What's Hot in POCUS this May and June!
- g&h CritCareEcho
- Jun 28
- 8 min read
Date: 28th June 2025
Welcome to our May-June Edition of "Whats Hot in POCUS"!
As summer approaches and we find ourselves midway through 2025, we're delighted to bring you another comprehensive roundup of the most compelling developments in critical care ultrasound.
This edition showcases the remarkable evolution of our field, from groundbreaking AI applications that are transforming how we acquire and interpret images, to innovative training methodologies that are reshaping POCUS education across specialties.
We've curated ten essential papers that span the breadth of contemporary POCUS practice - from enhanced lung ultrasound scoring systems and cardiac arrest protocols, to the expanding role of artificial intelligence and evolving training paradigms.
Let's explore what's shaping the future of point-of-care ultrasound!
1. Building a Point of Care Ultrasound (POCUS) Curriculum in Undergraduate Medical Education Through Stepwise Development and Assessment
🌐 Context and Background Despite widespread recognition that POCUS should be incorporated into undergraduate medical education, implementation remains challenging due to limited resources, lack of standardised curricula, and insufficient trained faculty. Most medical schools struggle to develop effective training programmes that meaningfully improve both knowledge and confidence.
🔍 Key Developments and Insights Hagood et al. present a compelling three-year stepwise approach to POCUS curriculum development during Internal Medicine clerkships. Their progressive model evolved from simulation-only training (which improved confidence but not knowledge) to resident-led workshops, and finally to faculty-led lectures combined with hands-on training. The final iteration achieved significant improvements, with overall knowledge scores improving from 49.9% to 66.7% (p<0.0001) amongst 102 students.
💡 Impact and Significance This study provides a practical blueprint for medical schools seeking to implement sustainable POCUS curricula with limited faculty resources. The stepwise methodology demonstrates how resident teachers can effectively bridge the gap between faculty availability and educational demand, whilst the rigorous assessment framework offers measurable outcomes for programme evaluation and improvement.
⚠️ Limitations The single-centre design limits generalisability, and the study relied on voluntary participation in surveys, potentially underestimating intervention impact. Variable resident expertise during workshops represents a confounding factor, and the assessment focused primarily on knowledge and confidence rather than practical skills acquisition, which remains challenging to evaluate systematically.
2. Seeing Ghosts: A Quality Improvement Intervention to Decrease Phantom Scanning Through Increased Image Archiving
🌐 Context and Background "Phantom scanning" - performing POCUS without saving representative images - represents a significant barrier to quality assurance and education in ultrasound training programmes.
🔍 Key Developments and Insights Kolbenson et al. implemented a quality improvement project using Plan-Do-Study-Act cycles to increase image archiving by internal medicine residents. Through education sessions and reminder systems, they achieved a dramatic increase from 0% to 76% of scans being archived, with 94% of residents saving at least one scan during the study period.
💡 Impact and Significance This intervention demonstrates practical, scalable methods for improving POCUS culture and establishing sustainable image archiving practices essential for quality assurance and educational feedback in residency programmes.
⚠️ Limitations The single-centre study design limits generalisability to other training programmes. Data collection relied on sign-out sheets which may introduce inaccuracies, and the intervention required ongoing reminders suggesting the need for sustained effort to maintain compliance.
3. A Theory-Informed Approach to Identify Barriers to Utilising Point-of-Care Ultrasound (POCUS) in Practice
🌐 Context and Background Despite widespread POCUS training, many clinicians fail to incorporate ultrasound skills into routine practice, leading to skill decay and missed opportunities to benefit patients.
🔍 Key Developments and Insights Hofmann et al. conducted a theory-informed qualitative case study of Focused Intensive Care Echo (FICE) utilisation, identifying how barriers form "vicious cycles" that perpetuate non-use. These cycles relate to enthusiasm, opportunity, support, participation, communication, and institutional norms, manifesting an underlying tension between POCUS training and patient care priorities.
💡 Impact and Significance This research provides theoretical framework for understanding implementation barriers and suggests low-resource mechanisms (variation, noticing, and "powering up" support) that could generate scalable solutions for improving POCUS utilisation post-training.
⚠️ Limitations The study was conducted in a single specialist heart and lung hospital, potentially limiting applicability to other healthcare settings. The focus on FICE may not translate directly to other POCUS applications or healthcare systems with different organisational structures.
4. AI Assisted Focused Cardiac Ultrasound in Preventive Cardiology
🌐 Context and Background Current cardiovascular risk prediction tools have limitations in sensitivity and accuracy across diverse populations, creating opportunities for enhanced risk stratification through novel approaches.
🔍 Key Developments and Insights Cohen et al. propose a framework integrating AI-assisted focused cardiac ultrasound (FoCUS) for preventive cardiology applications. The approach encompasses biological cardiac age assessment, valvular heart disease detection, occult atrial fibrillation screening, heart failure identification, and pulmonary hypertension assessment through automated AI-enhanced workflow.
💡 Impact and Significance This perspective envisions transformative potential for AI-enhanced cardiac POCUS in preventive medicine, potentially democratising cardiovascular screening and enabling earlier intervention through enhanced diagnostic accessibility and accuracy.
⚠️ Limitations As a perspective piece rather than empirical research, the proposed framework lacks clinical validation data. Implementation challenges, cost-effectiveness considerations, and regulatory pathways for AI integration into routine clinical practice are not adequately addressed.
5. Contemporary Applications of Artificial Intelligence and Machine Learning in Echocardiography
🌐 Context and Background Echocardiography faces ongoing challenges including inter-observer variability, time-consuming interpretation, and the need for expert training, creating opportunities for AI-driven solutions.
🔍 Key Developments and Insights Raissi-Dehkordi et al. provide a comprehensive review of AI applications in echocardiography, covering view classification, image segmentation, automated quantification, and disease prediction. The review encompasses LVEF assessment, valvular disease evaluation, and cardiomyopathy detection, highlighting both supervised and unsupervised learning approaches.
💡 Impact and Significance This review demonstrates the broad potential for AI to enhance diagnostic consistency, reduce interpretation time, and improve accuracy across diverse echocardiographic applications, whilst identifying key limitations requiring further research.
⚠️ Limitations As a narrative review, potential selection bias in literature inclusion exists. Many described AI algorithms require validation in larger, diverse populations before clinical implementation, and practical integration challenges remain underexplored.
6. Efficient Lung Ultrasound Severity Scoring Using Dedicated Feature Extractor
🌐 Context and Background AI-based lung ultrasound (LUS) severity scoring faces challenges from limited annotated datasets and the risk of information loss during frame-level feature aggregation for video-level assessment.
🔍 Key Developments and Insights Guo et al. propose MeDiVLAD, a novel pipeline using self-knowledge distillation to pre-train a vision transformer without labels, combined with dual-level VLAD aggregation for multi-level LUS severity scoring. The method achieved 82.47% accuracy with minimal fine-tuning and 82.60% accuracy at video-level classification.
💡 Impact and Significance This approach significantly reduces reliance on expert annotations whilst demonstrating superior performance compared to conventional supervised methods, potentially enabling automated LUS scoring in resource-limited settings with minimal training data.
⚠️ Limitations The study utilised a relatively small dataset (283 videos) with extreme class imbalance requiring score combination. Validation was limited to COVID-19 related lung pathology, potentially limiting generalisability to other respiratory conditions.
7. Point-of-Care Ultrasonography in the Critical Care Unit: An Update
🌐 Context and Background Critical care POCUS continues evolving rapidly with technological advances, new applications, and changing regulatory requirements for training programmes.
🔍 Key Developments and Insights Guevarra and Greenstein provide a comprehensive update on POCUS in critical care, highlighting AI integration for improved image acquisition and interpretation, emerging applications during cardiac arrest, VExUS scoring developments, and new ACGME mandates requiring POCUS competency in critical care fellowship training.
💡 Impact and Significance This update reinforces POCUS as essential in critical care whilst identifying key areas for development including AI-assisted quantitative assessments and the need for robust training programmes to meet new competency requirements.
⚠️ Limitations As a narrative review, systematic methodology for literature selection is lacking. The authors note limited robust research supporting VExUS integration into clinical practice, cautioning against premature adoption without stronger evidence.
8. Current Use, Training, and Barriers to Point-of-Care Ultrasound Use Across Multiple Specialties
🌐 Context and Background Understanding POCUS adoption patterns across medical specialties can inform strategic infrastructure investments and address common implementation barriers.
🔍 Key Developments and Insights Resop et al. conducted a comprehensive cross-sectional survey across all Veterans Affairs medical centres, comparing POCUS use, training needs, and barriers across emergency medicine, critical care, hospital medicine, anaesthesiology, and surgery. The study revealed significant increases in both POCUS use and training desires from 2015-2020, with training and infrastructure being the most common barriers.
💡 Impact and Significance This multi-specialty analysis provides valuable insights for healthcare systems planning POCUS infrastructure investments, highlighting the need for tailored training approaches and systematic barrier resolution across diverse clinical environments.
⚠️ Limitations Findings are limited to the VA healthcare system, potentially restricting generalisability. Data were self-reported by specialty chiefs rather than individual practitioners, and the survey ended early due to COVID-19, potentially affecting completeness.
9. The Expanding Point of Care Ultrasound (POCUS) Paradigm
🌐 Context and Background Traditional POCUS has been conceptualised through a binary lens focused on answering simple yes/no clinical questions, but the field is evolving toward more sophisticated applications.
🔍 Key Developments and Insights Wiskar advocates for an expanded POCUS paradigm moving beyond dichotomous decision-making to encompass complex, multi-organ assessments addressing nuanced clinical questions about haemodynamics, fluid management, and respiratory failure. This approach integrates POCUS data with other clinical parameters through Bayesian reasoning.
💡 Impact and Significance This paradigm shift positions POCUS as a comprehensive clinical tool capable of addressing complex clinical scenarios, expanding utility beyond simple rule-in/rule-out applications whilst requiring sophisticated operator understanding.
⚠️ Limitations As a perspective piece, the article presents expert opinion rather than empirical evidence. The expanded model introduces complexity and potential for misinterpretation, requiring substantial training in anatomic and physiologic principles that may not be feasible for all practitioners.
10. Echo-E³Net: Efficient Endo-Epi Spatio-Temporal Network for Ejection Fraction Estimation
🌐 Context and Background Left ventricular ejection fraction (LVEF) estimation remains time-consuming and operator-dependent, whilst many existing deep learning models are computationally demanding, hindering real-time clinical applications.
🔍 Key Developments and Insights Hedari et al. developed Echo-E³Net, an efficient spatio-temporal network for LVEF estimation featuring the Endo-Epi Cardial Border Detector (E²CBD) module and Endo-Epi Feature Aggregator (E²FA). The system achieved state-of-the-art performance with RMSE of 5.15 and R² of 0.82, using only 6.8 million parameters and 8.49G FLOPs, making it suitable for real-time point-of-care applications.
💡 Impact and Significance This lightweight yet highly accurate approach addresses the critical need for efficient LVEF estimation in resource-constrained POCUS environments, operating without pre-training or extensive data augmentation whilst maintaining clinical accuracy.
⚠️ Limitations Evaluation was conducted on a single dataset (EchoNet-Dynamic), and clinical validation in diverse patient populations and real-world POCUS settings is needed. The model's performance across different ultrasound systems and image qualities requires further investigation.
Conclusion
This May-June edition highlights the remarkable transformation occurring in point-of-care ultrasound, driven by artificial intelligence integration, standardised education initiatives, and evidence-based implementation strategies.
We're witnessing the emergence of AI as a clinical partner in POCUS — from automated image acquisition guidance to sophisticated diagnostic algorithms detecting subtle pathologies. Simultaneously, the field is maturing through standardised curricula and evidence-based training approaches that promise to democratise ultrasound skills across healthcare professions.
Key themes emerging from this edition include the critical importance of addressing implementation barriers post-training, the need for sustainable image archiving cultures, and the expanding paradigm of POCUS applications beyond traditional binary decision-making.
The integration of these technological advances with robust educational frameworks and systematic barrier resolution positions POCUS at the forefront of modern medical practice, offering unprecedented opportunities to improve patient care through enhanced diagnostic capabilities and broader accessibility.
Don't miss the next edition! Subscribe to our free monthly newsletter to stay up to date with the latest POCUS research: ghcritcareecho.co.uk
Have suggestions for papers you'd like us to cover? Drop us a message!
About the writer
Hannah Conway, a clinical-academic and National FUSIC Heart Lead for the UK.
Interests lie in PoCUS education, Echocardiography, RV injury and telemedicine
Follow me on Twitter/X for more PoCUS related educational content https://x.com/cardiacaccp
Do you want to learn echocardiography and how to apply this skill clinically?
Check out our upcoming courses below ⬇️
We have limited spaces on both courses - so register for your space today!
Comentarios