Language: Английский
Citations
29
Computerized Medical Imaging and Graphics, Journal Year: 2024, Volume and Issue: 115, P. 102383 - 102383
Published: April 17, 2024
Semi-supervised learning has made significant progress in medical image segmentation. However, existing methods primarily utilize information from a single dimensionality, resulting sub-optimal performance on challenging magnetic resonance imaging (MRI) data with multiple segmentation objects and anisotropic resolution. To address this issue, we present Hybrid Dual Mean-Teacher (HD-Teacher) model hybrid, semi-supervised, multi-task to achieve effective semi-supervised HD-Teacher employs 2D 3D mean-teacher network produce labels signed distance fields the hybrid captured both dimensionalities. This mechanism allows features 2D, 3D, or dimensions as needed. Outputs teacher models are dynamically combined based confidence scores, forming prediction estimated uncertainty. We propose regularization module encourage student results close uncertainty-weighted further improve their feature extraction capability. Extensive experiments of binary multi-class conducted three MRI datasets demonstrated that proposed framework could (1) significantly outperform state-of-the-art (2) surpass fully-supervised VNet trained substantially more annotated data, (3) perform par human raters muscle bone task. Code will be available at https://github.com/ThisGame42/Hybrid-Teacher.
Language: Английский
Citations
8
Radiotherapy and Oncology, Journal Year: 2024, Volume and Issue: 201, P. 110542 - 110542
Published: Sept. 17, 2024
Language: Английский
Citations
6
Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 91, P. 105881 - 105881
Published: Jan. 26, 2024
Language: Английский
Citations
5
Engineering Applications of Artificial Intelligence, Journal Year: 2024, Volume and Issue: 133, P. 108516 - 108516
Published: May 2, 2024
Language: Английский
Citations
4