Reproducible and Interpretable Machine Learning-Based Radiomic Analysis for Overall Survival Prediction in Glioblastoma Multiforme DOI Open Access
Abdulkerim Duman, Xianfang Sun, S Thomas

et al.

Cancers, Journal Year: 2024, Volume and Issue: 16(19), P. 3351 - 3351

Published: Sept. 30, 2024

Purpose: To develop and validate an MRI-based radiomic model for predicting overall survival (OS) in patients diagnosed with glioblastoma multiforme (GBM), utilizing a retrospective dataset from multiple institutions. Materials Methods: Pre-treatment MRI images of 289 GBM were collected. From each patient’s tumor volume, 660 features (RFs) extracted subjected to robustness analysis. The initial prognostic minimum RFs was subsequently enhanced by including clinical variables. final clinical–radiomic derived through repeated three-fold cross-validation on the training dataset. Performance evaluation included assessment concordance index (C-Index), integrated area under curve (iAUC) alongside patient stratification into low high-risk groups (OS). Results: model, which has highest level interpretability, utilized primary gross volume (GTV) one modality (T2-FLAIR) as predictor age variable two independent, robust RFs, achieving moderately good discriminatory performance (C-Index [95% confidence interval]: 0.69 [0.62–0.75]) significant (p = 7 × 10−5) validation cohort. Furthermore, trained exhibited iAUC at 11 months (0.81) literature. Conclusion: We identified validated based OS using multicenter Future work will focus use deep learning-based features, recently standardized convolutional filters tasks.

Language: Английский

MDU-Net: multi-scale densely connected U-Net for biomedical image segmentation DOI
Jiawei Zhang, Yanchun Zhang,

Yuzhen Jin

et al.

Health Information Science and Systems, Journal Year: 2023, Volume and Issue: 11(1)

Published: March 13, 2023

Language: Английский

Citations

85
Radiomics and artificial intelligence for precision medicine in lung cancer treatment DOI Creative Commons
Mitchell Chen, Susan J. Copley, Patrizia Viola

et al.

Seminars in Cancer Biology, Journal Year: 2023, Volume and Issue: 93, P. 97 - 113

Published: May 19, 2023

Lung cancer is the leading cause of cancer-related deaths worldwide. It exhibits, at mesoscopic scale, phenotypic characteristics that are generally indiscernible to human eye but can be captured non-invasively on medical imaging as radiomic features, which form a high dimensional data space amenable machine learning. Radiomic features harnessed and used in an artificial intelligence paradigm risk stratify patients, predict for histological molecular findings, clinical outcome measures, thereby facilitating precision medicine improving patient care. Compared tissue sampling-driven approaches, radiomics-based methods superior being non-invasive, reproducible, cheaper, less susceptible intra-tumoral heterogeneity. This review focuses application radiomics, combined with intelligence, delivering lung treatment, discussion centered pioneering groundbreaking works, future research directions area.

Language: Английский

Citations

76
Deep Learning in Breast Cancer Imaging: State of the Art and Recent Advancements in Early 2024 DOI Creative Commons
Alessandro Carriero, Léon Groenhoff,

Elizaveta Vologina

et al.

Diagnostics, Journal Year: 2024, Volume and Issue: 14(8), P. 848 - 848

Published: April 19, 2024

The rapid advancement of artificial intelligence (AI) has significantly impacted various aspects healthcare, particularly in the medical imaging field. This review focuses on recent developments application deep learning (DL) techniques to breast cancer imaging. DL models, a subset AI algorithms inspired by human brain architecture, have demonstrated remarkable success analyzing complex images, enhancing diagnostic precision, and streamlining workflows. models been applied diagnosis via mammography, ultrasonography, magnetic resonance Furthermore, DL-based radiomic approaches may play role risk assessment, prognosis prediction, therapeutic response monitoring. Nevertheless, several challenges limited widespread adoption clinical practice, emphasizing importance rigorous validation, interpretability, technical considerations when implementing solutions. By examining fundamental concepts synthesizing latest advancements trends, this narrative aims provide valuable up-to-date insights for radiologists seeking harness power care.

Language: Английский

Citations

27
A computed tomography angiography-based radiomics model for prognostic prediction of endovascular abdominal aortic repair DOI

ShihYau Grace Huang,

Dingxiao Liu, Kai Deng

et al.

International Journal of Cardiology, Journal Year: 2025, Volume and Issue: unknown, P. 133138 - 133138

Published: March 1, 2025

Language: Английский

Citations

2
Interstitial lung disease diagnosis and prognosis using an AI system integrating longitudinal data DOI Creative Commons
Xueyan Mei, Zelong Liu, Ayushi Singh

et al.

Nature Communications, Journal Year: 2023, Volume and Issue: 14(1)

Published: April 20, 2023

Abstract For accurate diagnosis of interstitial lung disease (ILD), a consensus radiologic, pathological, and clinical findings is vital. Management ILD also requires thorough follow-up with computed tomography (CT) studies function tests to assess progression, severity, response treatment. However, classification subtypes can be challenging, especially for those not accustomed reading chest CTs regularly. Dynamic models predict patient survival rates based on longitudinal data are challenging create due complexity, variation, irregular visit intervals. Here, we utilize RadImageNet pretrained diagnose five types multimodal transformer model determine patient’s 3-year rate. When history associated CT scans available, the proposed deep learning system help clinicians classify patients and, importantly, dynamically progression prognosis.

Language: Английский

Citations

26
Artificial Intelligence and Interstitial Lung Disease DOI Creative Commons

Ethan Dack,

Andreas Christe, Matthias Fontanellaz

et al.

Investigative Radiology, Journal Year: 2023, Volume and Issue: unknown

Published: April 12, 2023

Abstract Interstitial lung disease (ILD) is now diagnosed by an ILD-board consisting of radiologists, pulmonologists, and pathologists. They discuss the combination computed tomography (CT) images, pulmonary function tests, demographic information, histology then agree on one 200 ILD diagnoses. Recent approaches employ computer-aided diagnostic tools to improve detection disease, monitoring, accurate prognostication. Methods based artificial intelligence (AI) may be used in computational medicine, especially image-based specialties such as radiology. This review summarises highlights strengths weaknesses latest most significant published methods that could lead a holistic system for diagnosis. We explore current AI data use predict prognosis progression ILDs. It essential highlight holds information related risk factors progression, e.g., CT scans tests. aims identify potential gaps, areas require further research, combined yield more promising results future studies.

Language: Английский

Citations

25
AI-Driven Models for Diagnosing and Predicting Outcomes in Lung Cancer: A Systematic Review and Meta-Analysis DOI Open Access
Mohammed Kanan Alshammari, Hajar Alharbi, Nawaf Alotaibi

et al.

Cancers, Journal Year: 2024, Volume and Issue: 16(3), P. 674 - 674

Published: Feb. 5, 2024

(1) Background: Lung cancer's high mortality due to late diagnosis highlights a need for early detection strategies. Artificial intelligence (AI) in healthcare, particularly lung cancer, offers promise by analyzing medical data identification and personalized treatment. This systematic review evaluates AI's performance cancer detection, its techniques, strengths, limitations, comparative edge over traditional methods. (2) Methods: meta-analysis followed the PRISMA guidelines rigorously, outlining comprehensive protocol employing tailored search strategies across diverse databases. Two reviewers independently screened studies based on predefined criteria, ensuring selection of high-quality relevant role detection. The extraction key study details metrics, quality assessment, facilitated robust analysis using R software (Version 4.3.0). process, depicted via flow diagram, allowed meticulous evaluation synthesis findings this review. (3) Results: From 1024 records, 39 met inclusion showcasing AI model applications emphasizing varying strengths among studies. These underscore potential but highlight standardization amidst variations. results demonstrate promising pooled sensitivity specificity 0.87, signifying accuracy identifying true positives negatives, despite observed heterogeneity attributed parameters. (4) Conclusions: demonstrates showing levels However, variations underline standardized protocols fully leverage revolutionizing diagnosis, ultimately benefiting patients healthcare professionals. As field progresses, validated models from large-scale perspective will greatly benefit clinical practice patient care future.

Language: Английский

Citations

13
Artificial Intelligence-Based Treatment Decisions: A New Era for NSCLC DOI Open Access
Oraianthi Fiste, Ioannis Gkiozos, Andriani Charpidou

et al.

Cancers, Journal Year: 2024, Volume and Issue: 16(4), P. 831 - 831

Published: Feb. 19, 2024

Non-small cell lung cancer (NSCLC) is the leading cause of cancer-related mortality among women and men, in developed countries, despite public health interventions including tobacco-free campaigns, screening early detection methods, recent therapeutic advances, ongoing intense research on novel antineoplastic modalities. Targeting oncogenic driver mutations immune checkpoint inhibition has indeed revolutionized NSCLC treatment, yet there still remains unmet need for robust standardized predictive biomarkers to accurately inform clinical decisions. Artificial intelligence (AI) represents computer-based science concerned with large datasets complex problem-solving. Its concept brought a paradigm shift oncology considering its immense potential improved diagnosis, treatment guidance, prognosis. In this review, we present current state AI-driven applications management, particular focus radiomics pathomics, critically discuss both existing limitations future directions field. The thoracic community should not be discouraged by likely long road AI implementation into daily practice, as transformative impact personalized approaches undeniable.

Language: Английский

Citations

13
Exploring non-invasive precision treatment in non-small cell lung cancer patients through deep learning radiomics across imaging features and molecular phenotypes DOI Creative Commons
Xingping Zhang, Guijuan Zhang,

Xingting Qiu

et al.

Biomarker Research, Journal Year: 2024, Volume and Issue: 12(1)

Published: Jan. 25, 2024

Abstract Background Accurate prediction of tumor molecular alterations is vital for optimizing cancer treatment. Traditional tissue-based approaches encounter limitations due to invasiveness, heterogeneity, and dynamic changes. We aim develop validate a deep learning radiomics framework obtain imaging features that reflect various changes, aiding first-line treatment decisions patients. Methods conducted retrospective study involving 508 NSCLC patients from three institutions, incorporating CT images clinicopathologic data. Two radiomic scores network feature were constructed on data sources in the 3D region. Using these features, we developed validated ‘Deep-RadScore,’ model predict prognostic factors, gene mutations, immune molecule expression levels. Findings The Deep-RadScore exhibits strong discrimination features. In independent test cohort, it achieved impressive AUCs: 0.889 lymphovascular invasion, 0.903 pleural 0.894 T staging; 0.884 EGFR ALK, 0.896 KRAS PIK3CA, TP53, 0.895 ROS1; 0.893 PD-1/PD-L1. Fusing yielded optimal predictive power, surpassing any single feature. Correlation interpretability analyses confirmed effectiveness customized capturing additional phenotypes beyond known Interpretation This proof-of-concept demonstrates new biomarkers across can be provided by fusing multiple sources. holds potential offer valuable insights radiological phenotyping characterizing diverse alterations, thereby advancing pursuit non-invasive personalized

Language: Английский

Citations

9
Exploring a decade of deep learning in dentistry: A comprehensive mapping review DOI
Fatemeh Sohrabniya, Sahel Hassanzadeh-Samani,

Seyed AmirHossein Ourang

et al.

Clinical Oral Investigations, Journal Year: 2025, Volume and Issue: 29(2)

Published: Feb. 19, 2025

Language: Английский

Citations

1