CXR-LLaVA: a multimodal large language model for interpreting chest X-ray images

European Radiology, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 15, 2025

This study aimed to develop an open-source multimodal large language model (CXR-LLaVA) for interpreting chest X-ray images (CXRs), leveraging recent advances in models (LLMs) potentially replicate the image interpretation skills of human radiologists. For training, we collected 592,580 publicly available CXRs, which 374,881 had labels certain radiographic abnormalities (Dataset 1) and 217,699 provided free-text radiology reports 2). After pre-training a vision transformer with Dataset 1, integrated it LLM influenced by LLaVA network. Then, was fine-tuned, primarily using 2. The model's diagnostic performance major pathological findings evaluated, along acceptability radiologic radiologists, gauge its potential autonomous reporting. demonstrated impressive test sets, achieving average F1 score 0.81 six MIMIC internal set 0.56 external set. scores surpassed those GPT-4-vision Gemini-Pro-Vision both sets. In radiologist evaluations set, achieved 72.7% success rate reporting, slightly below 84.0% ground truth reports. highlights significant LLMs CXR interpretation, while also acknowledging limitations. Despite these challenges, believe that making our will catalyze further research, expanding effectiveness applicability various clinical contexts. Question How can be adapted interpret X-rays generate reports? Findings developed CXR-LLaVA effectively detects generates higher accuracy compared general-purpose models. Clinical relevance demonstrates support radiologists autonomously generating reports, reducing workloads improving efficiency.

Language: Английский

---

Evaluating pre-processing and deep learning methods in medical imaging: Combined effectiveness across multiple modalities

Alexandria Engineering Journal, Journal Year: 2025, Volume and Issue: 119, P. 558 - 586

Published: Feb. 10, 2025

Language: Английский

---

Graph embedding dimensionality reduction combined with improved APO optimized kELM for pneumonia recognition

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 108, P. 107909 - 107909

Published: April 22, 2025

Language: Английский

---

Reconstruction-based approach for chest X-ray image segmentation and enhanced multi-label chest disease classification

Artificial Intelligence in Medicine, Journal Year: 2025, Volume and Issue: 165, P. 103135 - 103135

Published: April 23, 2025

U-Net is a commonly used model for medical image segmentation. However, when applied to chest X-ray images that show pathologies, it often fails include these critical pathological areas in the generated masks. To address this limitation, our study, we tackled challenge of precise segmentation and mask generation by developing novel approach, using CycleGAN, encompasses affected pathologies within region interest, allowing extraction relevant radiomic features linked pathologies. Furthermore, adopted feature selection approach focus analysis on most significant features. The results proposed pipeline are promising, with an average accuracy 92.05% AUC 89.48% multi-label classification effusion infiltration acquired from ChestX-ray14 dataset, XGBoost model. applying methodology 14 diseases dataset resulted 83.12%, outperforming previous studies. This research highlights importance effective accurate diseases. promising underscore its potential broader applications

Language: Английский

---

Lung Diseases Classification Using Artificial Neural Network With Ant Colony Optimization

Advances in computational intelligence and robotics book series, Journal Year: 2025, Volume and Issue: unknown, P. 223 - 248

Published: March 7, 2025

The classification of medical data is the most difficult problem to solve among all research problems since it has more commercial significance in context health analytics. Several researchers have looked into using Artificial Intelligence (AI) for lung disease classification. This paper proposed a novel algorithm diagnosis various diseases. Already known existing algorithms some drawback noise removal and process. In this approach, better technique used remove unwanted noises input image. Hybridization Neural Network with Ant Colony Optimization based predict accurate obtain efficiency. suggested HANNACO was evaluated qualitatively obtained 95.30% accuracy, 93.72% minimum time duration 18 ms over current approaches such as Decision Tree, SVM, KNN.

Language: Английский

---

Deep learning-based computational approach for predicting ncRNAs-disease associations in metaplastic breast cancer diagnosis

BMC Cancer, Journal Year: 2025, Volume and Issue: 25(1)

Published: May 6, 2025

Non-coding RNAs (ncRNAs) play a crucial role in breast cancer progression, necessitating advanced computational approaches for precise disease classification. This study introduces Deep Reinforcement Learning (DRL)-based framework predicting ncRNA-disease associations metaplastic (MBC) using multi-dimensional descriptor system (ncRNADS) integrating 550 sequence-based features and 1,150 target gene descriptors (miRDB score ≥ 90). The model achieved 96.20% accuracy, 96.48% precision, 96.10% recall, 96.29% F1-score, outperforming traditional classifiers such as support vector machines (SVM) neural networks. Feature selection optimization reduced dimensionality by 42.5% (4,430 to 2,545 features) while maintaining high demonstrating efficiency. External validation confirmed specificity subtypes (87-96.5% accuracy) minimal cross-reactivity with unrelated diseases like Alzheimer's (8-9% accuracy), ensuring robustness. SHAP analysis identified key sequence motifs (e.g., "UUG") structural free energy (ΔG = - 12.3 kcal/mol) critical predictors, validated PCA (82% variance) t-SNE clustering. Survival TCGA data revealed prognostic significance MALAT1, HOTAIR, NEAT1 (associated poor survival, HR 1.76-2.71) GAS5 (protective effect, 0.60). DRL demonstrated rapid training (0.08 s/epoch) cloud deployment compatibility, underscoring its scalability large-scale applications. These findings establish ncRNA-driven classification cornerstone precision oncology, enabling patient stratification, survival prediction, therapeutic identification MBC.

Language: Английский

---

Ensemble of Deep Learning Architectures with Machine Learning for Pneumonia Classification Using Chest X-rays

Deleted Journal, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 13, 2024

Pneumonia is a severe health concern, particularly for vulnerable groups, needing early and correct classification optimal treatment. This study addresses the use of deep learning combined with machine classifiers (DLxMLCs) pneumonia from chest X-ray (CXR) images. We deployed modified VGG19, ResNet50V2, DenseNet121 models feature extraction, followed by five (logistic regression, support vector machine, decision tree, random forest, artificial neural network). The approach we suggested displayed remarkable accuracy, VGG19 obtaining 99.98% accuracy when forest or tree classifiers. ResNet50V2 achieved 99.25% forest. These results illustrate advantages merging in boosting speedy accurate identification pneumonia. underlines potential DLxMLC systems enhancing diagnostic efficiency. By integrating these into clinical practice, healthcare practitioners could greatly boost patient care results. Future research should focus on refining exploring their application to other medical imaging tasks, as well including explainability methodologies better understand decision-making processes build trust use. technique promises promising breakthroughs management.

Language: Английский

---

Advanced Diagnosis of Cardiac and Respiratory Diseases from Chest X-Ray Imagery Using Deep Learning Ensembles

Journal of Sensor and Actuator Networks, Journal Year: 2025, Volume and Issue: 14(2), P. 44 - 44

Published: April 18, 2025

Chest X-ray interpretation is essential for diagnosing cardiac and respiratory diseases. This study introduces a deep learning ensemble approach that integrates Convolutional Neural Networks (CNNs), including ResNet-152, VGG19, EfficientNet, Vision Transformer (ViT), to enhance diagnostic accuracy. Using the NIH dataset, methodology involved comprehensive preprocessing, data augmentation, model optimization techniques address challenges such as label imbalance feature variability. Among individual models, VGG19 exhibited strong performance with Hamming Loss of 0.1335 high accuracy in detecting Edema, while ViT excelled classifying certain conditions like Hernia. Despite strengths meta-model achieved best overall performance, 0.1408 consistently higher ROC-AUC values across multiple diseases, demonstrating its superior capability handle complex classification tasks. robust framework underscores potential reliable precise disease detection, offering significant improvements over traditional methods. The findings highlight value integrating diverse architectures complexities multi-label chest classification, providing pathway more accurate, scalable, accessible tools clinical practice.

Language: Английский

---

A Scoping Review of the Use of Blockchain and Machine Learning in Medical Imaging Applications

Lecture notes in networks and systems, Journal Year: 2024, Volume and Issue: unknown, P. 107 - 117

Published: Jan. 1, 2024

Language: Английский

---

Advances in medical image analysis: A comprehensive survey of lung infection detection

IET Image Processing, Journal Year: 2024, Volume and Issue: 18(13), P. 3750 - 3800

Published: Oct. 10, 2024

Abstract This research investigates advanced approaches in medical image analysis, specifically focusing on segmentation and classification techniques, as well their integration into multi‐task architectures for lung infections. begins by explaining key architectural models used tasks. The study extends to the enhancement of these through attention modules conditional random fields. Relevant datasets evaluation metrics, incorporating discussions loss functions are also reviewed. review encompasses recent advancements single‐task models, highlighting innovations semi‐supervised, self‐supervised, few‐shot, zero‐shot learning techniques. Empirical analysis is conducted both architectures, predominantly utilizing U‐Net framework, applied across multiple Results demonstrate effectiveness provide insights strengths limitations different approaches. contributes improved detection diagnosis infections offering a comprehensive overview current methodologies practical applications.

Language: Английский

---