Topic 8: Ultrasound Imaging PDF

Medical Imaging System Topic 8: Ultrasound Imaging By Dr. Mohammed A. Hassan  Reference Recommended text book Diagnostic Ultrasound: Physics and Equipment, 2nd ed., by Peter R. Hoskins (Editor), Kevin Martin (Editor), Abigail Thrush (Edit...

Medical Imaging System Topic 8: Ultrasound Imaging By Dr. Mohammed A. Hassan  Reference Recommended text book Diagnostic Ultrasound: Physics and Equipment, 2nd ed., by Peter R. Hoskins (Editor), Kevin Martin (Editor), Abigail Thrush (Editor) Cambridge University Press, 2010. 2  Objectives By the end of this topic, you should be able to: ▪ Define the term ultrasonography, its uses and advantages. ▪ Understand the concept of ultrasonography or echo ranging. ▪ Understand the physics of ultrasound waves. ▪ Understand the propagation of ultrasound beam and the lateral resolution. ▪ Know how to generate an ultrasound pulse, its characteristics and the axial resolution ▪ Understand how transducers work. ▪ Understand the function of each different parts of the ultrasonography system: beam forming, digitization, TGC, scan conversion and image formation. ▪ Understand the different modes of ultrasonography. 3  Introduction ▪ Ultrasonography is an imaging modality that utilizes certain range of sound wave called ultrasound: ▪ Different ranges of sound waves are illustrated in this figure In ultrasonography: ▪ A beam of sound waves is sent from the surface of the body. ▪ The image is reconstructed from reflections of ultrasound wave at tissue boundaries. ✓ So, the image is tomographic and contains morphological and structural information ✓ Functional imaging can be also obtained (blood flow for example) 4  Introduction Advantages and uses: ▪ Safe: neither ionizing radiation nor strong magnetic fields are required and hence, ✓ can be used for fetal imaging and measurements ✓ wide use in obstetrics and gynecology ▪ Real-time imaging and hence, ✓ can be used as functional imaging and for biopsy ✓ many cardiovascular applications ▪ Least expensive and most portable imaging modality ✓ handheld and cell phone-based units are currently available ▪ General imaging applications: ✓ liver cysts, aortic aneurysms and obstructive atherosclerosis in the carotids. 5  Concept of Ultrasonography ▪ An ultrasound beam is transmitted from external body surface to the internal tissue. ▪ Data are collected from the reflections of ultrasound wave at different tissue boundaries. ▪ In this figure, arrows represent reflected and transmitted waves at tissue boundaries. ✓ A boundary is a surface separating 2 media of different acoustic impedance. ▪ The time of detecting the echo can be used to calculate the Transducer distance at which the boundary is located ✓ That’s why it is called echo ranging and it is the concept upon which ultrasonography is based. ▪ So, different modes can be obtained: ✓ A-mode (signal) ✓ B-mode (image) ▪ Down-pointing arrows are parts of waves transmitted ✓ M-mode (motion image) to deeper tissues. ▪ Up-pointing arrows are the reflected waves (echo) ✓ Doppler and Power Doppler mode ▪ So, information related to structure could be obtained ✓ Color Flow Mapping Mode and other combined modes (tomogram) 6  Echo Ranging ▪ It is the detection of the distance from which echo is reflected 𝑡 back from a target. ▪ The amplitude of echo obtained at certain time depends on: 1. Range (distance) of the target from the transducer 2. Position and orientation of the ultrasound beam If a wave is transmitted at time 𝑡0 = 0 to a target at a distance 𝑑 from the transducer, then an echo from this target is detected after a time 𝑡 given by: 𝑡 = 2𝑑/𝑐 where 𝑐 is the speed of sound in the medium in which the 𝑑 Note the size of the arrows indicates the target is contained. intensities of the transmitted and reflected US beams: The beam is attenuated as being propagated 7  Ultrasound Imaging System The major parts of a basic ultrasound imaging system are shown in this graph. ▪ Starting from the Tx/Rx switch, the signal is transmitted or received from the transducer. ▪ The received signal is then conditioned with special amplification method because of different attenuations of the ultrasound signal. ▪ More processing is done before displaying the image. ▪ To understand the signals sent to or received from the transducer and why they need special amplification methods, the physics of ultrasound should be recalled. 8  Ultrasound Physics Ultrasound wave: is a pressure (mechanical) longitudinal wave: ✓ So, it needs a physical medium to propagate in. ✓ It is a longitudinal wave that oscillates in the direction of propagation. ✓ The frequency of ultrasound is above those of audible sounds (>20 kHz) ✓ The frequency of ultrasound used for medical imaging ranges from 2-15 MHz ▪ Like any wave, the US wave is characterized by frequency 𝑓, wavelength 𝜆 and phase shift 𝜑. As a pressure wave changing with time it is given by: 𝑝 = 𝑃𝑐𝑜𝑠(2𝜋𝑓𝑡 + 𝜑) ▪ The speed of ultrasound wave in medium depends on the characteristics of that medium and is given by: 𝑐 = 𝑘Τ𝜌 where 𝑘 is the stiffness (or Young’s modulus) 𝜌 is the density (attention: sometimes given in terms of compressibility). ▪ As 𝜌 of the tissue (medium) increases, the speed of sound decreases for const. stiffness. 9  Ultrasound Physics ▪ This means that different media have different acoustic impedance “𝑧” to sound propagation given by: 𝑧 = 𝜌𝑐 = 𝜌 𝑘 Τ𝜌 = 𝑘𝜌 ▪ This table shows the density, speed of sound and acoustic impedance of different tissues, air and water. ✓ Different soft tissues have sound speed and acoustic impedance close to that of water. ✓ The speed and impedance differ greatly in bone and air Material Density: 𝑔 𝑐𝑚−3 Speed: 𝑐(𝑚𝑠 −1 ) Impedance: 𝑧(𝑘𝑔 𝑚−2 𝑠 −1 ) Water 1 1480 1.48 × 106 Fat 0.95 1430 1.33 × 106 Blood 1.055 1575 1.66 × 106 Liver 1.06 1590 1.69 × 106 Muscle (along fiber) 1.065 1575 1.68 × 106 Bone: axial long. wave 1.9 4080 7.75 × 106 Air (at 20𝐶°) 344 430 10  Propagation of US Wave: Reflection In general, the angle of incidence 𝜃𝑖 equals the angle of reflection 𝜃𝑟. Reflection from large interface: ▪ Large interface occurs when the interface is large compared to 𝜆 of the sound wave and 𝜃𝑖 < 90∘. ▪ The Amplitude Reflection coefficient 𝑅𝐴 is given by: 𝑃𝑟 𝑍2 − 𝑍1 𝑅𝐴 = = 𝑃𝑖 𝑍2 + 𝑍1 where 𝑃𝑖 and 𝑃𝑟 are the amplitudes of the incident and reflected waves near the interface. ▪ The Intensity Reflection coefficient 𝑅𝐼 is given by: 𝐼𝑟 𝑅𝐼 = = 𝑅𝐴2 𝐼𝑖 Since 𝐼𝑖 = 𝐼𝑟 + 𝐼𝑡 , the Intensity Transmission coefficient 𝑇𝐼 is given by: 𝑇𝐼 = 1 − 𝑅𝐼 11  Propagation of US Wave: Reflection In general, the angle of incidence 𝜃𝑖 equals the angle of reflection 𝜃𝑟. Reflection from large interface: ▪ Large interface occurs when the interface is large compared to 𝜆 of the sound wave and 𝜃𝑖 < 90∘. ▪ The Amplitude Reflection coefficient 𝑅𝐴 is given by: 𝑃𝑟 𝑍2 − 𝑍1 𝑅𝐴 = = 𝑃𝑖 𝑍2 + 𝑍1 where 𝑃𝑖 and 𝑃𝑟 are the amplitudes of the incident and reflected waves near the interface. ▪ This means that the amplitude of the reflected wave (which is detected for imaging) depends on the acoustic impedances of the 2 media (2 tissues) separated by the surface and the amplitude of the incident. ✓ So, we can detect a differential signal to form an image 12  Reflection of US Wave: Scattering ▪ In the table, notice the value if the boundary is between bone and soft tissue (and also between air and soft tissues) ▪ Scattering of US wave ▪ When reflection occurs at very small interfaces (small target), it is called scattering ▪ Small targets (very small particles) make interfaces 𝐹𝑊𝐻𝑀 𝐹𝑊𝐻𝑀 Objects are just resolved in the image 𝑑 = 𝐹𝑊𝐻𝑀 2 wide beams (overlap above 0.5 of their max): 𝐹𝑊𝐻𝑀 Objects are not resolved in the image 𝑑 < 𝐹𝑊𝐻𝑀 24  Beam Width and Lateral Resolution ▪ If two small structures in the tissue are separated by a distance greater than the FWHM of the beam, then they can be resolved as separate structures and hence the 𝑙𝑎𝑡𝑒𝑟𝑎𝑙 𝑟𝑒𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛 = 𝐹𝑊𝐻𝑀 = 2.36𝜎 of the beam comes from one element in the array. ✓ The narrower the beam the better the lateral resolution ▪ Narrow beams come from narrow transducer (i.e of small radius 𝑟). But you know that sin 𝜃 = 0.61𝜆/𝑟, ✓ The narrower the beam the greater the angle of divergence which means ✓ less uniformity and more power losses (which is not recommended). ▪ So, for a practical plane disc source of a given frequency, the aperture is chosen to give the best compromise between beam width and beam divergence. ✓ Improvement (narrowing) to the beam width can be obtained by focusing ✓ Focusing is done by a concave transducer ▪ Focusing produces a concave wavefront that converges towards a focus and then diverge Focusing produces concave wavefront before the focus which diverge (convex) after the focus 25  Ultrasound Beam Focusing 𝑙 𝑓 No focusing: a disc transducer Focusing gets smaller near field length. 𝑓 < 𝑙 ▪ Beyond the focus, the waves become convex and the beam diverges again Focusing gets smaller near field length, but what about 𝜃? ▪ Strong Focusing is achieved when the focal length 𝑓 is given by: 𝑓 < 0.5𝑙 ✓ 𝜃(𝑤𝑖𝑡ℎ 𝑓𝑜𝑐𝑢𝑠𝑖𝑛𝑔) > 𝜃(𝑤𝑖𝑡ℎ𝑜𝑢𝑡 𝑓𝑜𝑐𝑢𝑠𝑖𝑛𝑔) which is not recommended. ▪ Optimum focusing is achieved when 𝑓 > 0.5𝑙 ✓ 𝜃 with and without focusing are comparable Focusing methods: ▪ Focusing by a curved source (a); ▪ Focusing by adding an acoustic lens to a plane source (b). (a) (b) ✓ Acoustic lens is made of materials of greater refraction indexes 26  Ultrasound Generation ▪ When a piezoelectric transduce is excited continuously with an oscillating signal, it produces a pressure wave at a specific frequency. ▪ Pulsed excitation gives a pressure wave that have different frequencies (should be similar to that of Sinc function) ✓ Because of interference between the components at different frequencies, the produced pulse is similar to that shown below. ✓ Transducers are characterized by a central frequency 𝑓0 and the band width (𝑓2 − 𝑓1 ) ✓ 𝑓1 and 𝑓2 are frequencies less and greater than 𝑓0 and have amplitudes relative to that of 𝑓0 of 3dB. ✓ Greater bandwidth means the A typical ultrasound pulse consists The pulse contains a range of frequencies transducer can work at multiple of a few cycles of oscillation: pulse nominated by a center frequency: pulse frequencies. represented in the time domain represented in the frequency domain 27  Ultrasound Pulse Important Definitions: ▪ The pulse spatial length PSL is given by: 𝑃𝑆𝐿 = 𝑛𝜆 where n is the number of cycles in the pulse duration ▪ The pulse duration PD is given by: 𝑃𝐷 = 𝑛𝜆/𝑐 ▪ The pulse average intensity 𝐼𝑃𝐴 the average value of intensity along the pulse duration) is given by: 𝑃𝐼𝐼 𝐼𝑃𝐴 = Side lobes 𝑃𝐷 where 𝑃𝐼𝐼 is the pulse intensity integral (integration of the waveform along the pulse duration). 28  Ultrasound Pulse ▪ The pulse repetition period 𝑇 is the time from the start of a pulse to the start of the next pulse. The pulse repetition rate = 1/𝑇 ▪ The temporal average intensity 𝐼𝑇𝐴 is the average intensity over the pulse repetition period (much smaller than 𝐼𝑃𝐴 because the pulse repetition period is approximately 1000 times the duration of the pulse itself (𝑃𝐷). ▪ 𝐼𝑇𝐴 indicates the rate at which energy deposition may accumulate in the tissue, resulting in heating effects. ▪ The highest value in the beam is the spatial peak intensity. ▪ The spatial average intensity 𝐼𝑆𝐴 is the average intensity over the area of the beam. 29  The SPL and Axial Resolution ▪ The axial resolution refers to the smallest detail that can be resolved in the direction of propagation. The axial resolution 𝑆𝑃𝐿 𝑛𝜆 𝑛𝑣 has a value given by: 𝐴𝑥𝑖𝑎𝑙 𝑅𝑒𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛 = = = where SPL is the Spatial Pulse Length. 2 2 2𝑓 So, the higher the frequency the better the axial resolution (smaller distinguished details) ▪ Typical values of axial resolution are 1.5mm at a frequency of 1MHz, and 0.3mm at 5MHz. ▪ However, attenuation of the ultrasound beam increases at higher frequencies and so there is a trade-off between penetration depth and axial spatial resolution. ▪ Axial resolution is generally around 4 times better than lateral resolution. 30  The SPL and Axial Resolution ▪ The incident and transmitted pulses are blue. ▪ The reflected pulses (from the 2 axially-arranged objects) are green. SPL The 2 objects are not resolved in the image 𝑑 < 𝑆𝑃𝐿/2 Black dots represent 2 axially-arranged objects with separation 𝑑. Not resolved because of overlapping between the 2 pulse reflected from the 2 surfaces. Gray dots is the image of 2 the objects The shorter the separation “𝑑” the greater the overlap and the worse the axial resolution 31  The SPL and Axial Resolution The 2 objects are just The 2 objects are resolved in the image clearly resolved in the image 𝑑 = 𝑆𝑃𝐿/2 𝑑 > 𝑆𝑃𝐿/2 Just resolved when 𝑑 = 𝑆𝑃𝐿/2 because there is no overlapping between the 2 pulse reflected from the 2 surfaces. Resolved when 𝑑 > 𝑆𝑃𝐿/2 because there is no overlapping between the 2 pulse reflected from the 2 surfaces. The greater the separation “𝑑” the less the overlap and the better the axial resolution 32  US Pulse Propagation Inside Tissues The propagation of US inside tissues is non-linear : ▪ As a rule, during +ve part (pressure) of the wave, particle motion is in the direction of propagation and the tissues becomes compressed. This leads to: ✓ an increase in tissue stiffness and hence, ✓ an increase in the speed of sound in that tissue according to the equation. ▪ Whereas during the –ve part of the wave (rarefaction), the wave motion is in the opposite direction so, ✓ the local speed is slightly reduced. This causes the propagation to be nonlinear ▪ The speed at which each part of the wave travels is related to: ✓the properties of the medium (its compressibility) and ✓the local particle velocity, which enhances or reduces the local speed. ▪ At high pressure amplitudes (>1 MPa), non-linear propagation effects become noticeable ▪ The non-linearity results in the generation of high frequency components called harmonics 33  Harmonic Imaging ▪ The generated harmonic components have multiple frequencies of the fundamental frequency 𝑓0 i.e. (2𝑓0 , 3𝑓0 , 4𝑓0 , etc) ✓ The harmonics can be utilized to get an improved image. The use of generated harmonics in imaging: ▪ As you know the higher the frequency the better the axial resolution (but greater attenuation because of scattering). ▪ If we send a fundamental wave (𝑓0 ), and detect the generated second harmonic (2𝑓0 ), the axial resolution is improved as the higher frequency is detected for image formation. ▪ Moreover, non-linear propagation occurs in the highest-amplitude (parts of the transmitted beam near the beam axis), ✓ The effective ultrasound beam (of second harmonic) is narrower than the conventional beam → improved lateral resolution ▪ Weaker parts of the beam (side lobes and edges of the main lobe) produce little harmonic energy. These harmonics are suppressed (attenuated early) in relation to the central part of the beam leads to noise reduction. 34  US Transducers and Beam Forming ▪ Assume a single element transducer, the image is built up line by imaging line on the body surface line as the beam is stepped along the imaging line (on the external body surface) as shown. ▪ So, an array of transducers can be used to make it faster. ▪ There are many shapes of array transducers (such as the linear array and the phased array transducers) and hence: ✓ Special handling via what is called beam forming is required ✓ Different image formation methods (pixel addressing) 35  US Transducers and Beam Forming ▪ The beam-former is the part of the system that determines the shape, size and position of the US beams by controlling electrical signals to and from the transducer array elements: ✓In transmission, it generates the electrical signals that drive each individual transducer element, ✓In reception, it combines the individual echo pulses received by a transducer element into a single echo sequence. ✓The echo sequence produced by the beam-former for each scan line is then amplified and processed in various ways before being used in the formation of the image 36  US Image Formation The type of transducer is considered not only for beam forming but also for image formation. For example, in case of linear array, pixel allocation in the array is simply done column by column (echo sequence from an element will fill a column) For phased array and sector scans, the angle of the element (w.r.t. the column direction) is considered to address the correct location inside the array as shown in this figure. Fig. 4:17. For sector transducers, image pixels on a line making an angle θ with the column direction are filled by this ray. 37  B-mode Instrumentation ▪ B-mode stands for Brightness mode i.e. a 2D image. ▪ The major processing blocks of an ultrasound image is shown: ✓ Transmit beam forming is done via transmit power control by the CPU or electronic circuits. ✓ At receiving stage, the beam is digitized (A/D converter) and then formed by a method set previously according to the chosen transducer and scan parameters. ✓ Formed received beams are then preprocessed by what is called TGC (Time Gain Control or Compensation. ✓ Then, the image is constructed by a scan conversion method, processed and displayed. 38  Ultrasound Imaging system The system can be divided into 3 major parts 1. Analogue front end: ✓ Transmits and receives signal to/from the transducer ✓ ADC (received beam) ✓ Beamforming (can be analog or digital) ✓ Amplification of the received signals TGC (time gain compensation) 2. Digital back end: a computer to ✓ Control the system (beam forming and others control maybe included) and, ✓ Pre-processing, scan conversion , processing and image display. 3. User interface ✓ For parameter setting, more imaging options, measurements and reporting. 39  Amplification of the Received Signal (TGC) ▪ The transmitted ultrasound pulses are attenuated while propagates through tissue. Echoes returning through tissue to the transducer are also attenuated. ✓ Hence, an echo from an interface at a greater depth in tissue is much smaller than that from a similar interface close to the transducer. ▪ In a B-mode image, the aim is to relate the display brightness to the strength of the reflection at each interface regardless of its depth. ▪ Because, the received signal strength will lead to image ambiguity, a method to amplify the echoes should compensate for such different attenuation. ▪ The deeper the surface, the longer the time taken by echo to reach the transducer. ▪ So, amplification should be calculated according to the time (echo from different depth can take) to arrive at the transducer. So, it is called ✓ Time-Gain Compensation (TGC), time gain control or swept gain 40  Time Gain Compensation (a) The gain applied by the TGC amplifier increases with time after transmission to compensate for the greater attenuation of echoes from deeper boundaries. (b) After TGC, echogenecities (𝑣𝑜𝑢𝑡 ) representing interfaces that have same reflectance are equal. (c) After TGC, echogenecities (𝑣𝑜𝑢𝑡 ) representing interfaces that have different reflectance are different. So, TGC logarithmically amplify the signal to compensate for the exponential attenuation with time (depth) to finally get different echogenicities due to differential acoustic impedance (reflectance) 41  Ultrasonography System Keyboard and Control Panel ALARA principle: (As Low As Reasonably-Accepted), TGC: Time Gain Compensation 42  Ultrasound Imaging system Transducers 43  US Signal Display Modes The received signal can be displayed in different modes: ❑ A-mode or Amplitude mode scan: ▪ produces a 1-D plot of the amplitude of the received echo vs. time. ✓ a noninvasive technique for measuring corneal thickness. ✓ Could also be used to measure aortic wall thickness A-Mode scan can be used for Axial ❑ B-mode or Brightness mode scan: Length Biometric. ▪ Echos are detected from surfaces along a line and along the depth of tissue. ▪ Echoes are used to construct a 2-D image. ▪ A pixel location corresponds to relative position within the body cross section. ▪ Brightness of pixels are related to strength of corresponding echoes B-mode scan produces : sectional image 44  M- Mode ▪ The M-mode or motion mode display produces a 2D image (not video) representing the motion of boundaries along a line with time ✓ Time history of single line at the same position over time ✓ Necessary for measurements of cardiac functions M-Mode (right) is the echo of different boundaries along the white line shown at the image (left) for a period of time (x-axis) 45  Doppler Mode The Doppler-mode is based on Doppler shift which states that : ▪ The frequency 𝑓of the echo received from a moving object differs from frequency 𝑓0 of the incident wave such that: 𝑐±𝑣𝑟 𝑓= 𝑓0 where 𝑐 is velocity of wave in the medium, 𝑣𝑟 and 𝑣𝑠 are the velocities of receiver (moving object) 𝑐±𝑣𝑠 and source respectively relative to the medium (+ve if moving away) ▪ Doppler mode measures and visualizes blood flow and speed in vessels: ✓ Continuous wave (CW) Doppler ✓ Pulsed wave (PW) Doppler ✓ Used for blood flow measurement Fig. 7.5 Spectral display: a display of the Doppler frequency shift versus time. The Doppler waveform from the femoral artery is shown. Vertical distance from the baseline corresponds to Doppler shift, while the greyscale indicates the amplitude of the detected ultrasound with that particular frequency. 46  Power Doppler Mode ❑ Power Doppler Mode: color coding based on intensities ▪ Blood flow is indicated by one color (usually orange) ▪ The higher the velocity, the higher the brightness of such color ▪ So, pale color (yellowish) indicates high velocity. Power Doppler mode: paler colors represent higher velocities. 47  Color Flow Mapping Mode ❑ CFM Mode: Color Flow Mapping Mode ▪ Spatial map overlaid on a B-mode gray-scale image that depicts an estimate of blood flow mean velocities. ✓ Direction of flow is encoded in colors (blue: the direction is away from the transducer and red: towards it) ✓ Amplitude of mean velocity is indicated by the brightness, ✓ Turbulence is indicated by a third color (often green). CFM Mode: Color Flow Mapping Mode 48  Other Display Modes Secondary Modes ▪ Duplex ✓ Presentation of two modes simultaneously: usually 2D and pulsed wave Doppler ▪ Triplex ✓ Presentation of three modes simultaneously: usually 2D, color flow, and pulsed Doppler ▪ 3D ✓ Display or Surface/volume rendering used to visualize volume composed of multiple 2D slices. ▪ 4D ✓ A 3D image moving in time 49 Thank you