🖥️

Ultrasound Image Processing Overview

Sep 23, 2025,

Overview

This lecture covers essential image processing and contemporary features in ultrasound, focusing on their functions, types, impact on image quality, and whether they are pre- or post-processing. The unit is divided into two parts: 15a discusses image processing and contemporary features, while 15b will address how these features affect spatial, contrast, and temporal resolution.

Image Processor & Scan Converter

  • The image processor receives scanline data from the signal processor (also called the receiver). At this stage, the data is a long string of digital information representing individual scan lines, not yet an image.
  • The scan converter digitizes and stores this data, converting it from a vertical format into a horizontal format suitable for display.
  • The scan converter fills a pixel matrix with data, creating one complete frame. This process occurs at least 30 times per second, producing real-time imaging by continuously refreshing the display.
  • Pre-processing occurs during live imaging (write mode). Adjustments like gain, compression, and TGCs are made at this stage, affecting how the data is written and stored.
  • Post-processing occurs after the image is frozen (read mode). At this point, only display changes can be made; the original data cannot be altered.
  • Analogy: Pre-processing is like editing a book before it is published; post-processing is like reading the published book—you can only change how you interpret or view it, not the original content.
  • Once the image is frozen, any changes made are reversible, as the original data is saved in memory. However, changes made during pre-processing overwrite the original data and cannot be undone.

Magnification (Zoom)

  • Magnification, or zoom, allows the sonographer to enlarge a region of interest (ROI) for closer examination. There are two types:
    • Write Magnification (Write Zoom):
      • A pre-processing function performed during live scanning.
      • Steps: Acquire a live image → select the zoom button → adjust the ROI box → activate zoom again → the machine rescans only the ROI using all scan lines and pixels.
      • This method dedicates the same number of scan lines and pixels to a smaller area, significantly improving spatial (detail) resolution. If the ROI is shallower than the original image depth, temporal resolution can also improve, as fewer pulses are needed.
      • Example: Zooming in on a moving object, like a puppy’s nose, provides high detail and clarity, showing fine textures and features.
      • Write magnification is preferred whenever possible because it maintains image detail and can improve both spatial and temporal resolution.
    • Read Magnification (Read Zoom):
      • A post-processing function applied after the image is frozen.
      • Steps: Freeze the image → turn the zoom knob → the pixels in the ROI are simply enlarged, with no rescanning.
      • This method does not add new information; it just makes existing pixels larger, resulting in a more pixelated and blurry image with loss of detail.
      • Example: Zooming in on a frozen image leads to visible pixelation and less clarity, making it harder to distinguish fine details.

Fill-In Interpolation

  • Fill-in interpolation (also called pixel interpolation) is a pre-processing function.
  • Used in sector scans, where scan lines start from a common point and diverge, creating gaps in the far field.
  • The machine automatically estimates and fills in pixel values for these gaps by averaging the values of surrounding pixels, resulting in a smoother and more continuous image.
  • This process is not user-controlled and is designed to be seamless, so the sonographer typically does not notice when it occurs.

B Color (Brightness Color)

  • B color changes the grayscale map to a color map, allowing the sonographer to select color schemes that enhance tissue differentiation and visibility.
  • This is a post-processing function, applied to stored digital data after acquisition.
  • B color improves contrast resolution, as the human eye is better at distinguishing color differences than grayscale differences.
  • The sonographer can choose from various color maps to suit personal preference or to highlight subtle anatomical differences and borders, making it easier to detect abnormalities.

Panoramic Imaging

  • Panoramic imaging expands the field of view by allowing the sonographer to slide the transducer along the body, creating a long, continuous image that covers a larger area than the transducer’s footprint.
  • The machine matches overlapping scan lines and stitches frames side by side, effectively “gluing” multiple frames together to form a wide image.
  • This technique is especially helpful for imaging large structures, vessels, or areas that exceed the width of a single frame, such as muscles or long vessels.
  • Panoramic imaging provides a comprehensive view that would not be possible with a single frame, making it easier to assess the full extent of anatomical structures.

Compounding Techniques

  • Compounding techniques combine multiple images to improve image quality by reducing noise and artifacts and enhancing resolution. There are three main types:
    • Spatial Compounding:
      • A pre-processing function.
      • Averages images acquired from different angles by steering the beam in various directions. Each frame is created with scan lines directed at a different angle, and the resulting frames are averaged together.
      • This process reduces artifacts (such as shadows), smooths the image, and improves spatial resolution. However, it decreases temporal resolution because multiple frames are needed to create one compounded image.
      • Example: With spatial compounding on, images appear smoother and have fewer artifacts, but some diagnostic artifacts (like shadowing from stones or tumors) may be less visible.
      • Spatial compounding is often enabled by default but can be turned off if needed to better visualize certain artifacts.
    • Temporal Compounding (Persistence):
      • Also a pre-processing function.
      • Superimposes frames taken over time from the same angle, layering them to reduce noise and increase the signal-to-noise ratio.
      • Commonly used in color Doppler imaging to improve color fill within vessels or heart chambers.
      • Improves spatial and contrast resolution but decreases temporal resolution, as more frames are required for averaging.
      • Example: High persistence settings result in smoother, more filled color Doppler images, while low persistence may show more noise and less color fill.
    • Frequency Compounding:
      • Processes echoes from multiple frequencies simultaneously, rather than sequentially.
      • Improves spatial and contrast resolution by averaging images created from different frequencies, but does not degrade temporal resolution since all frequencies are processed at once.
      • Unlike spatial and temporal compounding, frequency compounding does not require the acquisition of multiple sequential frames; instead, it uses the transducer’s bandwidth to process multiple frequencies in parallel.

Frequency Tuning & Coded Excitation

  • Frequency Tuning:
    • Uses the transducer’s bandwidth to assign high frequencies to the near field, mid frequencies to the mid field, and low frequencies to the far field within a single image.
    • This technique improves lateral and axial resolution, especially in the near and mid fields, by using higher frequencies where possible.
    • Frequency tuning is a feature of the beam former, not the image processor, and is not an averaging or compounding technique.
  • Coded Excitation:
    • Involves sending a long, coded pulse made up of short segments, which are decoded upon return to the transducer.
    • This technique improves axial resolution, signal-to-noise ratio, spatial and contrast resolution, and penetration depth.
    • Coded excitation allows for better image quality without increasing the number of pulses sent, as the machine can extract more information from each coded pulse.
    • This is typically a built-in feature of modern ultrasound machines and is not user-controlled.

Edge Enhancement

  • Edge enhancement is a pre-processing function that sharpens image borders and makes anatomical boundaries more distinct.
  • The machine detects interfaces between different tissues and adds subtle highlights and shadows on either side of these boundaries, creating the appearance of sharper edges.
  • This enhancement improves contrast resolution and helps the sonographer better visualize anatomical borders and interfaces.

Contemporary Features: Elastography & Cardiac Strain Imaging

  • Elastography:
    • Assesses tissue stiffness, providing valuable information for evaluating masses or fibrosis, such as in the liver, breast, or thyroid.
    • Two main types:
      • Strain Elastography: The sonographer applies manual pressure with the transducer, and the machine measures how the tissue deforms. The result is a qualitative map showing soft (e.g., blue) versus hard (e.g., red) areas.
      • Shear Wave Elastography: The machine sends a strong pulse, generating shear waves that move sideways through the tissue. The speed of these waves is measured, providing a quantitative value of tissue stiffness (in kilopascals). Faster shear waves indicate stiffer tissue.
    • Elastography is especially useful for assessing liver fibrosis and distinguishing between benign and malignant lesions.
  • Cardiac Strain Imaging:
    • Measures myocardial deformation to assess heart muscle function and detect areas of dysfunction.
    • Two methods:
      • Speckle Tracking: The machine tracks specific areas (speckles) in the myocardium to measure movement and deformation.
      • Tissue Doppler: Assigns colors to tissue to visualize motion and deformation, allowing assessment of how different segments of the heart contract and relax.
    • Cardiac strain imaging evaluates the effectiveness of myocardial contraction and relaxation, helping to identify areas of infarction or abnormal function.

3D Rendering

  • 3D rendering compiles ultrasound data from three planes to create a three-dimensional image, providing a more realistic and detailed view of anatomy.
  • Data can be acquired by manually sweeping the transducer or using specialized transducers that capture all three planes simultaneously.
  • 3D rendering is a post-processing function, as manipulation and visualization occur after the data is acquired and stored.
  • This technique enhances anatomical visualization and is especially useful for complex structures or when a comprehensive view is needed.

Key Terms & Definitions

  • Pre-processing: Adjustments made during live scanning (write mode) that affect how data is acquired and stored in memory.
  • Post-processing: Changes made to frozen, stored images (read mode) that affect only the display, not the original data.
  • Spatial Resolution: The ability to distinguish small details and structures in the image.
  • Temporal Resolution: The ability to accurately display moving structures in real time, depending on frame rate.
  • Contrast Resolution: The ability to differentiate between varying echo intensities, enhancing the visibility of subtle differences in tissue.
  • Region of Interest (ROI): The specific area selected for detailed imaging or analysis, often used in magnification or elastography.
  • Signal-to-Noise Ratio (SNR): The proportion of useful signal compared to background noise, directly affecting image clarity and quality.

Action Items / Next Steps

  • Complete the activity in the workbook to reinforce understanding of these concepts.
  • Answer the review questions in the nerd check section to test your knowledge.
  • Prepare for upcoming units on dynamic range, harmonics, and contrast imaging, which will be covered in more detail in future lessons, as these are also important image processing techniques.

Full transcript