INDIVIDUALIZED LUNG-WATER-BASED INTELLIGENT MONITORING OF CARDIO-PULMONARY BIOMARKERS.

Ph.D. School in Engineering “L. Da Vinci”

PhD Thesis

PhD course on Automatic, Robotics and Bioengineering

XXVI (2011-2013)

SSD: ING-INF/06

INDIVIDUALIZED LUNG-WATER-BASED

INTELLIGENT MONITORING

OF CARDIO-PULMONARY BIOMARKERS

Student: Rossella Raso

Supervisors: Eng. Giovanni Pioggia, MD Eugenio Picano,

MD Luna Gargani

1 Introduction 5 2 The B-lines state of the art 8 2.1

The extra-vascular lung water leakage

. . . 8 2.2

Clinical methods to quantify pulmonary edema

12

2.2.1

Chest radiography

. . . 13 2.2.2

Computed tomography

. . . 14 2.2.3

Nuclear magnetic resonance (NMR) imaging

. . . 15 2.2.4

Positron emission tomography

. . . 15 2.2.5

Electrical impedance tomography (EIT)

. . . 17 2.2.6

Double Indicator dilution and thermodilution

meth-ods

. . . 19 2.2.7

Discussion

. . . 21 2.3

Chest sonography

. . . 22 3 The lung ultrasound: an innovative EVLW assessment

method 24

3.1

B-lines ecographic and physical development

3.2

B-lines clinical develompent principles

. . . 37

3.3

Ultrasounds theory

. . . 39

3.4

Ultrasonography

. . . 42

3.4.1

Producing a sound wave

. . . 44

3.4.2

Receving the echoes to generate (reconstruct)

and display the images

. . . 47

3.4.3

The ultrasound waves inside the human body

. . 48

3.4.4

Pros and cons of ultrasonography

. . . 50

3.5

Chest ultrasonography

. . . 52

4 The software iULC© Copyright #D007915 28/01/2013 56 4.1

Physiology of the sense of sight

. . . 56

4.2

The Hough and the Radon mathemarical

trans-formations

. . . 60

4.3

The iULC© algorithm

. . . 66

4.4

Knowledge-based models and Artificial Neural

Networks

. . . 73

4.4.1

Biological neural networks

. . . 75

4.4.2

The Occam principle

. . . 81

4.4.3

The formal neuron

. . . 83

4.4.4

Architectures and learning rules

. . . 85

4.4.5

Supervised learning

. . . 87

4.4.6

Unsupervised learning

. . . 93

4.4.7

The Probabilistic Neural Network (PNN)

. . . 98

4.5

The identification of the model

. . . 99

4.7

Network classification performances statistical

assessment

. . . 109 4.8

Study limitations

. . . 117 4.9

The ultrasound portable system framework

. . 117 5 Heart rate and heart rate variability assessment in anorexia

nervosa adolescents. 128 5.1

Methods

. . . 132

Extravascular lung water is a crucial parameter for the management of many different pathological conditions, especially heart failure [1]. Acute heart fail-ure syndrome (AHFS) is a major public health problem. It is defined as a gradual or rapid change in heart failure (HF) signs and symptoms, which of-ten results in an unplanned hospitalization and a need for urgent therapy. Heart failure (HF) is the most frequent cause of hospitalization among people older than 65 years, and over the past decade the rate of hospitalization for HF has greatly increased. Despite new treatments and improvement in survival, hospitalizations in HF have steadily increased over the last 30 years [2]. In pa-tients with impending acute HF, there is a relatively long incubation period of days and weeks, characterized by lung water accumulation; in the congestion cascade, pulmonary congestion can be detected well before the appearance of clinical signs and symptoms [15]. Detection and treatment of hemodynamic congestion before it is clinically evident may be useful under many perspectives, the most important is that they may prevent hospitalization and progression of HF and, also, could lead to hospitalization costs reduction; in fact, this would be economically notable, since it is estimated that expenditures for HF range between 1% and 2% of the total healthcare budget. These improvements are especially important, since each episode of worsening congestion, resulting in

hospitalization for acute HF syndrome, might contribute to the progression of left ventricular dysfunction, which in turn would increase the risk for future cardiac events. In a near future, we hope that extravascular lung water assess-ment could be performed by non-invasive wearable devices, allowing remote continuous monitoring of pulmonary congestion.

Evaluation of the lung has been generally considered off-limits for ultrasound, since it is standard textbook knowledge that ultrasound energy is rapidly dissi-pated by air [3]. This is why ultrasound is not considered useful for evaluating the pulmonary parenchyma in physiologic conditions, where the high differ-ence in acoustic impedance between air and surrounding tissues determines the complete reflection of the ultrasound beam and does not allow any image to be visualized. The presence of extra-vascular lung water (EVLW) or inter-stitial fibrosis opens the previously locked pulmonary acoustic window, since water- or collagen-thickened interlobular septa create the adequate acoustic mismatch able to trigger the phenomenon of sonographic reverberation that generates B-lines.

B-lines (previously called ULC, Ultrasound Lung Comets) are discrete laser-like vertical hyperechoic reverberation artifacts that arise from the pleural line, extend to the bottom of the screen without fading and move synchronously with respiration [4]. B-lines, obtained with non-ionizing, patient-friendly lung ultrasound (LUS) evaluation, have been recently proposed as an emerging echo-graphic marker for the assessment of pulmonary interstitial syndrome, reveal-ing the thickened pulmonary interstitium [5]. They are present both in water-thickened pulmonary interstitium (watery B-lines), such as in pulmonary con-gestion due to congestive heart failure (CHF) [4, 6] and in collagen-thickened pulmonary interstitium (fibrotic B-lines), such as in pulmonary fibrosis and

interstitial lung diseases [4, 7, 8].

The current technology for measuring EVLW can be inaccurate (chest X-ray), cumbersome (nuclear medicine and radiology techniques) or invasive (indicator dilution technique) [9, 10]. The simple and patient-friendly way of monitor-ing the pulmonary interstitial syndrome by ultrasounds would be crucial in many patient populations, from CHF [11] to dialysis [12], from acute lung injury/acute respiratory distress syndrome (ALI/ARDS) [13] to interstitial pulmonary fibrosis [7], since B-lines evaluation could allow a close follow-up to assess dynamic changes during pharmacological treatment and other interven-tions [14].

In less than 10 years, the proposal to use B-lines to evaluate pulmonary conges-tion in CHF patients has moved from the research setting to the clinical arena and has now entered recommendation papers [14, 15, 16]. This application of echography is especially valuable, since it is very easy, does not require the expertise necessary for the echocardiographic examination and interpretation [17], is fast to perform, portable, repeatable, non-ionizing and independent of cardiac acoustic window. However, since LUS is a sonographic technique, it suffers from ultrasound-related limitations, such as operator-dependence and lack of precise quantification. Our aim was to develop a soft computing-based B-lines analysis for the objective, automated and quantitative classification of the severity of the pulmonary interstitial syndrome in an operator-independent fashion. Soft computing-based models are capable to analyze complex medi-cal data, exploiting meaningful relationships to help physicians in diagnosis, treatment and prediction of the clinical outcomes.

2.1

The extra-vascular lung water leakage

An owed gratitude goes to doctor Luna Gargani since the B-lines state of the art concepts included in is chapter were freely influenced and inspired by her dissertation thesis “Comete polmonari, un segno ecografico di acqua extravas-colare polmonare: confronto con i peptidi natriuretici cardiaci nella diagnosi differenzaiale di dispnea”, discussed at the faculty of Medicine of the University of Pisa in 2006.

Although about 80% of the lung is made up of water, gas-exchanging air spaces are protected by various barriers and drains. In multiple disease states, through injury or pressure (or both), these protective mechanisms fail, resulting in the abnormal accumulation of extravascular lung water (EVLW) leading to the so-called phenomenon of “congestion”. Congestion is a main feature in patients with heart failure (HF). Analogous to fluid mechanics, if the pulmonary vas-cular bed is considered as a confined system, congestion can be expressed as increased weight of the fluid column. As force is transmitted through a fluid as a pressure wave, the pressure across pulmonary capillaries, that is pulmonary capillary wedge pressure (PCWP), is a good estimate of the pressure across that fluid column. Increased PCWP represents a condition of “hemodynamic

congestion”. Increased PCWP can lead to redistribution of excess fluid within the lungs, resulting in interstitial and alveolar edema, that characterize what we can define “pulmonary congestion” [2].

In the human body the extra-vascolar fluid filtration maily concentrate within the pulmonary micro-circulation. Usually this transfer happens through little gaps (junctions) among the endotelial cells, known as inter-cellular junctions. Once fluids and proteins have filtered into the interstitium can’t go further into the air, because the epithelium of the alveola has very dense inter-cellular junctions [18]. Both hemodynamic and pulmonary congestion may be present in absence of signs and symptoms of congestion, that characterize “clinical congestion”. The development of signs and symptoms represents the main reason for hospitalization in HF patients; this often occurs several days after the onset of PCWP elevation [19], as also shown by increase in intrathoracic fluid as early as 18 days before hospitalization, documented by intrathoracic impedance monitoring [20].

Figure 2.1.1: Schematic differences in normal pulmonary circulation, hemodynamic and pulmon-ary congestion.

An increase of capillary pressure defines detachments of inter-cellular junctions leading to a leakage of the filtered fluid which is a pathological state named interstitial lung edema. The principle paradigm describing fluid flux in the lung is the equation expressed in 1896 by the english physiologist E. H. Starling [21]. The equation describes the mechanism of fluids transportation across semi-permeable membranes, such as capillary walls and how liquids and proteins can cross a generic membrane depending on the equilibrium between hydrostatic and osmotic pressures at the two sides of the membrane and on thier own permeability.

“Lymph flow” is a term that summarize those mechanisms responsible for the returning of extravasated fluid to the vascular compartment:

EV LW = (Lp×S)[(P c − P i) − σv(Πc − Πi)] − lymphflow

where EVLW is expressed in ml, Lp (the hydraulic conductivity for water) in cm/min/mmHg, S (the surface area) in cm2, Pc and Pi (respectively the

hydrostatic pressure within the capillary and the interstitial spaces) in mmHg, Πc and Πi (respectively the oncotic pressure within the capillary and the inter-stitial spaces) in mmHg and σv (the reflection coefficient for protein) in absolute number (no units) [22].

Why the measurement of EVLW is important from a prognostic and diagnostic point of view? The main reasons are that EVLW leakage phenomenon leads to a pathologic condition clinically known as “pulmonary edema” and that to discover a method to measure EVLW in laboratory in an accurate, sensitive, reproducible, non-invasive and inexpensive [9] way should be crucial. However

such a technique has not been described yet. Imaging and indicator dilution techniques comprise the most common strategies for measuring lung water at the bedside. The most accurate (within 10% of the gravimetric gold standard) and most reproducible (< 5% between-test variation) are also, unfortunately, the most expensive and most difficult to implement for purposes of large-scale clinical trials or for routine clinical practice [23]. The standard chest radio-graph remains the best screening test for the detection of pulmonary edema. The assessment and the monitoring of pulmonary congestion in patients with HF is of great importance for clinical applications [24, 25]. The effort to develop and validate techniques to detect pulmonary edema even before it be-comes apparent clinically, or that could be used to provide information about the natural history of pulmonary edema or its response to therapeutic inter-vention, still continues. A proper assessment of this parameter in patients suffering of heart failure is crucial in each phase of their clinical management. Three different stages of congestion may be distinguished: emodinamic conges-tion, represented by the increase of replenishment pressures of the left ventricle; pulmonary congestion, which might be asymptomatic still; the clinical conges-tion, that shows by signs and sympthoms of pulmonary and systemic liquids accumulation [26]. Signs and sympthoms of the clinical congestion phase usu-ally demonstrate late, while it has been proved that the accumulation of lung water might forerun them by weeks [20]. A timely identification of pulmonary congestion might prepare for pharmacologic and behavioral actions instead that might prevent the appearance of clinic congestion which often worsen the patient’s clinical history and causes his admission to hospital. To monitor AHF patients is fundamental to adjust the pharmacological treatment and to opti-mize health care procedure since a wrong fluid management could be fatal [27].

It seems attractive for complement conventional Doppler-echocardiography in the fast evaluation of patients with known or suspected HF and dyspnoea as a presenting symptom in the emergency department (for the differential di-agnosis of dyspnoea), in-hospital management (for serial evaluations in the same patient and for tailoring diuretic therapy), home care (with hand-held echocardiography) and stress echocardiography laboratory (as a sign of acute pulmonary congestion during stress) [14].

2.2

Clinical methods to quantify pulmonary

edema

The measurable variable within a pixel or a voxel of an image or a volume from a lung is the concentration, e.g. expressed in ml EVLW/ml lung. A common issue related to all imaging methods is that since the lung is an air containing structure, the amount of lung parenchyma within each voxel can change, de-pending on the underlying state of lung inflation (lung volume). To quantify changes in images of EVLW in absolute terms, the signal over the entire organ must be integrated. Most imaging methods, except positron emission tomog-raphy (PET), for evaluating pulmonary edema (Table 2.2.1) produce estimates of total water content or concentration (i.e. vascular plus extravascular water). The data from such methods can be misinterpreted if the blood volume of the lungs is not constant. No modality can resolve composition of edema on den-sity alone since the edema, blood and inflammatory white cells are virtually identical, leading in general to an overestimation of EVLW per se. Certainly no modality can differentiate between intra- and extracellular water.

Measures Quantitation Accuracy Reproducibilty (COV) Sensitivity CXR LD Poor Unknown Unknown Moderate

CT LD Excellent Unknown Unknown High NMR TLW Fair Underestimates by 40 % 5-10 % Poor PET EVLW Excellent Underestimates by 10-15 % <5 % High

ID EVLW Good-excellent Overestimates by 10-20 % 4-8 % Moderate

Table 2.2.1:None of the methods can distinguish whether an increase in EVLW represents non-cellular pulmonary edema or non-cellular water from an inflamatory infiltrate. CXR, chest X-ray; CT, computed tomography; NMR, nuclear magnetic resonance; PET, positron emission tomography; ID, indicator diluition methods; LD, lung density; TLW, total lung water (of a region on an image); COV, coefficient of variation.

2.2.1

Chest radiography

A chest radiography is commonly used to detect whether or not pulmonary edema is present, to describe its overall distribution within the lung, and to evaluate associated findings to infer a probable etiology. It can also be used, at least semi-quantitatively, to estimate the amount of pulmonary edema that is present as well. Certain characteristic signs, such as pulmonary congestion, Kerley’s lines, and an interstitial pattern to the radiographic densities, are as-sociated with only modest increases in EVLW (perhaps as little as 30% above normal values) [28]. Cook et al. [29] and Staub et al. [30] have proposed that the sequence of fluid accumulation during acute pulmonary edema is quantal, meaning that the individual alveolus can exist only in a state of being either airfilled and expanded or fluid-filled and collapsed. Intermediate forms are un-stable and tend immediately to either expand or collapse. Increased variation in radiographic density with respiration in pulmonary edema can be attributed to such quantal alveolar behaviour: the more severe is the edema, the greater is the variation. It has also been observed that variation in the radiological

density during the respiratory cycle is increased in patients with pulmonary edema compared with normal subjects [31].

Another feature of the chest radiograph which make such an interpretation possible is that as EVLW increases the radiographic densities occupy a greater fraction of the total lung airspace [32]. Furthermore as edema worsens and water displaces air in any given lung region, the density of the infiltrate also worsens, becoming more and more white. Nevertheless accuracy, meant as the amount of EVLW present, is significantly limited by acquisition techniques and clinical issues that override standardization procedures especially in the critically ill [33]. Under clinically relevant conditions, the correlation of EVLW by chest radiography to other established techniques has been weak [34].

2.2.2

Computed tomography

The principle advantages of using X-ray computed tomography (CT) over con-ventional radiography are that the density of the infiltrates can be determined quantitatively and the spatial distribution of edema in transverse sections can be defined. Lung density can be quantified with X-ray CT because the arbi-trary Hounsfield units used for CT display can be calibrated against objects or substances of known density. Experimentally a direct proportionality was found in CT-derived estimates of lung density increased by 69% when gravi-metric measurements of lung weight increased by about 250% [35]. Obviously such procedure has the disadvantage that it is not portable and involves ex-posure to ionizing radiation.

2.2.3

Nuclear magnetic resonance (NMR) imaging

Another emerging approach to estimating lung water content is based on the ability to align hydrogen nuclei (protons) of water in the direction of an exter-nally applied magnetic field [36]. When a subject lays within a magnetic field and is then irradiated with electromagnetic radiation, resonance develops from the absorption and subsequent release of energy as the radiofrequency stimulus pulse is applied and discontinued. This energy can be detected with appro-priately placed antennas, producing a signal of varying strength, depending on the strength of the magnetic field and the frequency of the radiofrequency pulse. The spin-echo stimulus pulse sequence is the only one to date that has been employed to measure lung water. Numerous studies have reported a good correlation between NMR-determined estimates of lung water and esti-mates from the gold-standard gravimetric method [37, 38, 39, 40, 41, 42]. An intrinsic issue related to NMR imaging is that normal or mildly edematous lung produces relatively little signal using conventional spin-echo sequences on 1.5 Tesla imagers typically used for clinical purposes [36, 41] causing the underestimation of true lung water in absolute terms by as much as 20-40% [43].

2.2.4

Positron emission tomography

Lung water can also be measured by external residue detection techniques, af-ter separately adminisaf-tering radioactively labeled tracers that distribute within the total and intravascular water spaces of the lung. Emissions are then de-tected with a device such as a gamma camera or a PET scanner. PET is

widely held to be the gold standard for measuring EVLW (amongst the nu-clear medicine techniques) because a tomographic image can then be created and normalized for the attenuation of the structure being imaged using a trans-mission scan [44]. Lung water content can be measured either directly, or esti-mated from tissue density measurements [45]. With this approach, the water fraction of lung tissue must be assumed (0.82-0.84 ml/g). A small correction of about 2% for differences in tissue versus blood density can also be incorporated [46]. When lung water is measured directly instead of lung density, a sample of sterile water is labeled with a positron-emitting isotope, such as oxygen-15 (H215O) (half-life = 2.06 min), and then administered intravenously.

Inaccura-cies from areas of hypoperfusion are made less significant since the15O labeled

water is allowed to equilibrate within tissue water over a 3-4 min period. If the activity data in the PET image are scaled to simultaneously obtained activity in the blood, the image can be displayed as a quantitative regional map of lung water distribution [47]. An analogous approach is used to measure the blood volume concentration in the images using15_{O (or, alternatively,}11_C)

la-beled carbon monoxide instead of 15_{O water. If} 15_{O carbon monoxide is used,}

trace amounts of C15O are inhaled as a gas, binding immediately to blood

hemoglobin. After a few minutes, to allow equilibration within the body’s blood volume, another PET scan is obtained. When again normalized to ac-tivity measurements in blood and corrected for attenuation, the image is a regional display of blood volume. With the assumption that 84% of blood at normal hematocrits is water, the blood water content in a lung region can be subtracted from the total lung water concentration, yielding a derived image of extravascular water concentration [48]. The total time required to measure EVLW with PET is about 45 minutes, but repeat measurements can begin

in as little as 10-15 minutes from the previous one. EVLW measurements by PET correlates well with those obtained by gravimetrics (r = 0.86-0.93), even though corrections for potential differences in peripheral versus capillary hematocrit, or for differences in tissue versus blood density were not included into the clinical records [48, 49]. Perhaps because such corrections were not in-corporated, PET underestimated EVLW by about 10-15% systematically with respect to the gravimetric estimates. Advantages of PET measurements are high reproducibility (coefficient of variation for repeat measurements < 5%) and linearity. The method also shows exquisite sensitivity: as little as 1 ml additional extravascular water can be detected with PET [49]. Despite these impressive performance characteristics, take into account that PET imaging is too expensive, like NMR, and not widely distributed among medical cen-ters. Also positron-emitting isotopes produce ionizing radiation although the amounts used in any one study are quite low.

2.2.5

Electrical impedance tomography (EIT)

Air and liquid offer different resistances to the flow of electricity through the body. Measuring thoracic bioelectrical impedance in response to a low ampli-tude alternating electric current passed through the body yields a value for resistivity which can be correlated to gravimetric EVLW after correction for weight [50, 51, 52]. In transthoracic bioetectrical impedance analysis (BIA) an alternating electrical current is passed through biologic tissue and the resis-tance to that current measured [53]. This resisresis-tance is inversely proportional to the amount of water contained by the tissues within the electric field. By measuring the electrical resistance of the right lung it is possible to calculate

a standardized index of right lung water termed resistivity which correlates with both total lung water and ext ravascular lung water obtained by direct postmortem gravimetry [52]. Whole body measurements are done using stan-dard right-sided, tetrapolar Ag/AgCl spot ECG electrode placement to deliver an imperceptible constant 800 pA alternating current at 50 KHz frequency at room temperature [54]. One of the measuring protocols uses two distal sig-nal source electrodes placed on the dorsal surfaces of the right hand and foot just proximal to the metacarpal-phalangeal and metatarsalphalangeal joints, respectively and voltage sensing detector electrodes placed between the medial and lateral malleoli at the ankle and at the pisiform prominence on the wrist. Whole body data are collected through one analyzer software package that uses measured resistance and reactance values to calculate total body, intracellular and extracellular water based on proprietary equations. Such transthoracic BIA measured changes in lung water appeared to provide worthwhile addi-tional information to body weights, chest exam and fluid balance by both mon-itoring the dynamic changes that occurred during treatment and by helping to establish meaningful therapeutic endpoints. Transthoracic bioimpedance measurements of lung water are simple to perform, can be monitored con-tinuously, and yield objective data [52]. Although less precise than invasive measurements of lung water, the growing awareness of the morbidity and pos-sible mortality that may result from invasive monitoring suggests that using a less precise but safer noninvasive technique may be preferable [55].

2.2.6

Double Indicator dilution and thermodilution

methods

The indicator dilution technique is an invasive procedure to measure EVLW even at bedside. It is based either on the so-called “mean transit time” or “slope-volume’” approaches to analyze the temperature-time or concentration-time data [56, 57]. In the indicator dilution method the evaluation of volumes is based on the simultaneous administration of a freely diffusible (cold bolus of saline or glucose solution) and a non-diffusible (indocyanine green dye) indi-cator. While the indicators distribute within the intra and extravascular com-partiments, it is possibile to evaluate intrathoracic compartimental volumes, although each have the same flow but through different volumes of distribution within intravascular (indocyanine green dye) and both intravascular and ex-travascular spaces (cold bolus of saline or glucose solution). The difference in the mean transit times of the two indicators is therefore extravascular thermal volume (ETV). In the slope-volume method, a slope for the linear decay of the thermodilution curve is determined by mixing within the largest volume through which the thermal indicator passes (lungs). When multiplied by the cardiac output, pulmonary thermal volume (PTV) can be calculated. Further correction for intrathoracic blood volume yields a value for EVLW. This can be achieved through injection of a single thermal indicator, obviating the need to use indocyanine green dye [58, 59]. The execution of this technique needs both a central venous catheter and a catheter inside a big arterial vessel (prefer-ably the femoral artery) to be placed. The latter is a thermistor tipped optic fibre catheter for the colorant marker dilution analysis. The measurement is made by a 0,3 mg/kg of indocyanine green dye injection as quickly as possibile

into a given little volume of 5% glucose solution at a temperture lower than 10°C within the central venous catheter or into the proximal opening of the Swan-Ganz catheter [9]. The dilution curves of the indicators are recorded simultanously from the descending aorta by the thermistor at the tip of the optical fiber inside the artery. EVLW is estimated through a simple calcu-lus from these curves. Since the extravascular water content of myocardium and non-pulmonary blood vessels is small relative to the extravascular water content of the lung, ETV and EVLW are usually considered to be equiva-lent. Many studies have shown that ETV usually closely approximates EVLW [57, 60]. The thermal capacitance of the non-aqueous structures may, however, be significant, leading to overestimates of EVLW of 10-15% [61]. At least two previous studies [56, 57] have both pointed out that the measurement of ETV is only equal to EVLW if the relative transit times of the thermal indicator through red cells versus plasma, the relative specific heats of extravascular tis-sue versus plasma, the density of blood, and the fraction of extravascular mass represented by water are all taken into account. Without such corrections, ETV should consistently overestimate EVLW by as much as 24% in normal lungs. Interestingly, as the lungs become more edematous, a greater fraction of the extravascular mass becomes water, and the error introduced by ignoring these factors, which is the case with commercially available devices, actually goes down. The advantages of measuring EVLW by the single or double indi-cator dilution methods are several; the methods are simple to implement, safe, reproducible, and repeatable. On the other hand, although they are consid-ered gold standard procedures for EVLW assessment they are not often used in clinic because somewhat invasive since they require central venous as well as arterial catheterization. In addition, the accumulation of extravascular water

in any portion of lung that is downstream from a large vascular obstruction cannot be detected [57]. An analogous problem exists for lung regions that are simply poorly perfused, for instance as a result of using positive end expiratory pressure (PEEP) [56, 57, 62].

2.2.7

Discussion

None of the listed methods for measuring EVLW, other than chest radiog-raphy, have been widely incorporated into clinical practice. One reason is undoubtedly that no one has shown that a measurement of EVLW per se is needed for sound clinical decision making during the treatment of pulmonary edema. Similarly, no one has shown that incorporating such methods into rou-tine clinical practice will affect patient outcome. Although the potential value of having a quantitative measure of pulmonary edema seems obvious, such as a treatment endpoint surrogate for mortality in clinical trials, and various studies have suggested how such measurements might be used in clinical de-cision making [61], a convincing outcome study demonstrating benefit is still needed. The standard chest radiograph remains the best screening test for the detection of pulmonary edema, but it requires radiology facilities, specific reading expertise, it uses ionising energy, and poses a significant logistic bur-den. ULCs assessment provides an appealing simple, non-radiologic and low cost bedside alternative to available methods and it appears to be reasonably well correlated with extra-vascular lung water assessed by chest radiography [63] and other highly complex methods such as computerized tomography and thermodiluition techniques [64]. It is theoretically appealing for detecting and quantifying extravascular lung water, a key parameter in the serial evaluation

of the cardiologic patient with heart failure [65].

2.3

Chest sonography

It has been recently shown that lung ultrasound (LUS) may represent a sim-ple and useful tool for the semi-quantitative evaluation of many pulmonary conditions in cardiovascular disease, especially to assess pulmonary congestion in patients with heart failure. However this procedure is still relatively new and modestly experienced thus it is not universally acknowledged and applied. Evaluation of the lung has been generally considered off-limits for ultrasound [3] since it is standard textbook knowledge that ultrasound energy is rapidly dissipated by air [66]. This is why ultrasound is not considered useful for eval-uating the pulmonary parenchyma in physiologic conditions, where the high difference in acoustic impedance between air and surrounding tissues deter-mines the complete reflection of the ultrasound beam and does not allow any image to be visualized. The figure 2.3.1 shows in the upper side an operator during a lung ultrasound acquisition; the underlying images represent how a normal lung looks like in comparison to one involved with pulmonary edema; on their left and right it is possible to see, respectively, drawings of the reflec-tion of the ultrasounds in the two cases and of the genarareflec-tion of the scattering artifact, the so-called “comet-tail”. The presence of extra-vascular lung water (EVLW) or interstitial fibrosis opens the previously locked pulmonary acoustic window, since water- or collagen-thickened interlobular septa create the ade-quate acoustic mismatch able to trigger the phenomenon of sonographic re-verberation that generates B-lines. Multiple B-lines are present in pulmonary congestion, and may help in the detection, semiquantification and monitoring of extravascular lung water, in the differential diagnosis of dyspnea, and in the

prognostic stratification of chronic heart failure and acute coronary syndromes [67, 68].

Figure 2.3.1: A picture from a lung ultrasound acquisition and some drawings representing the comparative reflection pattern of an ultrasound beam.

In patients with cardiogenic pulmonary edema, semiquantification of disease severity may be obtained by evaluating the number of B-lines as this is di-rectly proportional to the severity of congestion. In patients with cardiogenic pulmonary edema, B-lines should be evaluated because it allows monitoring of response to therapy. In patients with increased extravascular lung water, assessment of lung reaeration can be assessed by demonstrating a change (de-crease) in the number of B-lines [4].

EVLW assessment method

B-lines obtained with non-ionizing, patient-friendly lung ultrasound (LUS) evaluation, have been recently proposed as an emerging echographic marker for the assessment of pulmonary interstitial syndrome, including pulmonary congestion, after many studies have confirmed the doctor Picano’s pioneering intuitive idea developed within the Institute of Clinical Physiology of the CNR [63, 69]. Scientific litterature quotes the first identification of ultrasound lung comets (ULCs) is unconnected to lungs; instead it goes back up in 1981 when an ultrasound scan of a human liver nestled with metal bullets was described. At a naked-eye examination ULCs appear like an echoes array which globally resambles a comet tail [70]. One year later a study revealed an explanation in theory to such phenomenon, testing some hypothesis related to ultrasound physics, especially to the so-called reverberation [71]. Other studies about ULCs, dated from 1986, describe them like a set of as close as indistinguish-able echoes. As a whole, these echoes generate a structure very similar to the tail of a comet that is why it is called “comet-tail” artifact. This study supported that ULCs are artifacts rising after reverberation of the ultrasound beam [72]. Water- or collagen-thickened interlobular septa create the adequate

acoustic mismatch able to trigger the phenomenon of sonographic reverbera-tion that generates ULCs. Pulmonary interstitium, divided into a central and a peripheral area, was branched into three compartments:

1. one axial or central compartment, pertinent to main vessels and air paths;

2. peripheral compartment, which covers the pleura and forms interlobular septa that isolate the minor lobes;

3. parenchymal compartment, enclosing alveolar ducts and producing a con-nective tissue net that widely links axial and peripheral compartments [73].

Figure 3.0.1: B-lines demonstration in precence of interstitial lung disease (cardiac pulmonary edema). A: linear probe scan; B: convex probe scan.

Comets arise from the peripheral interstitial compartment where the interlob-ular septa divide the secondary lobes. Each secondary lobe is the smallest portion of lung surrounded by connective tissue; it has an irregular polyhedral shape and a variable diameter from 1 up to 2.5 cm. Interlobular septa are

well-developed at the periphery of the lung where they connect and join with the vascular layer of the pleura (see figure 3.0.2).

Figure 3.0.2: Hystologic appearance of a secondary lobe.

Interlobular septa include pulmonary veins and lymphatic vessels which drain the conterminous lobular tissue and are bounded on both sides from a thin elas-tic layer of tissue which continues with the inner elaselas-tic pleural leaf. Arterial pulmonary and bronchiolar distal branches are located in the central section of lobules so they never reach the surface, instead the pulmonary venules and lymphatic vessels are place at the periphery of the lobule thus they can flow along interlobular septa. They are numerous in the apical, frontal and lateral regions of the upper lobe and in the frontal and lateral regions of the right middle lobe, in the lingula and in the inferior lobe. Here they are 100 micron thick and they can be identified by a naked-eye inspection of the pleura, of the surface of the lung or by a HRCT scan. In the middler regions of the lung the septa may be ponderously identified and they usually are absent at all. The lung ultrasound scans only permit the assessment of the most ex-ternal portion (the cortical region) of the lung, with the exception of lungs damaged by atelectasis (atelectasis is defined as the collapse or closure of the lung resulting in reduced or absent gas exchange due to a deflation of the

alve-oli). The pulmonary cortex is the surface of the lung (3-4 cm thick) and it is mainly made up of secondary lobules. A healthy secondary lobule can’t be assessed with thorax radiography. Interlobular septa may become visible as demarcated lines after their own thickening due to the precence of liquid (in case of pulmonary edema) or tissue (in case of interstitial lung disease). In this event, secondary lobules can be assessed among two of these lines within the lung. The thickening of the interstitium causes a sort of radiologically distinguishable opacity although it might depend on different reasons: leakage of edematous liquid, inflamatory exudation, obstruction of lymph vessels and more. When this happens, chest radiography reveals linear opacities, the so-called Kerley lines or B-lines [75].

Figure 3.0.3: Pleural line representation and its horizontal parallel reverberation lines.

These peculiar lines are peripheral, they rise from the base and plumb spread from the pleural surface. The pleural line is a hyperechogenic line visible be-tween two ribs and half a cm lower. It shows the lung-wall interface, i. e. the interface between chest wall and lung surface [68]. Two opposed types of artifacts arising from the pleural line can be differentiated. One type is hori-zontal, the other vertical. The “horizontal artifact” may be a convenient term

for the repetition of the pleural line reverberating at regular intervals, yielding parallel, roughly horizontal hyperechogenic lines. The figure 3.0.3 shows the pleural line (large arrows) visible between two ribs and roughly horizontal par-allel reverberation lines (small arrows). The distance between two horizontal lines is equal to the distance between the skin and the lung surface. The same pattern was observed in normal subjects. Comet-tail artifacts are roughly ver-tical hyperechogenic narrow-based repetition artifacts, as shown in the figure 3.0.4.

Figure 3.0.4: Multiple comet-tail artifacts arising from the pleural line in a patient with pulmon-ary edema.

The comet-tail artifact described here extends to the edge of the screen (whereas short comet-tail artifacts may exist in other regions), and arises only from the pleural line. Comet-tail artifacts arising from the pleural line can be localized or disseminated to the whole lung surface, or again isolated or multiple. Lung sliding is a to-and-from movement observed at the level of the pleural line synchronized with respiration. B-lines can be evaluated with both cardiac and abdominal ultrasound (US) linear or convex probe mounted also on protable devices. The procedure setting is easy to learn, a few hours training could be enough, and it can be performed also by non-medical staff [76]. It was found

that ULCs are equally reliable in the hands of highly experienced echocardi-ologist using expensive and complex technology and in the hands of absolute beginners with hand-held echocardiography. ULCs are easy both to obtain and to measure (learning curve of <10 examinations) and fast to perform (<3 minutes), require very limited technology, even without a second harmonic or Doppler and are not restricted by cardiac acoustic window limitations or pa-tient decubitus. The main limit to this method is that B-lines are not specific for EVLW since they are a sign of thickened pulmonary interstitium and thus they are not feasible in case of interstice diffuse (spread, scattered) disease. In an emergency B-lines can give useful information very timely and also they have a negative predictive value very close to 100% to exclude the presence of pulmonary edema [77]. Modern lung ultrasound is mainly applied not only in critical care, emergency medicine, and trauma surgery, but also in pulmonary and internal medicine. The most interesting advantage of this technique is its extreme simplicity of execution.

Figure 3.0.5: Four patients with different ULCs presence, that may be classified (scored) into four levels of severity: absence (normal lung, healthy control), mild, moderate and severe.

The figure 3.0.5 shows one clinical case: with chest sonography, the normally air inflated lung is black, moderately diseased lung (with interstitial water) is black and white, with white lines corresponding to lung comets, and markedly

diseased lung (with alveolar edema) is white (diffusely bright). The normal signal is no signal at ultrasounds and the echo image shows only a horizontal hyperecogenic line representing the pleural line. (figure 3.0.6/A, red arrow).

Figure 3.0.6: A: the red arrow pinpoints the pleural line, that is clearly visible with ultrasounds in the normal lung inflated with air; B: the white arrows highlight the presence of B-lines due to the impedance difference between air and water.

As the accumulation of EVLW increases and subsequentially the quantity of air decreases, vertical hyperecogenic bands show upon the image; they are the so-called B-lines or ultrasound lung comets, whose number increases along with the quantity of EVLW, giving the visualized pulmonary tissue a whiter appearance (figure 3.0.6/B, white arrows). The abnormal signal mirrors water accumulation can be scored on the basis of the number of comets on the chest which is also directly related to an independent assessment of extravascular lung water by semi-quantitative chest X-ray [63] or invasive thermo-dilution method [64].

ULCs are based on the principle that ultrasound is reflected by an interface between different acoustic impedance. In normal conditions, the ultrasound beam finds the lung air, that means high impedance and no acoustic mismatch on its pathway through the chest [69]. In the presence of extravascular lung

water, the ultrasound beam finds sub-pleural interlobular septa thickened by edema, that means a low impedance structure surrounded by air and with a high acoustic mismatch. The reflection of the beam creates a phenomenon of reverberation. The result is a wedge – shaped signal with a narrow origin in the near field of the image [68].

3.1

B-lines ecographic and physical development

principles

The initial reflection of almost the whole ultrasound beam, the so-called “acous-tic blockage” produced from the air to the ultrasound waves, prevent leftover waves from remaining available to further reflection, below the first air - soft tissue interface. The loss of ultrasound energy achieved as the ultrasound waves cross the medium, including the absorption and dispersion phenomena, is known as “attenuation”. The halfpower distance is the parameter used to express the entity of the attenuation of the ultrasound waves across the tissues. It expresses the distance that an ultrasound wave can cover across a specific tissue or medium until its width reduces half of its value in coincidence with the site of first impact. This value is also function of the frequency of the beam. The halfpower distance of the air at 2 MHz is equal to 0.08 cm, while the halfpower distance of water is equal to 380 cm at the same frequency. This difference explains why within a normally inflated lung, half of the available ultrasound energy is already halved after less than 1 mm. Some otherwise hidden structures are made recognizable when pulmonary diseases open a sort of acoustic window allowing a deep assessment of the pulmonary tissue. In fact, only when air is excluded from the pulmonary tissue ultrasounds may

be used to visualize it even more clearly as the quantity of air inside the lung dicreases.

The physical origin of the lung comets is due to the thickening of the inter-lobular interstitium of the secondary lobules under the pleura, as it is already confirmed and well acknowledged from CT scans. An increase in the thickness of these structures causes a difference between the acoustic impedance values of the air versus the tissue. As a consequence, the ultrasound beam is reflected and a first echo spawns immediately under the pleura. The complexion of the comets phenomenon depends on another physical circumstance that is the riverberation. When the ultrasound beam is reflected from an interface and comes back to the transducer, the transducer itself acts as a secondary reflect-ing surface. Here, the rebound ultrasound waves are reflected one more time, they travel through again their primary path, hit the interface again and then they finally come back to the transducer. Hence the selfsame ultrasound waves beam produces a new signal detached from the transducer twice the distance covered by the original echoes. The echoes producing this extra second signal, the so-called reverberation echoes, can be weaker than the original echoes [78]. The drawing in figure 3.1.1 shows how the reverberations are induced. The ul-trasound beam leaves the transducer, hits one side of the container and comes back to the transducer. The rebound ultrasound waves produce the echo “B”. Some of the rebound waves are riflected from the transduced and go back on the container. Here the transducer reflects for the second time the rebound waves generating a weaker echoe B’, that is the reverberation. This may cause the appearance of an another reflecting interface far from the transducer twice the distance between the transducer itself and the outermost side of the con-tainer.

Figure 3.1.1: Representation of an ultrasound beam backscatter produced inside a test tube.

A reverberation is achivable at any interface where a very high difference in the acoustic impedance occurs. Any interface dividing two different media is char-acterized by the so-called reflection coefficient which depends on the difference in the acoustic impedance values of those media. The acoustic impedance of the air is about 0.42 m−2 _{kg s}−1_{, a considerably lower value with respect to}

the values respectively related to the bone (7350 m−2 _{kg s}−1_{), the parenchyma}

(1732,5 m−2 _{kg s}−1_{) and the water (1554 m}−2 _{kg s}−1_{). This is to explain why}

the reflection coefficient of an air – soft tissue interface is extremely high and almost the whole acoustic wave is reflected as a consequence. Such hypothesis about the phyical genesis of the ULCs was confirmed in 1982 by an analysis of ultrasound patterns belonging to different subjects of different dimensions and materials and to a liver of a dog in which two 2 mm diameter bullets had been shot [71]. Both a 3.5 and a 5 MHz probes were employed to acquire the ultrasound signal from the phantoms plunged into the water. Usually the reflection of the reverberation echoes gets greater as the difference of the acoustic impedance increases and it is due to many variables such as the shape, the dimension, the composition and the orientation of the ultrasound beam.

The images derived from the liver expressly nested with the bullets are very similar to those previously described by Wendell [70] one year before by the depiction of a bullet causin the so-called “comet-tail” artifact. The analysis of the other objects confirm that the echoes get stronger (more intense) as the difference of the acoustic impedance between the object itself and the environ-ment increases, the more the object is perpendicular to the ultrasound beam, the bigger are the dimensions of the object and the more acoustic interfaces the beam comes across.

The sub-pleural side of a thickened pulmonary septa is too thin to be visual-ized through the reflection of an ultrasound beam [67] but adequate to interfere with it by reason of the difference of acoustic impedence with respect to the surrounding air otherwise absent in normal septa. This primary reflection of the ultrasound beam induces a reverberation phenomenon characterized by multiple reflections separated by a time interval that can be interpreted as a distance. All those multiple reflections (reverberations), being very close one to each other, appear in the whole like a featured comet-tail shaped band. The figure 3.1.2 explains the mechanism just described.

Figure 3.1.2: EWLV-thickened interlobular septa cause a different reverberation pattern of the ultrasound beam.

A healthy and normally inflated lung generates a classic ultrasound image with a clearly visible pleural line whose thickness is proper to create a difference of acoustic impdance able to reflect an ultrasound beam even in a normal condition. The pleural line is a primary reference in the thoracic echography. The pleura is less than 2 mm thick and its radiological appearance is bright (white) and regular. Below the pleural line, a peculiar echogenicity can be visualized by an ultrsound scan. It is possibile to distinguish (figure 3.1.3) horizontal lines steadily and equally spaced becoming less bright while relieving from the pleura: they are the reverberations of the pleural line itself, also called from authors “A-lines”.

Figure 3.1.3: Ultrasound image of a normal lung: note the pleural line and its equally spaced reverberations becoming less bright while relieving from it.

An illustrative example can be made with the following image, where it is possible to distinguish three different compartments: the air into the normally inflated lung, the water into the lungh with congestion and the connectival tissue settled in the fibrotic lung.

Figure 3.1.4: (A) healthy lung inflated with air up to 81% and its related CT, echographic and hystological representations. (B) edematous lung with increased quantity of ex-travascular water an its related CT, echographic and hystological representations. (C) fibrotic lung with an increased quantity of connectival tissue its related CT, echographic and hystological representations.

LUS examination can be performed using any commercially available 2-D scan-ner, with any transducer (phased-array, linear-array, convex, microconvex).

Figure 3.1.5: An example of 2D (bidimensional) probe.

The examination can be performed with any type of echographic platform, from fully equipped machines to pocket size ones [76]. Patients can be in the near-supine, supine or sitting position, as clinically indicated [69]. All the chest can be easily scanned by ultrasound, just laying the probe along the intercostal spaces. However, some specific methods have been proposed: ultrasound scanning of the anterior and lateral chest may be obtained on the right and left hemithorax, from the second to the fourth (on the right side to the fifth) intercostal spaces, and from the parasternal to the axillary line [63]. Usually the most popular probes exploited during an echographyc lung scan are those with a lower acoustic frequency, from 2.5 and 3.5 up to 5 MHz, able analyse the most deep structures but with a lower spatial resolution and the linear probes with a higher acoustic frequency, from 7.5 to 10 MHz, able to supply more detailed images but only of superficial structures.

3.2

B-lines clinical develompent principles

Ultrasound lung comets occur in well established clinical conditions, which may be classified into cardiac and non-cardiac. All the clinical cardiac condi-tions leading to the appearing of ULCs are characterized by an increase of the

filling pressures of the left ventricle with a further consequential transmission of the hydrostatic pressure, accordingly with the Starling equation, which may produce a leakege of some liquid into the pulmonary interstitum. Heart failure, uncontrolled arterial hypertension and aortic and mitral valve dysfunction be-long to the clinical cardiac conditions class. The non cardiac clinical conditions are mostly of pulmonary origin: diffuse interstitium disease with sediment of collagene may cause bilateral diffuse comets; inflamatory conditions like pneu-monia and bronchial pneupneu-monia may cause localized comets into the region of interest.

Figure 3.2.1: Differential diagnosis of different pathologies which may cause the presence of lung comets.

3.3

Ultrasounds theory

Ultrasound is an oscillating mechanic (pressure) wave with a frequency greater than the upper limit of the human hearing range. The upper frequency limit in humans is due to limitations of the middle ear, which acts as a low-pass fil-ter. Ultrasound is thus not separated from audibel sound based on differences in physical properties, only the fact that humans cannot hear it. Although this limit varies from person to person, it is approximately 20 KHz in healthy, young adults. Ultrasound devices operate with frequencies from 20 KHz up to several GHz.

Ultrasounds are characterized by a frequency, a wave length, a propagation velocity, an intensity (measured in decibel - dB) and they usually undergo an attenuation process while they are transmitted through a medium due to its acoustic impedance. Ultrasonic devices are used to detect objects and mea-sure distances. Ultrasonic imaging (sonography) is used in both veterinary medicine and human medicine. Industrial employments of ultrasounds are cleaning, mixing and acceleration of chemical processes. Organisms such as bats and porpoise use ultrasound for locating prey and obstacles [79]. Ultra-sound can be used for medical imaging, detection, measurement and cleaning. Ultrasound used for medical purposes produces sound waves that are beamed into the body causing return echoes that are recorded to visuzalize structures beneath the skin. As any other oscillating wave ultrasounds can be reflected, refracted and diffracted. The ability to measure different echoes reflected from a variety of tissues allows a shadow picture to be constructed. The technol-ogy is especially accurate at seeing the interface between solid and fluid filled spaces.

of piezoelectric crystals located on the transducer surface. Electrical stimu-lation causes mechanical distortion of the crystals resulting in vibration and production of sound waves (i.e. mechanical energy). The conversion of elec-trical to mechanical (sound) energy is called the converse piezoelectric effect (G. Lippman, 1881). The first demonstration of the direct piezoelectric effect was in 1880 by the brothers Pierre and Jacques Curie [80]. They combined their knowledge of pyroelectricity with their understanding of the underlying crystal structures that gave rise to pyroelectricity to predict crystal behavior, and demonstrated the effect using crystals of tourmaline, quartz, topaz and Rochelle salt (sodium potassium tartrate tetrahydrate). Quartz and Rochelle salt exhibited the most piezoelectricity. They later observed the converse ef-fect [81], which was predicted from thermodynamic principles by Lippmann in 1881 [82].

Figure 3.3.1: The piezoelectric effect causes crystal materials like quartz to generate an electric charge when the crystal material is compressed, twisted or pulled. The reverse also is true, as the crystal compresses or expands when an electric voltage is applied.

The direct effect may be used in sensing applications, while the indirect effect may be used in actuation and acoustic transduction. The piezoelectric effect is

caused by an asymmetry in the unit cell and the resulting relationship between mechanical distortion and electric dipole separation. The effect may be quan-tified through the use of suitable piezoelectric coefficients, the measurement of which has been described in various standards on piezoelectricity [83]. The direct effect may be used in sensing applications, while the indirect effect may be used in actuation and acoustic transduction. The nature of the piezoelectric effect is closely related to the occurrence of electric dipole moments in solids. The latter may either be induced for ions on crystal lattice sites with asymmet-ric charge surroundings or may directly be carried by molecular groups. The dipole density or polarization (cm

m3) may be easily calculated for crystals by

summing up the dipole moments per volume of the crystallographic unit cell [84]. As every dipole is a vector, the dipole density P is a vector field. Dipoles near each other tend to be aligned in regions called Weiss domains. The do-mains are usually randomly oriented, but can be aligned using the process of poling, a process by which a strong electric field is applied across the material, usually at elevated temperatures. Not all piezoelectric materials can be poled [85]. Of decisive importance for the piezoelectric effect is the change of po-larization P when applying a mechanical stress. This might either be caused by a re-configuration of the dipole-inducing surrounding or by re-orientation of molecular dipole moments under the influence of the external stress. Piezo-electricity may then manifest in a variation of the polarization strength, its direction or both, with the details depending on:

1. the orientation of P within the crystal;

2. crystal symmetry;

The change in P appears as a variation of surface charge density upon the crystal faces, i.e. as a variation of the electric field extending between the faces caused by a change in dipole density in the bulk. Piezoelectric materials also show the opposite effect, called converse piezoelectric effect, where the application of an electrical field creates mechanical deformation in the crystal.

3.4

Ultrasonography

Diagnostic sonography (ultrasonography) is an ultrasound-based diagnostic imaging technique used for visualizing subcutaneous body structures including tendons, muscles, joints, vessels and internal organs for possible pathology or lesions. The practice of examining pregnant women using ultrasound is called obstetric sonography and is widely used. Ultrasound images (sonograms) are made by sending a pulse of ultrasound into tissue using an ultrasound trans-ducer, the probe. The sound reflects and echoes off parts of the tissue; this echo is recorded and displayed as an image to the operator. Many different types of images can be formed using ultrasound. The most well-known type is a B-mode image, which displays a two-dimensional cross-section of the tissue being imaged. Other types of image can display blood flow, motion of tissue over time, the location of blood, the presence of specific molecules, the stiffness of tissue or the anatomy of a three-dimensional region.

Ultrasounds can also be used therapeutically, to break up gallstones and kid-ney stones (lithotripsy) or to heat and destroy diseased or cancerous tissues. Compared to other prominent methods of medical imaging, ultrasonography has several advantages. It provides images in real-time, rather than after an acquisition or processing delay, it is portable and can be brought to a sick

patient’s bedside, it is substantially lower in cost, and it does not use harmful inonizing radiation. Drawbacks of ultrasonography include various limits on its field of view including difficulty imaging structures behind bone and its relative dependence on a skilled operator.

Typical diagnostic sonographic scanners operate in the frequency range of 2 to 18 MHz, though frequencies up to 50–100 MHz have been used experimentally in a technique known as biomicroscopy in special regions, such as the ante-rior chamber of the eye [86]. The choice of frequency is a trade-off between spatial resolution of the image and imaging depth: lower frequencies produce less resolution but image deeper into the body. Higher frequency sound waves have a smaller wavelength and thus are capable of reflecting or scattering from smaller structures. Higher frequency sound waves also have a larger atten-uation coefficient and thus are more readily absorbed in tissue, limiting the depth of penetration of the sound wave into the body. Sonography is effective for imaging soft tissues of the body. Superficial structures such as muscles, tendons, breast, thyroid and parathyroid glands, and the neonatal brain are imaged at a higher frequency, 7–18 MHz, which provides better axial and lat-eral resolution. Deeper structures such as liver and kidney are imaged at a lower frequency 1–6 MHz with lower axial and lateral resolution but greater penetration.

Ultrasonography is widely used in medicine. It is possible to perform both di-agnosis and therapeutic procedures, using ultrasound to guide interventional procedures (for instance biopsies or drainage of fluid collections). Sonogra-phers are medical professionals who perform scans which are then typically interpreted by radiologists, physicians who specialize in the application and interpretation of a wide variety of medical imaging modalities, or by

cardiolo-gists in the case of cardiac ultrasonography (echocardiography). Sonographers typically use a hand-held probe that is placed directly on the patient and moved over him/her. Increasingly, clinicians and other healthcare professionals who provide direct patient care are using ultrasound in their office and hospital practices, for efficient, low-cost, dynamic diagnostic imaging that facilitates treatment planning while avoiding any ionising radiation exposure.

3.4.1

Producing a sound wave

When a sound wave is produced, strong, short electrical pulses from the ultra-sound machine make the transducer ring at the desired frequency. The ultra-sound is focused either by the shape of the transducer, a lens in front of the trans-ducer, or a complex set of control pulses from the ultrasound scanner machine. The probe needs to be in contact with the surface of the body district un-der examination and a coupling gel has to be spread between the probe and the surface of the tissue. The ultrasound gel acts as both a lubricant and an energy conductor. This gel guarantees the further transmission of the sound wave from the sufrace of the probe through the body. Otherwise the air would prevent the wave to penetrate into the tissue and would dissipate all its energy. The reason why a coupling gel is needed is that all materials (human tissues in this case) will present an impedance to the passage of sound waves. The specific impedance of a tissue will be determined by its density and elasticity. In order for the maximal transmission of energy from one medium to another, the impedance of the two media needs to be the same. Clearly in the case of US passing from the generator to the tissues and then through the different tissue types, this can not actually be achieved. The greater the difference in impedance at a boundary, the greater the reflection that will occur, and

there-fore, the smaller the amount of energy that will be transferred. The difference in impedance is greatest for the steel/air interface which is the first one that the US has to overcome in order to reach to body. To minimise this differ-ence, a suitable coupling medium has to be utilised. If even a small air gap exists between the transducer and the skin, the proportion of US which will be reflected approaches 99.998% which in effect means that there will be no transmission.

Figure 3.4.1: More than 99.998% of ultrasound rays is reflected at the tissue/air interface due to the high difference of impedance (impedance gap) between the two means. Trans-mission is made possible throughout the application of a coupling gel over the US probe surface tip in contact with the tissue to scan.

The primary job of the coupling medium is to facilitate transmission of the ultrasound energy from the machine head to the tissues. Given an ideal cir-cumstance, this transmission would be maximally effective with no absorption of the US energy, nor any distortion of its path, etc. This ideal is almost im-possible to achieve, but the type of coupling medium employed does make a

difference. Furthermore the ultrasound beam needs to be focused; focusing is an adjustment proceduce which produces an arc-shaped sound wave from the face of the transducer. The wave travels into the body and comes into focus at a desired depth.

Figure 3.4.2: Medical sonographic instrument.

Older technology transducers focus their beam with physical lenses. Newer technology transducers use phased-array techniques to enable the sonographic machine to change the direction and depth of focus. The main advantage of using phased-array probes is the possibility to adjust depth and number of focuses (that is also possible to realize with electronic probes also) and the angle of incidence of the beam without moving the probe.

efficiently into the body (usually seeming to be a rubbery coating, a form of impedance matching). Almost all piezoelectric transducers are made of ce-ramic. The sound wave is partially reflected from the layers between different tissues. Specifically, sound is reflected anywhere any density change in the body occurs: e.g. blood cells in blood plasma, small structures in organs, etc. Some of the reflections return to the transducer. The percentage of re-flected wave carries the information about the difference of acoustic impedance between two tissues isolated from a surface. This is equal to R = (z1−z2)2

(z1+z2)2.

3.4.2

Receving the echoes to generate (reconstruct) and

display the images

The return of the sound wave to the transducer results in the same process that it took to send the sound wave, except in reverse. The rebound sound wave vibrates the transducer, the transducer turns the vibrations into electri-cal pulses that travel to the ultrasonic scanner where they are processed and transformed into a digital image.

The sonographic scanner must determine three things from each received echo:

1. how long it took the echo to be received from when the sound was trans-mitted;

2. how long it took the echo to be received from when the sound was re-flected;

3. how strong the echo was.

It could be noted that sound wave is not a click, but a pulse with a specific carrier frequency. Moving objects change this frequency on reflection, so that

it is only a matter of electronics to have simultaneous doppler sonography. Once the ultrasonic scanner determines these three things, it can locate which pixel in the image to light up and to what intensity and at what hue if fre-quency is processed. Transforming the received signal into a digital image may be explained by using a blank spreadsheet as an analogy. First picture a long, flat transducer at the top of the sheet. Send pulses down the columns of the spreadsheet (A, B, C, etc.). Listen at each column for any return echoes. When an echo is heard, note how long it took for the echo to return. The longer the wait, the deeper the row (1,2,3, etc.). The strength of the echo determines the brightness setting for that cell (white for a strong echo, black for a weak echo, and varying shades of grey for everything in between.) When all the echoes are recorded on the sheet, we have a greyscale image.

Images from the sonographic scanner can be displayed, captured, and broad-cast through a computer using a frame grabber to capture and digitize the analog video signal. The captured signal can then be post-processed on the computer itself.

3.4.3

The ultrasound waves inside the human body

Whenever a sound wave encounters a material with a different density (acous-tic impedance), part of the sound wave is reflected back to the probe and is detected as an echo. The time it takes for the echo to travel back to the probe is measured and used to calculate the depth of the tissue interface causing the echo. The greater is the difference between acoustic impedances, the larger is the echo. If the pulse hits gases or solids, the density difference is so great that most of the acoustic energy is reflected and it becomes impossible to see

deeper.

The frequencies used for medical imaging are generally in the range of 1 to 18 MHz. Higher frequencies have a correspondingly smaller wavelength, and can be used to make sonograms with smaller details. However, the attenuation of the sound wave is increased at higher frequencies, so in order to have better penetration of deeper tissues, a lower frequency (3–5 MHz) is used.

Seeing deep into the body with sonography is very difficult. Some acoustic energy is lost every time an echo is formed, but most of it (approximately 0.5_{cm∗depth*MHz}dB is lost from acoustic absorption. The speed of sound varies as it travels through different materials, and is dependent on the acoustical impedance of the material. However, the sonographic instrument assumes that the acoustic velocity is constant at 1540 m

s. An effect of this assumption

is that in a real body with non-uniform tissues, the beam becomes somewhat de-focused and image resolution is reduced.

To generate a 2D-image, the ultrasonic beam is swept. A transducer may be swept mechanically by rotating or swinging. Or a 1D phased array transducer may be used to sweep the beam electronically. The received data is processed and used to construct the image. The image is then a 2D representation of the slice into the body.

3D images can be generated by acquiring a series of adjacent 2D images. Com-monly a specialised probe that mechanically scans a conventional 2D-image transducer is used. However, since the mechanical scanning is slow, it is diffi-cult to make 3D images of moving tissues. Recently, 2D phased array trans-ducers that can sweep the beam in 3D have been developed. These can image faster and can even be used to make live 3D images of a beating heart. Doppler ultrasonography is used to study blood flow and muscle motion. The different

detected speeds are represented in color for ease of interpretation, for example leaky heart valves: the leak shows up as a flash of unique color. Colors may alternatively be used to represent the amplitudes of the received echoes.

3.4.4

Pros and cons of ultrasonography

Strenghts

1. It images muscles, soft tissues and bone surfaces very well and is partic-ularly useful for delineating the interfaces between solid and fluid-filled spaces;

2. it renders live images, where the operator can dynamically select the most useful section for diagnosing and documenting changes, often enabling rapid diagnoses. Live images also allow for ultrasound-guided biopsies or injections, which can be cumbersome with other imaging modalities;

3. it shows the structure of organs;

4. it has no known long-term side effects and rarely causes any discomfort to the patient;

5. small, easily carried scanners are available; examinations can be per-formed at the bedside;

6. relatively inexpensive compared to other modes of investigation, such as computer X-ray tomography or magnetic resonance imaging;

7. spatial resolution is better in high frequency ultrasound transducers than it is in most other imaging modalities.