Echocardiography and Cardiac Output Monitoring: Towards Techquity

We explore how echocardiography and cardiac output monitoring are becoming more accessible through AI-enabled ultrasound tools and the recent integration of pulse contour analysis into standard multiparameter bedside monitors.

Echocardiography has become an indispensable diagnostic tool in the intensive care unit (ICU), providing critical insights into cardiac function and systemic haemodynamics (Mayo et al. 2022). Despite this, the use of advanced echocardiography-guided management remains inconsistent, with key barriers including insufficient operator training and restricted access to equipment.

Cardiac output (CO) monitoring also plays a pivotal role in identifying the underlying causes of haemodynamic instability and guiding appropriate therapeutic interventions. Guidelines and expert consensus statements recommend routine CO monitoring in high-risk surgical patients (Scott et al. 2024; Saugel et al. 2025). Yet, only around 10% of patients with significant comorbidities or undergoing major surgery are monitored during the procedure (Ahmad et al. 2015; Molliex et al. 2019). A recent survey of European Society of Anaesthesiology and Intensive Care members revealed that the main reasons for this gap are the limited availability and high cost of CO monitoring technologies (Flick et al. 2023).

Two emerging technological innovations may help overcome these barriers: AI-assisted echocardiography and the integration of pulse contour algorithms into standard multiparameter bedside monitors. Together, these developments have the potential to democratise access to high-quality haemodynamic assessment across diverse care settings.

AI-Guided Echocardiography for Everyone

The non-invasive nature and immediate feedback of echocardiography make it a cornerstone in the rapid assessment and management of critically ill patients presenting with conditions such as circulatory shock, cardiac dysfunction, and respiratory failure. However, its use is operator-dependent. The ability to acquire high-quality diagnostic images and perform precise quantitative assessments requires extensive, specialised training and significant hands-on experience. There is a steep learning curve in advanced echocardiography, even for skilled intensivists. The potential variability in operator skill directly translates to inconsistencies in diagnostic accuracy and reproducibility of measurements, which can undermine clinical decision-making. Compounding this challenge, the high-acuity and time-sensitive nature of the ICU often presents significant time and resource constraints. Decision making needs to be rapid and accurate; the luxury of comprehensive evaluation by dedicated echocardiographists is frequently unavailable. The global scarcity of skilled personnel creates a tangible gap in the optimal delivery of patient care.

Artificial intelligence (AI) is emerging as a transformative solution, providing critical information that might otherwise be unavailable or significantly more challenging to obtain for less experienced operators. Machine learning (ML) algorithms, meticulously trained on vast, diverse datasets of echocardiographic images, are now capable of automating several technically demanding aspects of the ultrasound examination. For instance, AI-enabled tools can autonomously identify and optimise standard ultrasound views, providing real-time guidance to users to enhance image quality (Figure 1). More importantly, they can automatically measure in a few seconds the subaortic velocity time integral (VTI) (Gonzalez et al. 2022), the left ventricular ejection fraction (Varudo et al. 2022), and speckle tracking variables such as the left ventricular global longitudinal strain (Leeson and Fletcher 2021). This sophisticated automation significantly mitigates intra-operator variability and markedly enhances the reproducibility of measurements, even when performed by clinicians with nascent echocardiographic skills (Varudo et al. 2022). The capacity to quickly and accurately quantify parameters like VTI, which is invaluable for assessing fluid responsiveness and calculating cardiac output, brings specialised haemodynamic assessments within routine clinical reach.

The recent COMPASS-AI survey highlighted considerable enthusiasm for AI in POCUS (Wong et al. 2025). A majority of respondents agreed or strongly agreed that AI could improve the speed of diagnosis in the ICU (74%) and substantially reduce inter-operator variability in ultrasound interpretation (89%). Interestingly, the survey also elucidated significant adoption barriers. Training and education emerged as the most frequently cited barrier (27%), underscoring the perceived lack of sufficient training resources and standardised credentialing.

Beyond facilitating the acquisition of diagnostic information, AI also holds potential to train healthcare professionals in performing echocardiography. By functioning as a "virtual tutor," AI-guided systems can provide instantaneous, constructive feedback during image acquisition, prompting users to modify probe position or angulation for achieving optimal views (Figure 1). This interactive and adaptive guidance lowers the entry barrier to performing competent echocardiography, allowing trainees and less experienced clinicians to gain confidence and proficiency where traditional mentorship is infrequent or unavailable. By guiding and enabling the acquisition of high-quality images and accurate measurements, AI tools can empower practitioners to confidently integrate echocardiography into their daily practice, improving their diagnostic acumen. Notably, the COMPASS-AI survey acknowledged legitimate concerns about AI errors potentially leading to incorrect diagnoses (54% agreement) and emphasised the necessity of verifying all AI-generated findings (74% agreement). Nevertheless, the utility of AI in democratising access to echocardiography and serving as a training aid proposes a future where advanced, high-quality echocardiography is potentially accessible for most critical care practitioners.

Perioperative Cardiac Output Monitoring for the Many

Hypotension is common during and after high-risk surgery and is associated with postoperative morbidity and mortality. A recent statement from the European Society of Anaesthesiology and Intensive Care emphasised that "the treatment of hypotension should be based on underlying causes, including vasodilation, hypovolaemia, bradycardia, and cardiac dysfunction" (Saugel et al. 2025). CO monitoring is essential for discriminating between vasodilation, hypovolaemia, and cardiac dysfunction and selecting the appropriate treatment (Michard et al. 2025). Indeed, in patients with hypotension and preserved CO, the underlying mechanism is vasodilation; thus, the logical treatment is the administration of vasopressors and, when possible, a reduction in the depth of anaesthesia. In patients with hypotension and low CO, the condition is often related to hypovolaemia. In this case, fluid administration leads to a significant increase in CO. If this does not occur (fluid non-responder patient), cardiac dysfunction must be suspected.

Many high-risk surgical patients, such as those with significant comorbidities (ASA III-IV) or undergoing major surgery with substantial blood loss, have a radial arterial catheter placed for continuous blood pressure monitoring during and immediately after the procedure. In this context, pulse contour algorithms enable continuous CO monitoring by analysing the arterial pressure waveform. In a 2023 European survey, 92% of anaesthesiologists report pulse contour methods as their preferred approach for CO monitoring during non-cardiac surgery (Flick et al. 2023). However, as previously mentioned, only about one in ten high-risk surgical patients currently benefit from CO monitoring (Ahmad et al. 2015; Molliex et al. 2019), a gap mainly due to limited device availability and high costs (Flick et al. 2023).

Traditionally, CO monitoring required a dedicated haemodynamic system, incurring both capital and recurring costs associated with specialised pressure transducers. In 2023 and in France alone, these recurring costs were estimated to exceed €65 million per year (Michard et al. 2023). They could surpass €1 billion annually across Europe if current recommendations for routine CO monitoring in high-risk surgical patients were widely implemented (Michard et al. 2024).

Among the existing pulse contour algorithms, the PRAM (Pressure Recording Analytical Method) emerges as one of the most reliable for computing stroke volume and CO. Indeed, studies have repeatedly reported strong agreement with clinical reference standards, such as pulmonary thermodilution (Greiwe et al. 2020) and Doppler-echocardiography (Scolletta et al. 2016). The PRAM algorithm also offers the unique advantage of analysing waveforms recorded with any standard low-cost pressure transducer. Recently, PRAM has been integrated into standard multiparameter bedside monitors, enabling continuous CO monitoring on demand by simply tapping the screen (Figure 2). This innovation eliminates the need for external modules or specialised pressure transducers, substantially reducing equipment and consumable costs.

This forward-thinking move by the MedTech industry marks a significant step toward "techquity" and the democratisation of advanced haemodynamic monitoring. By enhancing accessibility and affordability, it not only helps clinicians provide care consistent with current guidelines but also aids in reducing the monitoring's carbon footprint (Michard et al. 2024).

Conclusion

Over the past decade, the miniaturisation and decreasing cost of ultrasound devices have facilitated their clinical adoption. Today, AI-enabled tools enable automatic and reproducible haemodynamic measurements, even in the hands of less experienced clinicians. In parallel, the integration of pulse contour algorithms into standard multiparameter bedside monitors now allows continuous CO monitoring in all patients who have an arterial catheter in place.

Together, these innovations aim to bring clinical practice closer to current expert guidelines, expand access to advanced haemodynamic monitoring, and ultimately reduce disparities in care, a concrete step toward true techquity in perioperative and critical care medicine.

Conflict of Interest

FM is the founder and managing director of MiCo, a consulting and research firm based in Switzerland. AW has received honoraria from GE Healthcare and Mindray for the delivery of educational material.