Better With Ultrasound: Transesophageal Echocardiography

Deep Learning Based Automatic Left Ventricle Segmentation from the Transgastric Short-Axis View on Transesophageal Echocardiography: A Feasibility Study

Segmenting the left ventricle from the transgastric short-axis views (TSVs) on transesophageal echocardiography (TEE) is the cornerstone for cardiovascular assessment during perioperative management. Even for seasoned professionals, the procedure remains time-consuming and experience-dependent. The current study aims to evaluate the feasibility of deep learning for automatic segmentation by assessing the validity of different U-Net algorithms. A large dataset containing 1388 TSV acquisitions was retrospectively collected from 451 patients (32% women, average age 53.42 years) who underwent perioperative TEE between July 2015 and October 2023. With image preprocessing and data augmentation, 3336 images were included in the training set, 138 images in the validation set, and 138 images in the test set. Four deep neural networks (U-Net, Attention U-Net, UNet++, and UNeXt) were employed for left ventricle segmentation and compared in terms of the Jaccard similarity coefficient (JSC) and Dice similarity coefficient (DSC) on the test set, as well as the number of network parameters, training time, and inference time. The Attention U-Net and U-Net++ models performed better in terms of JSC (the highest average JSC: 86.02%) and DSC (the highest average DSC: 92.00%), the UNeXt model had the smallest network parameters (1.47 million), and the U-Net model had the least training time (6428.65 s) and inference time for a single image (101.75 ms). The Attention U-Net model outperformed the other three models in challenging cases, including the impaired boundary of left ventricle and the artifact of the papillary muscle. This pioneering exploration demonstrated the feasibility of deep learning for the segmentation of the left ventricle from TSV on TEE, which will facilitate an accelerated and objective alternative of cardiovascular assessment for perioperative management.

Point-of-Care Echocardiography in the Difficult-to-Image Patient in the ICU: A Narrative Review

OBJECTIVES

The objective of this narrative review was to address common obstacles encountered in the ICU to acquiring quality and interpretable images using point-of-care echocardiography.

DATA SOURCES

Detailed searches were performed using PubMed and Ovid Medline using medical subject headings and keywords on topics related to patient positioning, IV echo contrast, alternative subcostal views, right ventricular outflow tract (RVOT) hemodynamics, and point-of-care transesophageal echocardiography. Articles known to the authors were also selected based on expert opinion.

STUDY SELECTION

Articles specific to patient positioning, IV echo contrast, alternative subcostal views, RVOT hemodynamics, and point-of-care transesophageal echocardiography were considered.

DATA EXTRACTION

One author screened titles and extracted relevant data while two separate authors independently reviewed selected articles.

DATA SYNTHESIS

Impediments to acquiring quality and interpretable images in critically ill patients are common. Notably, body habitus, intra-abdominal hypertension, dressings or drainage tubes, postoperative sternotomies, invasive mechanical ventilation, and the presence of subcutaneous emphysema or lung hyperinflation are commonly encountered obstacles in transthoracic image acquisition in the ICU. Despite these obstacles, the bedside clinician may use obstacle-specific maneuvers to enhance image acquisition. These may include altering patient positioning, respiratory cycle timing, expanding the subcostal window to include multilevel short-axis views for use in the assessment of RV systolic function and hemodynamics, coronal transhepatic view of the inferior vena cava, and finally point-of-care transesophageal echocardiography.

CONCLUSIONS

Despite common obstacles to point-of-care echocardiography in critically ill patients, the beside sonographer may take an obstacle-specific stepwise approach to enhance image acquisition in difficult-to-image patients.

Prevalence of high-risk aortic arch atherosclerosis features on computed tomography angiography in embolic stroke of undetermined source

INTRODUCTION

Embolic stroke of undetermined source (ESUS) comprises a heterogenous group. There is a need to further identify etiologies within this group to guide management strategies. We examined the prevalence of aortic arch atherosclerosis (AAA) on CT angiography (CTA) in patients with embolic stroke of undetermined source (ESUS) to characterize high-risk plaque features.

METHODS

All patients from two prospective multicenter acute ischemic stroke studies (INTERRSeCT and PRove-IT) were included if the CTA adequately imaged the proximal aortic arch and the stroke etiology was recorded. Three readers blinded to stroke etiology analyzed the following AAA plaque features on baseline CTA at the time of stroke: 1) thickness in millimetres (mm); 2) morphology (none, smooth, ulcerated, or protruding); 3) location within the aortic arch (proximal, transverse, or distal); and 4) calcification (none, single small, multiple small, single large, or diffuse extensive).

RESULTS

We included 1063 patients, of which 293 (27.6%) had ESUS (mean age 67.5 years; 46.4% men; median NIHSS 12; 80.6% large vessel occlusion). Mean AAA thickness was significantly larger in ESUS patients (3.8 mm) compared to non-ESUS patients (3.0 mm; p<0.0001) and to a subgroup of patients with large artery atherosclerosis (2.9 mm; p=0.003). ESUS patients had a significantly higher proportion of ulcerated or protruding plaques (17.4% vs 10.3%; risk ratio 1.7, 95% C.I. 1.2-2.4, p=0.002). The location of AAA in the ESUS group was the ascending aorta in 37.9%, transverse arch in 42.3%, and descending aorta in 84.6%. Although AAA was mostly located in the distal aortic arch, ulcerated or protruding plaques were least common in the distal arch (p=0.002). There was no difference between ESUS and non-ESUS patients in plaque location (p=0.23) or calcification grade (p=0.092).

CONCLUSION

ESUS patients in our study had thicker AAA and a higher prevalence of ulcerated or protruding plaques located more proximally within the aortic arch. High-risk plaque features may suggest a causal role of AAA in the ESUS population with visible intracranial occlusions.

Focused Ultrasonography in Cardiac Arrest

Rapid diagnostic tools available to the emergency physician caring for cardiac arrest patients are limited. Focused ultrasound (US), and in particular, focused echocardiography, is a useful tool in the evaluation of patients in cardiac arrest. It can help identify possible causes of cardiac arrest like tamponade and pulmonary embolism, which can guide therapy. US can also yield prognostic information, with lack of cardiac activity being highly specific for failure to achieve return of spontaneous circulation. US may also be used to aid in procedural guidance. Recently, focused transesophageal echocardiography has been used in the emergency department setting.

Indications, Clinical Impact, and Complications of Critical Care Transesophageal Echocardiography: A Scoping Review

BACKGROUND

Critical care transesophageal echocardiography (ccTEE) is an increasingly popular tool used by intensivists to characterize and manage hemodynamics at the bedside. Its usage appears to be driven by expanded diagnostic scope as well as the limitations of transthoracic echocardiography (TTE) - lack of acoustic windows, patient positioning, and competing clinical interests (eg, the need to perform chest compressions). The objectives of this scoping review were to determine the indications, clinical impact, and complications of ccTEE.

METHODS

MEDLINE, EMBASE, Cochrane, and six major conferences were searched without a time or language restriction on March 31^st, 2021. Studies were included if they assessed TEE performed for adult critically ill patients by intensivists, emergency physicians, or anesthesiologists. Intraoperative or post-cardiac surgical TEE studies were excluded. Study demographics, indication for TEE, main results, and complications were extracted in duplicate.

RESULTS

Of the 4403 abstracts screened, 289 studies underwent full-text review, with 108 studies (6739 patients) included. Most studies were retrospective (66%), performed in academic centers (84%), in the intensive care unit (73%), and were observational (55%). The most common indications for ccTEE were hemodynamic instability, trauma, cardiac arrest, respiratory failure, and procedural guidance. Across multiple indications, ccTEE was reported to change the diagnosis in 52% to 78% of patients and change management in 32% to79% patients. During cardiac arrest, ccTEE identified the cause of arrest in 25% to 35% of cases. Complications of ccTEE included two cases of significant gastrointestinal bleeding requiring intervention, but no other major complications (death or esophageal perforation) reported.

CONCLUSIONS

The use of ccTEE has been described for the diagnosis and management of a broad range of clinical problems. Overall, ccTEE was commonly reported to offer additional diagnostic yield beyond TTE with a low observed complication rate. Additional high quality ccTEE studies will permit stronger conclusions and a more precise understanding of the trends observed in this scoping review.

Left ventricle segmentation in transesophageal echocardiography images using a deep neural network

PURPOSE

There has been little progress in research on the best anatomical position for effective chest compressions and cardiac function during cardiopulmonary resuscitation (CPR). This study aimed to divide the left ventricle (LV) into segments to determine the best position for effective chest compressions using the LV systolic function seen during CPR.

METHODS

We used transesophageal echocardiography images acquired during CPR. A deep neural network with an attention mechanism and a residual feature aggregation module were applied to the images to segment the LV. The results were compared between the proposed model and U-Net.

RESULTS

The results of the proposed model showed higher performance in most metrics when compared to U-Net: dice coefficient (0.899±0.017 vs. 0.792±0.027, p<0.05); intersection of union (0.822±0.026 vs. 0.668±0.034, p<0.05); recall (0.904±0.023 vs. 0.757±0.037, p<0.05); precision (0.901±0.021 vs. 0.859±0.034, p>0.05). There was a significant difference between the proposed model and U-Net.

CONCLUSION

Compared to U-Net, the proposed model showed better performance for all metrics. This model would allow us to evaluate the systolic function of the heart during CPR in greater detail by segmenting the LV more accurately.

Rationale and clinical applications of 4D flow cardiovascular magnetic resonance in assessment of valvular heart disease: a comprehensive review

BACKGROUND

Accurate evaluation of valvular pathology is crucial in the timing of surgical intervention. Whilst transthoracic echocardiography is widely available and routinely used in the assessment of valvular heart disease, it is bound by several limitations. Although cardiovascular magnetic resonance (CMR) imaging can overcome many of the challenges encountered by echocardiography, it also has a number of limitations.

MAIN TEXT

4D Flow CMR is a novel technique, which allows time-resolved, 3-dimensional imaging. It enables visualisation and direct quantification of flow and peak velocities of all valves simultaneously in one simple acquisition, without any geometric assumptions. It also has the unique ability to measure advanced haemodynamic parameters such as turbulent kinetic energy, viscous energy loss rate and wall shear stress, which may add further diagnostic and prognostic information. Although 4D Flow CMR acquisition can take 5-10 min, emerging acceleration techniques can significantly reduce scan times, making 4D Flow CMR applicable in contemporary clinical practice.

CONCLUSION

4D Flow CMR is an emerging CMR technique, which has the potential to become the new reference-standard method for the evaluation of valvular lesions. In this review, we describe the clinical applications, advantages and disadvantages of 4D Flow CMR in the assessment of valvular heart disease.

Is There a Role for Transesophageal Echocardiography in the Perioperative Trauma Patient?

Purpose of Review

This review article summarizes the advantages and potential uses of focused transesophageal echocardiography (TEE) in the perioperative period for trauma patients. We suggest a locally developed TEE protocol for trauma and provide strategies to achieve widespread use of TEE in the anesthesia care of trauma patients.

Recent Findings

In recent years, TEE has gained interest as an additional modality as point-of-care ultrasound (POCUS) for the resuscitation of acutely ill patients in whom transthoracic echocardiography is not feasible or non-diagnostic. Nevertheless, its use among non-cardiac anesthesiologists is still limited compared to the more traditional POCUS applications.

Summary

A goal-directed, focused TEE can be performed at the bedside in different locations and mechanically ventilated patients. TEE provides relevant diagnostic information to guide the resuscitation of acutely injured patients, particularly to identify life-threatening hidden thoracic injuries in the scenario of patients with multi-system trauma.

Feasibility, utility, and safety of fully incorporating transesophageal echocardiography into emergency medicine practice

INTRODUCTION

Transthoracic echocardiography (TTE) is a standard procedure for emergency physicians (EPs). Transesophageal echocardiography (TEE) is known to have great utility in patients who are critically ill or in cardiac arrest and has been used by some EPs with specialized ultrasound (US) training, but it is generally considered outside the reach of the majority of EPs. We surmised that all of our EPs could learn to perform focused TEE (F-TEE), so we trained and credentialed all of the physicians in our group.

METHODS

We trained 52 EPs to perform and interpret F-TEEs using a 4-h simulator-based course. We kept a database of all F-TEE examinations for quality assurance and continuous quality feedback. Data are reported using descriptive statistics.

RESULTS

Emergency physicians attempted 557 total F-TEE examinations (median = 10, interquartile range = 5-15) during the 42-month period following training. Clinically relevant images were obtained in 99% of patients. EPs without fellowship or other advanced US training performed the majority of F-TEEs (417, 74.9%) and 94.3% (95% confidence interval [CI] = 91.4%-96.3%) had interpretable images recorded. When TTE and TEE were both performed (n = 410), image quality of TEE was superior in 378 (93.3%, 95% CI = 89.7%-95%). Indications for F-TEE included periarrest states (55.7%), cardiac arrest (32.1%), and shock (12.2%). There was one case of endotracheal tube dislodgement during TEE placement, but this was immediately identified and replaced without complication.

CONCLUSION

After initiating a mandatory group F-TEE training and credentialing program, we report the largest series to date of EP-performed resuscitative F-TEE. The majority of F-TEE examinations (75%) were performed by EPs without advanced US training beyond residency.

Transesophageal Echocardiography Simulator Training: A Systematic Review and Meta-analysis of Randomized Controlled Trials

SUMMARY STATEMENT

We aimed to assess the learning effects of novice transesophageal echocardiography (TEE) simulator training and to identify gaps in existing studies. We performed a systematic review and meta-analysis of randomized controlled trials (RCTs) comparing the learning effects of novice TEE training with versus without simulators, searching published articles and proceedings in 6 major databases in June 2019. We included 9 RCTs (268 participants). Compared with nonsimulator training, TEE simulator training resulted in higher skill and knowledge posttraining test scores with large effect sizes (standardized mean difference = 0.81 for skill, 1.61 for knowledge; low-certainty evidence) and higher training satisfaction with a small effect size (standardized mean difference = 0.36; very low-certainty evidence). No RCTs reported training budget or patient outcomes. Additional well-designed studies with low risk of bias and large sample sizes are needed to provide reliable and robust findings and develop more effective TEE simulation-based training curricula.

The Utility of Point-of-Care Transesophageal Echocardiography in the Intensive Care Unit for Detecting the Cause of Hemolytic Anemia After Ascending Aorta Replacement

Hemolysis is well-known as a complication after mitral valve surgery, and hemolytic anemia after valvular surgery in the intensive care unit (ICU) usually leads ICU physicians to a careful imaging examination. However, hemolytic anemia following aortic replacement rarely has been described and little is known. The authors present a patient with hemolytic anemia after ascending aortic replacement for type A aortic dissection repair. In the patient, transthoracic echocardiography (TTE), contrast-enhanced computed tomography (CT), and laboratory tests failed to identify the cause of hemolytic anemia. Transesophageal echocardiography (TEE) finally revealed a stenosis of the proximal anastomosis of the graft, with turbulent flow due to protrusion into the lumen. This finding was thought to represent the cause of hemolysis and was sufficient to convince the surgeons to repair the lesion. In the repair surgery, inversion of the inner felt strip that had been applied to the proximal anastomosis of the replacement ascending aorta was confirmed as shown with TEE, and the constricted anastomosis was repaired. Although TTE and CT generally have been used to identify the lesion of mechanical hemolysis after cardiac surgery, point-of-care TEE could have promptly identified the causative lesion in the present patient. Point-of-care TEE in the ICU could become a first-line imaging examination for postoperative hemolysis in cardiac patients.

Ultrasonography in the Critical Care Unit

Purpose of Review

This article summarizes the utility and evidence supporting the use of ultrasound exams in the intensive care unit.

Recent Findings

Point-of-care ultrasonography (POCUS) is widely used by intensivists managing critically ill patients whereby they can accurately and rapidly assess for many pathologies such as pneumothorax, pulmonary edema, hydronephrosis, hemoperitoneum, and deep vein thrombosis among others. Basic and advanced critical care echocardiography, including transesophageal echocardiography, are routinely performed to determine the etiology of hemodynamic instability in undifferentiated shock and to guide subsequent therapy. The use of POCUS in the assessment of volume status is controversial with studies demonstrating that respiratory variation of the IVC is not reliable and with analysis of aortic blood flow velocity after passive leg raising maneuvers being the most promising.

Summary

Point-of-care ultrasonography allows frontline clinicians to make real-time diagnoses and treatment decisions. This article will provide the reader with a broad overview of this important topic.

Electronic supplementary material

The online version of this article (10.1007/s11886-020-01393-z) contains supplementary material, which is available to authorized users.

Transesophageal echocardiography instead or in addition to transthoracic echocardiography in evaluating haemodynamic problems in intubated critically ill patients

Background

Transesophageal echocardiography (TEE) performed by intensivists is increasingly used in critically ill patients. However, TEE is usually not the preferred monitoring tool, especially when transthoracic echocardiography (TTE) appears to have addressed the clinical problems. As a result, it remains largely unknown whether TEE is a clinically valuable replacement or supplement for TTE as a primary tool in evaluating haemodynamic problems in critically ill surgical patients. The purpose of this study was to assess the diagnostic and therapeutic value of TEE instead or in addition to TTE in critically ill surgical patients with hemodynamic instability.

Methods

A prospective observational study was conducted. A total of 68 consecutive patients were enrolled from December 2016 to February 2018. TEE was routinely performed in addition to TTE, and the imaging data from TTE and TEE were successively disclosed to two different primary physicians, who reported any resulting changes in management. The two physicians were required to reach a consensus if there was any disagreement. The results of the additional TEE examination were compared with the clinical findings and TTE information. The image quality of TTE views was classified as a good (score 2), suboptimal (score 1) or poor view (score 0). According to the scores of TTE images, the patients were divided into two groups: patients with adequate TTE views (score ≥6) and inadequate TTE views (score <6).

Results

The results of additional TEE examination were classified into four categories. TEE failed to provide additional information about the initial diagnosis and therapy (class 1) in 26 patients (38.2%). Of the remaining 42 patients (61.8%), TEE instead or in addition to TTE revealed new findings or led to significant changes in therapy, as TTE supplied inadequate information. TEE used in addition to TTE led to a new diagnosis without therapeutic implications (class 2) in 11 patients (16.2%) and made a major clinical contribution leading to a therapeutic change (class 3) in 23 patients (33.8%). TEE used instead of TTE determined the diagnosis and therapy in 8 patients (11.8%) whose haemodynamic problems could not be addressed by TTE (class 4). In total, TEE had critical therapeutic benefits (class 3 and 4) that was not provided by TTE in 31 patients (45.6%). Of particular concern was that TEE had a higher proportion of therapeutic benefits to patients with inadequate TTE views than those with adequate TTE views (54.3% vs. 27.3%, P=0.036).

Conclusions

TEE as a feasible clinical tool is useful for critically ill surgical patients with hemodynamic instability, especially for the patients with inadequate TTE views. TEE instead or in addition to TTE could provide valuable information for diagnosis, which may bring significant therapeutic benefits.