Artificial intelligence versus radiologists in detecting early-stage breast cancer from mammograms: a meta-analysis of paradigm shifts

Polish Journal of Radiology, Journal Year: 2025, Volume and Issue: 90, P. 1 - 8

Published: Jan. 21, 2025

Purpose Early detection of breast cancer is crucial for improving patient outcomes. With advancements in artificial intelligence (AI), there growing interest its potential to assist radiologists interpreting mammograms early detection. AI algorithms offer the promise increased accuracy and efficiency identifying subtle signs cancer, potentially complementing expertise enhancing screening process early-stage Material Methods A systematic literature review was conducted identify select original research reports on diagnosis by versus conventional using accordance with PRISMA guidelines. Data were analysed Review Manager version 5.4. <i>P</i>-value <i>I²</i> used test significance differences. Results This meta-analysis included 8 studies data from a total 120,950 patients. Regarding sensitivity AI, pooled analysis 6 sensitivities ranging 0.70 0.89 yielded 0.85. However, ranged 0.63 0.85, an overall 0.77. As specificity, both groups had closer results. Conclusions The comparison between systems detecting highlights as valuable tool screening. While have shown promising results terms efficiency, they should be viewed complementary rather than replacements.

Language: Английский

---

A critical assessment of artificial intelligence in magnetic resonance imaging of cancer

npj Imaging, Journal Year: 2025, Volume and Issue: 3(1)

Published: April 9, 2025

Given the enormous output and pace of development artificial intelligence (AI) methods in medical imaging, it can be challenging to identify true success stories determine state-of-the-art field. This report seeks provide magnetic resonance imaging (MRI) community with an initial guide into major areas which AI are contributing MRI oncology. After a general introduction intelligence, we proceed discuss successes current limitations when used for image acquisition, reconstruction, registration, segmentation, as well its utility assisting diagnostic prognostic settings. Within each section, attempt present balanced summary by first presenting common techniques, state readiness, clinical needs, barriers practical deployment setting. We conclude new advances must realized address questions regarding generalizability, quality assurance control, uncertainty quantification applying cancer maintain patient safety utility.

Language: Английский

---

Artificial intelligence utilization in cancer screening program across ASEAN: a scoping review

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

Published: April 15, 2025

Cancer remains a significant health challenge in the ASEAN region, highlighting need for effective screening programs. However, approaches, target demographics, and intervals vary across member states, necessitating comprehensive understanding of these variations to assess program effectiveness. Additionally, while artificial intelligence (AI) holds promise as tool cancer screening, its utilization region is unexplored. This study aims identify evaluate different programs ASEAN, with focus on assessing integration impact AI A scoping review was conducted using PRISMA-ScR guidelines provide overview usage ASEAN. Data were collected from government ministries, official guidelines, literature databases, relevant documents. The use reviews involved searches through PubMed, Scopus, Google Scholar inclusion criteria only included studies that utilized data January 2019 May 2024. findings reveal diverse approaches Countries like Myanmar, Laos, Cambodia, Vietnam, Brunei, Philippines, Indonesia Timor-Leste primarily adopt opportunistic Singapore, Malaysia, Thailand organized Cervical widespread, both methods. Fourteen review, covering breast (5 studies), cervical (2 colon (4 hepatic (1 study), lung oral study) cancers. Studies revealed stages screening: prospective clinical evaluation (50%), silent trial (36%) exploratory model development (14%), promising results enhancing accuracy efficiency. require more targeting appropriate age groups at regular meet WHO's 2030 targets. Efforts integrate Thailand, show optimizing processes, reducing costs, improving early detection. technology enhances identification during detection management region.

Language: Английский

---

Benchmarking Deep Learning Algorithms for Breast Cancer Detection: A Comprehensive Review and Evaluation Across Public Imaging Datasets

InfoScience Trends, Journal Year: 2025, Volume and Issue: 2(4), P. 11 - 24

Published: April 14, 2025

Language: Английский

---

Ethical and Operational Approaches for Addressing Gender Bias in AI Health Technology in African Contexts

Oxford University Press eBooks, Journal Year: 2025, Volume and Issue: unknown

Published: March 20, 2025

Abstract An urgent need exists to define comprehensive ethical and operational frameworks designed address gender bias in artificial intelligence–driven health applications within unique healthcare environments of Africa. These would the critical challenge disparities that exist AI systems, which can worsen outcomes for women. This article outlines proactive approaches mitigate deployed African settings. It expands on complexities surrounding consent deployment highlights transparency about AI’s role patient care strict data governance policies are sensitive vulnerabilities The examines key challenges related technologies how existing may not be suited context. With this background, it advocates gender-sensitive gender-transformative design frameworks, mandating multidisciplinary teams with studies, Afro-feminist expertise. Additionally, collected from FemTech Africa offer complementary solutions by integrating into systems improve accuracy inclusivity. also draws attention gender-differentiated datasets ensure algorithms trained context-specific data. further proposes establishment oversight committees strong representation deep expertise ethics. enforce compliance standards promote accountability. If efforts made biases implementation these have potential contribute improved However, impact depends responsible development AI, robust governance, consideration local contexts.

Language: Английский

---

Exploring the Impact of Artificial Intelligence on Patient Care: A Comprehensive Review of Healthcare Advancements

Scholars Academic Journal of Pharmacy, Journal Year: 2024, Volume and Issue: 13(02), P. 67 - 81

Published: Feb. 28, 2024

Artificial Intelligence (AI) is revolutionizing healthcare by transforming disease identification, treatment, and management. Healthcare organizations are rapidly adopting AI technologies to improve patient outcomes, streamline operations, optimize costs. Utilizing a broad toolkit comprising Robotics, Computer Vision, Natural Language Processing, Machine Learning, has made significant advancements across various domains. AI-driven diagnostic systems showcased for their precision in analyzing medical images, enabling early detection of diseases such as cancer. Personalized treatment plans preventive treatments possible predictive analytics, which uses large amounts data predict the course identify those who at risk. This leads an improvement care. Beyond clinical applications, reshaping delivery through solutions like telemedicine, virtual consultations, remote monitoring. Virtual Health Assistants, empowered AI, deliver personalized health information, medication reminders, lifestyle guidance, enhancing engagement adherence. Telemedicine employ algorithms enhance resource allocation, expedite appointment scheduling, supply superior services isolated populations. Hence, AI’s potential productivity, encourage creativity, solve difficult problems sophisticated analysis automation what it so important many sectors.

Language: Английский

---

Application of Artificial Intelligence (AI) System in Opportunistic Screening and Diagnostic Population in a Middle-income Nation

Current Medical Imaging Formerly Current Medical Imaging Reviews, Journal Year: 2024, Volume and Issue: 20

Published: March 1, 2024

Objective:: This study evaluates the effectiveness of artificial intelligence (AI) in mammography a diverse population from middle-income nation and compares it to traditional methods. Methods:: A retrospective was conducted on 543 mammograms 467 Malays, 48 Chinese, 28 Indians nation. Three breast radiologists interpreted examinations independently two reading sessions (with without AI support). Breast density BI-RADS categories were assessed, comparing accuracy, sensitivity, specificity, positive predictive value (PPV), negative (NPV) results. Results:: Of mammograms, 69.2% had lesions detected. Biopsies performed 25%(n=136), with 66(48.5%) benign 70(51.5%) malignant. Substantial agreement assessment between radiologist software (κ =0.606, p < 0.001) category =0.74, 0.001). The performance comparable PPV, NPV or alone, + AI, alone 81.9%,90.4%,56.0%, 97.1%; 81.0%, 93.1%,55.5%, 97.0%; 90.0%,76.5%,36.2%, 98.1%, respectively. enhances accuracy lesion diagnosis reduces unnecessary biopsies, particularly for 4 lesions. results synthetic almost similar original 2D mammography, AUC 0.925 0.871, Conclusion:: may assist accurate lesions, enhancing efficiency mixed opportunistic screening diagnostic patients. Key Messages:: • use population-based cancer has been validated high-income nations, reported improved performance. Our evaluated usage an tool setting multi-ethnic application potentially leading reduced biopsies. integration into workflow did not disrupt trained radiologists, as there is substantial inter-reader density.

Language: Английский

---

DCE-MRI Radiomic analysis in triple negative ductal invasive breast cancer. Comparison between BRCA and not BRCA mutated patients: Preliminary results

Magnetic Resonance Imaging, Journal Year: 2024, Volume and Issue: 113, P. 110214 - 110214

Published: July 22, 2024

The research aimed to determine whether and which radiomic features from breast dynamic contrast enhanced (DCE) MRI could predict the presence of BRCA1 mutation in patients with triple-negative cancer (TNBC). This retrospective study included consecutive histologically diagnosed TNBC who underwent DCE-MRI 2010–2021. Baseline DCE-MRIs were retrospectively reviewed; percentage maps wash-in wash-out computed lesions manually segmented, drawing a 5 mm-Region Interest (ROI) inside tumor another mm-ROI contralateral healthy gland. Features for each map ROI extracted Pyradiomics-3D Slicer considered first separately (tumor gland) then together. In analysis more important status classification selected Maximum Relevance Minimum Redundancy algorithm used fit four classifiers. population 67 86 (21 BRCA1-mutated, 65 non BRCA-carriers). best classifiers BRCA Support Vector Classifier Logistic Regression models fitted both gland features, reaching an Area Under ROC Curve (AUC) 0.80 (SD 0.21) 0.79 0.20), respectively. Three higher BRCA1-mutated compared BRCA-mutated: Total Energy Correlation gray level cooccurrence matrix, measured maps, Root Mean Squared, tumor. showed feasibility potential radiomics predicting mutational status.

Language: Английский

---

Сontemporary Medical Decision Support Systems Based on Artificial Intelligence for the Analysis of Digital Mammographic Images

Journal of radiology and nuclear medicine, Journal Year: 2023, Volume and Issue: 104(2), P. 151 - 162

Published: Aug. 7, 2023

The relevance of implementing artificial intelligence (AI) technologies in the diagnosis breast cancer (BC) is associated with a continuing high increase BC incidence among women and its leading position structure incidence. Theoretically, using AI possible both at stage screening clarifying diagnosis. article provides brief overview systems used clinical practice discusses their prospects Advances machine learning can be effective to improve accuracy mammography by reducing missed cases false positives.

Language: Английский

---

Exploring the Potential of Artificial Intelligence in Breast Ultrasound

Critical Reviews™ in Oncogenesis, Journal Year: 2023, Volume and Issue: 29(2), P. 15 - 28

Published: Sept. 18, 2023

Breast ultrasound has emerged as a valuable imaging modality in the detection and characterization of breast lesions, particularly women with dense tissue or contraindications for mammography. Within this framework, artificial intelligence (AI) garnered significant attention its potential to improve diagnostic accuracy revolutionize workflow. This review article aims comprehensively explore current state research development harnessing AI's capabilities ultrasound. We delve into various AI techniques, including machine learning, deep well their applications automating lesion detection, segmentation, classification tasks. Furthermore, addresses challenges hurdles faced implementing systems diagnostics, such data privacy, interpretability, regulatory approval. Ethical considerations pertaining integration clinical practice are also discussed, emphasizing importance maintaining patient-centered approach. The holds great promise improving accuracy, enhancing efficiency, ultimately advancing patient's care. By examining identifying future opportunities, contribute understanding utilization encourage further interdisciplinary collaboration maximize practice.

Language: Английский

---