Abstract
The field of echocardiography has evolved due to miniaturization, from increasing functionality of high end devices to the portability of hand held devices which have a footprint similar to a smart phone.
This chapter seeks to inform the reader on the fundamental capabilities and limitations of hand-held echocardiography and the types of devices available. We explore the learning requirements for all levels of users and validation studies pertinent to hand held echocardiography. A guide is provided on protocols which may be used to perform and report hand held echocardiography. We discuss the qualitative and semi quantitative methods available to maximise the diagnostic yield of hand held echocardiography. Finally we discuss the specific roles of this modality currently and its potential role in the future.
Despite and perhaps because of its diminutive size, hand held echocardiography represents a revolution in the field of cardiac imaging and auscultation: For the first time since its introduction in 1816, a true hand held “stetho-scope.”
Keywords
handheld echocardiography of HHE, HHE protocol, learning curve, specific roles of HHE, validation of HHE
Introduction
Although Moore’s law of increasing computing capabilities (including processing power and memory) has propelled the development of all imaging modalities, its most tangible manifestation has been in the field of ultrasonography. Unrestrained by the physical limitations (gantry size and power requirements) of the other modalities, echocardiography machines have evolved from large, cumbersome pieces of equipment to handheld echocardiography (HHE) devices that are the size of a mobile phone ( Fig. 46.1 ).
Three aspects of miniaturization have been particularly important for HHE. The display interface has benefited tremendously from technological evolution—cathode ray tube monitors have been replaced by lightweight, high-resolution liquid crystal display (LCD) screens. Developments in microprocessors have led to a shift in the balance between hardware and software so that it is closer to the transducer, usurping some of the functionalities previously performed by the scanner ( Fig. 46.2 ). Likewise, there has been a progressive drop in the size of digital beamforming components from 1 μm to 100 nm. Freeing the handheld device from the ECG leads through fixed time acquisitions or more complex tracking iterations using mitral annular movement of speckle tracking has led to increased portability and reduced size.
The inextricable connection between energy transmitted by a system and the information gained has led powerful, high-end systems to hold an advantage over their battery-dependent counterparts. These power-to-performance issues have benefited from greater efficiency of systems and greater integration, although the limitations imposed by their size continue to leave HHE at one end of the spectrum of ultrasound devices.
Current Handheld Devices
A modern HHE device is characteristically lightweight, portable, and can fit into a coat pocket—in contrast to previous miniaturized models. These devices provide B mode grayscale imaging and in some cases color Doppler. Unlike their fully functional mobile (but non-HHE) counterparts, which are essentially complete echocardiographic devices, most HHE devices do not provide spectral Doppler. They also have smaller screens and a display that is lower in resolution than standard echocardiographic devices ( Fig. 46.3 ). Their various properties are described in Table 46.1 . It is important to remember that comparison with standard echocardiography is only reasonable if these devices are touted as a replacement for echocardiography rather than an extension of the physical examination.
MobiUS™ SP1 | Vscan V1.2 | ACUSON P10™ | SignosRT | |
---|---|---|---|---|
Company | MobiSante | GE Healthcare | Siemens | Signostics |
Weight (g) | 330 | 390 | 725 | 392 |
Display Size (Inches) | 4.1 | 3.5 | 3.7 (Diagonal) | 4.5 (height) |
Imaging options | Gray scale | Grayscale Color Doppler | Grayscale | Gray scale M-mode PW Doppler |
Transducer Frequency (MHz) | 3.5–5.0 (and 7.5) Mechanical Single element | 1.7–3.8 Phased Array | 2–4 Phased array | 3.0–5.0 |
Interface with PC | USB synch | Micro SD card | Software | Micro SD card |
Battery Capacity (minutes) | 60 | 90 | 100 | 120 |
The limitations of HHE are related to the imaging modes, processing, display, and ability to do measurements. None of the devices have continuous-wave Doppler, and only one has pulsed-wave Doppler capability. Because Doppler wave-form analysis is a cornerstone in the severity assessment of valvular and diastolic heart disease, this represents an important (and potentially avoidable) limitation. The high-resolution display of high-end devices provides high-fidelity images that are difficult to reproduce on HHE devices. Finally, post-acquisition analysis and measurements are a key component of analysis. Although three devices enable measurement of distances and area, volumetric assessments are not possible. Adjustment of imaging parameters such as zoom, changing focal point, narrowing sector width to improve frame rate, changing mechanical index for contrast studies, harmonic imaging, changing dynamic range or grayscale maps, and changing frequency are all currently unavailable on such devices. Future iterations may overcome some if not all of these limitations, but currently there is a clear difference between a standard echocardiogram and HHE equipment.
The Learning Curve
The acquisition of information from the traditional physical examination is less reliable than in former times. HHE is a potential replacement for bedside diagnosis, but ultrasonography has not been as well taught and is currently restricted to certain physicians, surgeons, and sonographers. The relative cost, portability, and applicability of HHE devices make them ideal for more widespread dissemination. However, for those unfamiliar with echocardiography, there needs to be a learning process and assessment of competence. The recommended training requirements for performance and reporting of echocardiography are summarized in Table 46.2 .
Level | Duration of Training (m) | Cumulative duration of training (m) | Minimal No. of TTE Exams Performed | Minimal No. of TTE Exams Interpreted |
---|---|---|---|---|
1 | 3 | 3 | 75 | 150 |
2 | 3 | 6 | 150 | 300 |
3 | 6 | 12 | 300 | 750 |
Clearly Level 2– and 3–trained individuals will readily adapt to HHE; both American Society of Echocardiography (ASE) and European Association of Echocardiography (EAE) recommend that experienced cardiologists and sonographers should be able to use handheld devices. Miniaturization of technology has outpaced training and accreditation guidelines, and ultrasonography has moved from the field of radiologists and subspecialty physicians to residents and general physicians. The ASE advises additional training for Level 1–trained individuals and EAE recommends additional training for cardiologists not fully conversant with echocardiography. The American College of Emergency Physicians (ACEP) has addressed the role of focused cardiac ultrasound (FCU) in the emergency department and provided guidelines on the acquisition and interpretation of ultrasound images on a range of diagnostic possibilities. At an earlier stage in training, HHE can also be used as teaching aide to visualize anatomy and physiology at the bedside. Wittich demonstrated that 79% of medical students could produce a satisfactory parasternal long axis (PLAX) image within 3 weeks of didactic and practical sessions. It seems feasible to use HHE to acquire images and interpret basic cardiac function early during medical school.
There have been multiple previous comparisons of HHE and physical examination—a recent example is summarized in Table 46.3 . There is clear incremental value of HHE over the traditional physical examination; the area under the receiver operating characteristic (ROC) curve was 1.97 for physical examination, 2.42 for ECG and physical examination, and 6.23 for HHE-based echocardiography.
Echocardiogram Findings | HHE % Correct | PE % Correct | Incremental % | P |
---|---|---|---|---|
Normal LV function | 89 | 58 | 31 | <0.0001 |
Abnormal LV function | 96 | 35 | 61 | <0.0001 |
Normal RV function | 94 | 57 | 37 | <0.0001 |
Abnormal RV function | 68 | 21 | 47 | 0 |
Pulmonary hypertension absent | 92 | 89 | 3.1 | 0.36 |
Pulmonary hypertension present | 53 | 42 | 10 | 0.33 |
Valve disease, mild or absent | 94 | 91 | 3.5 | 0.23 |
Valve disease, moderate or severe | 71 | 31 | 39 | 0.00 |
Miscellaneous findings absent | 77 | 64 | 13 | 0.02 |
Miscellaneous findings present | 47 | 3 | 44 | <0.0001 |
Handheld ultrasonography can be applied to any organ system, although for the purposes of this chapter it will refer solely to the examination of the cardiovascular system. FCU is a specific adaptation—involving HHE—that provides a “Focused examination of the cardiovascular system performed by a physician using an ultrasound as an adjunct to the physical examination to recognise specific ultrasonic signals that represent a narrow list of potential diagnosis in specific clinical settings.” This must be differentiated from limited transthoracic echocardiography, which refers to a reduced number of images performed on a standard and thus a high-end machine by an echocardiographer with appropriate qualifications and interpreted by a cardiologist with the necessary level of experience. The training recommendations on FCU by the ASE comprise three educational components—a didactic component (ultrasound physics, basic cardiac anatomy and views), practical training (image acquisition and correction of technique by experienced echocardiographers), and image interpretation. Recent literature has demonstrated that electronic modules were equivalent to didactic teaching, but there is no substitute for “hands-on” sonographer-based training, or the use of a training log to track the number of successful echocardiographic interrogations for specific cardiac pathologic conditions. There is considerable variation in the duration of training programs depending on the level of experience of operators, opportunity to scan, time, and resources. The length of the learning curve is variable. General practitioners with 8 hours of supervised training using HHE were able to assess left ventricle (LV) function with a sensitivity of 83% and a specificity of 78%. A regression model based on more than 230 HHE examinations performed and interpreted by 30 residents and audited against cardiologists’ measurements suggested improvements every 10 scans and that 30 scans would result in a minimal overall difference. It must be noted that with all R values of less than 0.2, the fit and predictive value of the model was limited.
Validation of Handheld Devices
Table 46.4 summarizes the literature reporting sensitivity, specificity, or agreement between HHE and standard echocardiography. Reported levels of sensitivity, specificity, agreement, and weighted Kappa values, depend on multiple factors including technology used (availability of color Doppler, ability to make measurements), skill level of the operator and interpreter (echocardiographer/cardiologist vs. internist/general resident) and the exact target. Categorical evaluations (presence vs. absence of pathology) are likely to be more robust than quantitative evaluations (e.g., severity of a valvular lesion). The variability in Table 46.4 also reflects that in addition to the progressive development of HHE, there has been an evolution of standard echocardiography over time, with a shifting paradigm toward quantitative analysis using tools such as strain, 3D volumetric analysis, and a focus on quantitative assessment of Doppler waveforms. The variation in agreement despite evolution of handheld technology does not necessarily represent a decline in diagnostic ability; rather, it is testament to the evolution of high-end echocardiography.
Author | Year | Device | Operator | Concordance with SE (Ag, k, r) | Accuracy (Ac, Sn, Sp) | Comments |
---|---|---|---|---|---|---|
Fukuda | 2009 | Acuson P10 | Sonographer | r 0.87–0.98 | (RWMA) Sn 88%, Sp 95% | Valves not assessed (no color Doppler) |
Galderisi | 2010 | Vscan | Experts/Trainee | k 0.84 | Sn 97%, Sp 84% | Calcification k 1.00 (AS not reported) |
Andersen | 2011 | Vscan | Cardiologists | r 0.62–1.00 | Sn 63%–100%, Sp 68%–100% | Moderate correlation with AS (r 0.62), poor specificity for LA size (Sp 68%) |
Gianstefani | 2011 | Vscan | Sonographers | Ag 79%, k 0.47 | X | No k or Ag data for valvular disease, reported good concordance |
Giusca | 2011 | AcusonP10 | Cardiology trainees | k 0.56–0.81 | Sn 56%–71%, Sp 90%–100% | No color Doppler |
Lafitte | 2011 | Vscan | Expert physician | k 0.64–0.91 | x | Moderate agreement with aortic root size (k 0.64) |
Liebo | 2011 | Vscan | Fellow/Physician | x | Ac (0.58–0.91) | Lowest accuracy for IVC size (0.54) |
Prinz | 2011 | Vscan | Cardiologists | k 0.21–1.00 | x | Severity AS (k 0.21), qualitative valve disease (k 0.9 any regurgitation), (K 1.0 any AS/calcification) |
Razi | 2011 | Vscan | Residents | Ag LVSD 86%–98% | EF < 40% Sn 94%, Sp 94% | LV function study |
Reant | 2011 | Vscan | Residents | k 0.86–0.90 | x | LV function, MR and PE assessment |
Amiel | 2012 | Vscan | Sonographers | k 0.75 | x | LV function study |
Biais | 2012 | Vscan | Experienced physician | k 0.70–0.90 | Sn 77%–94%, Sp 96%–100% | Good agreement for severe RV dilatation (k 0.87) |
Kimura | 2012 | AcusonP10 | Sonographer | x | x | EPSS Ac 82%, Sn 47%, Sp 98%, LA Ac 64%, Sn 79%, Sp 52% |
Mjolstad | 2012 | Vscan | Cardiologist/Internist | Overall ≥0.85 | Sn/Sp ≥89% (valvular) | LV sz/fn (Sn/ Sp 97%/ 99%) LA r 0.65,IVC r 0.68 |
Prinz | 2012 | Vscan | Sonographer | r 0.60–1.0 | x | LV Fn r >0.6, valvular regurgitation k 0.10–0.90 |
Abe | 2013 | Vscan | Sonographer | AS score k 0.85 | Mod-Sev AS Sn 84%, Sp 90% | AS study |
Kitada | 2013 | Vscan | Expert physician | Ag 90% | X | Cost-effectiveness study |
Mjolstad | 2013 | Vscan | Residents | r 0.44–0.86 | Sn 40%–92%, Sp 81%–94% | Poor sensitivity for Rv functional assessment |
Testuz | 2013 | Vscan | Cardiologists | k 0.46–0.90 | X | RV size and LV size k 0.46, K 0.59 (function and valvular k ≥0.60) |
Beaton | 2014 | Vscan | Pediatric cardiologist | X | Sn 90%, Sp 93% for RHD | Modified assessment of RHD |
Cullen | 2014 | Vscan | Sonographers | k 0.49–0.91 | X | LVH, atrial size, PE k ≤0.55, valvular and fn ≥0.61 |
Khan | 2014 | Vscan | Cardiology fellows | Ag 90%–97% | Sn 79%–96%, Sp 92%–99% | Cost-effectiveness study/good overall agreement |
Mehta | 2014 | Vscan | Cardiology fellows | % correct- 82% | x | LV fn (89% N, 96% Abn), RV fn (N 94%, abn 68%), overall valve (absent/mild 94%, mod-sev 71%) |
Riley | 2014 | Vscan | Cardiologists | k 0.82 | Sn 75% | Pediatric patients |
Di Bello | 2015 | Vscan | Cardiologists | k 0.82, Ag 94% | Sn 94%, Sp 88% | Comparison to physical examination. Incremental area under ROC curve. |
A consensus statement published by the ASE found that the use of HHE for assessment of LV enlargement (6 studies), LV hypertrophy (8), LV systolic function (19), left atrium (LA) enlargement (9), right ventricle (RV) enlargement (3), RV systolic function (6), pericardial effusion (11), and inferior vena cava (IVC) size (2) have all been well validated. Similarly, recent literature has attempted to categorize accuracy of HHE as excellent (Sn ≥ 90%, Sp ≥ 95%) including studies by nonexperts, good (Sn ≥ 90%, Sp ≥ 95%) by experts, fair (Sn ≈ 80%, Sp≈ 80%), and variable. Using this categorization, HHE for detection of pericardial effusion was considered to have excellent accuracy. Good accuracy was demonstrated for LV size, LV systolic function, regional wall motion abnormalities (RWMA), ultrasound lung comets, pleural effusion, and abdominal aortic aneurysm (AAA) detection. The accuracy was deemed fair for LA size and assessment of the presence and severity of aortic and mitral valve disease. The accuracy of HHE for RV and IVC assessment was considered to be quite variable.
Suggested Protocol
HHE may be used for a targeted assessment, for example, IVC diameter, a limited examination (apical views only), or a more comprehensive assessment. Any protocol recommendations need to balance the comprehensiveness of an examination with expediency.
Clearly the clinical context will determine the best approach and duration of echocardiography. In an emergent setting such as cardiac arrest, HHE can be used to guide resuscitation. Images can be acquired in the 5- to 10-second window during pulse check or between defibrillations, with the aim to exclude potential causes, including hypovolemic shock (small cavity with contact of ventricular walls), saddle pulmonary embolism (dilated, dysfunctional RV, systolic flattening of septum), cardiogenic shock (markedly reduced ejection fraction [EF]), or pericardial tamponade. The detection of a significant pericardial effusion can be achieved easily and rapidly as shown in Fig. 46.4 .
The international consensus has recommended a systematic approach of obtaining multiple views in a protocoled manner. HHE does not require the execution of all views used on standard echocardiography. The suggested views for an FCU examination include subcostal long axis, subcostal IVC, parasternal long axis, parasternal short axis, and apical four chamber. Clearly these recommendations should be seen as the minimum necessary number of views and have been designed to limit duration of the examination in time-sensitive scenarios. When appropriate, we also recommend the acquisition of apical two- and three-chamber views. In a time-limited examination, there is a need to employ whichever views are necessary to answer the clinical question, informed by the insight that all may not be possible with HHE ( Fig. 46.5 ). If the device is equipped with color, then this should be employed in each view as well.
Documentation of results is also necessary. The ASE has recommended that in addition to identification data and results, limitations of the study and recommendations for additional studies should be reported. A suggested template is shown in Fig. 46.6 .
Quantitative Assessment with Handheld Echocardiography
The size of a regurgitant jet by color Doppler and its temporal resolution are influenced by transducer frequency, settings such as gain, output power, Nyquist limit, and depth of the sector. HHE devices do not have the ability to significantly change these settings. The predominant strength of handheld echocardiography lies in its immediate ability to facilitate bedside decisions based on “eye-balling” chamber size, function, and severity of valvular lesions. However, a semiquantitative approach can also be implemented depending on the features available on a device.
Visual estimation and template matching have been the predominant manner with which to evaluate EF. However, semiquantitative approaches including linear measurements ( Fig. 46.7 ), with the Teichholz or Quinones formula to calculate EF, or mitral annular systolic plane excursion (mitral annular plane systolic excursion [MAPSE]; normal MAPSE is greater than 11 mm for women and greater than 13 mm for men) can be employed. In the future, models incorporating speckle tracking technology, initialized and constrained by a deformable model may offer an automated approach to assessment of the severity of LV dysfunction with HHE. The assessment of volume status may be assisted by evaluation of IVC size, and pulmonary congestion may be identified by the detection of “lung comets” caused by fluid in the interlobular septa ( Fig. 46.8 ). The use of IVC collapsibility and lung comets facilitate differentiation of cardiogenic from pulmonary causes of dyspnea.
As a means to facilitate more accurate determination of valvular heart disease, Beaton et al. have modified the classification of pathological valvular lesions by the World Heart Federation (WHF) ( Table 46.5 ). HHE-acquired examples of mitral regurgitation (MR) (see Fig. 46.7 ) and aortic regurgitation ( Fig. 46.9 ) can be assessed using this methodology. The importance of nonstandard views is demonstrated in a patient with carcinoid syndrome with moderate pulmonary and severe tricuspid regurgitation ( Fig. 46.10 ).
Criteria | Pathologic MR | Pathologic AR |
---|---|---|
Seen in 2 views | Seen in 2 views | |
2012 WHF criteria | In at least 1 view jet length >2 cm | In at least 1 view, jet length >1 cm |
Velocity >3 m/sec for one complete envelope | Velocity >3 m/s in early diastole | |
Pansystolic jet in at least 1 envelope | Pansystolic jet in at least 1 envelope | |
Modified criteria | ||
Seen in 2 views | Seen in 2 views | |
In at least 1 view jet length >2 cm | In at least 1 view, jet length >1 cm | |
Pansystolic jet (by color Doppler) | Pansystolic jet (by color Doppler) |