Aplikasi Pada Komputer Radiology

Summary

The video explains how computer hardware and software support digital image acquisition, processing, display, and analysis in radiology.

1. Radiology Computer Applications and Equipment

Specialized radiology software controls imaging systems and helps process or reconstruct images. It connects the computer’s CPU with the additional devices used by each imaging modality.

Equipment varies by modality. Examples include:

• Computed radiography (CR): An image reader and imaging plate
• Digital radiography (DR): A detector plate
• Computed tomography (CT): Integrated detectors
• Magnetic resonance imaging (MRI): Radiofrequency (RF) components, including transmitter and coil systems

The software and hardware are designed to handle the data and reconstruction needs of each imaging system.

2. Analog-to-Digital Conversion and Digitization

Imaging devices produce signals that must be converted from analog form into digital data. This is done by an analog-to-digital converter (ADC).

The lecture describes analog signals from image receptors and sensors, including phosphor materials, CT detectors or ionization chambers, scintillation crystals in nuclear medicine, and piezoelectric crystals in ultrasound. The ADC converts these varying signals into binary numbers; this conversion is called digitization.

In radiography, differences in signal intensity or attenuation are represented as different digital values, which contribute to the gray levels in the resulting image. A comparator, identified as one component of an ADC, compares incoming signals or voltages and indicates which is larger.

The lecture identifies three important ADC parameters:

1. Bit depth (digitization depth): Determines how many gray levels can be represented. For example, 10-bit data can represent (2^{10} = 1{,}024) levels.
2. Dynamic range: The range of input signals that can be represented. Together with bit depth, it affects image contrast and latitude.
3. Digitization rate: How quickly analog data can be converted into digital data.

3. Pixels, Voxels, and Image Matrices

A pixel (“picture element”) represents a location in a two-dimensional image. A voxel (“volume element”) represents a three-dimensional volume and has length, width, and height.

Each pixel or voxel is assigned a digital value corresponding to signal intensity or attenuation. Images are arranged in matrices. Increasing the matrix size generally increases spatial resolution, but also increases the amount of data.

• A (256 \times 256) matrix contains 65,536 pixels.
• A (512 \times 512) matrix contains 262,144 pixels.

The image matrix size is distinct from the physical size of the object or field of view being imaged.

4. Specialized Image Processors

The lecture describes a processor array, or dedicated image processor, that handles large amounts of imaging data and repetitive calculations more quickly than a general-purpose CPU.

A graphics card is offered as an analogy: it is designed to handle graphics tasks, while radiology processors are designed for medical-image processing. These processors may use mathematical operations such as Fourier transforms and matrix calculations.

They are particularly useful in data-intensive systems such as CT and MRI, where faster processing can shorten image-reconstruction tasks.

5. Image Display and Histograms

Radiology images are generally displayed in grayscale. The video notes monitor refresh rates of about 30–60 Hz to reduce flicker.

A frame buffer temporarily stores image data and supports display adjustments, such as windowing, without changing the image data held in memory.

A histogram represents the distribution of gray levels in a digital image and helps describe its brightness or exposure distribution.

6. Digital Contrast and Brightness Controls

In digital radiography, window width controls the range of gray values displayed and therefore affects contrast:

• A narrower window produces higher contrast, but can hide values outside the selected range and increase the apparent prominence of noise.
• A wider window displays a broader range of gray values and latitude, generally with lower contrast.

Window level shifts the displayed range and is related to image brightness or density.

In CT, pixel values are commonly expressed using Hounsfield units (HU). The lecture refers to the different values associated with materials such as water, bone, and tissue. The number of gray levels available depends on the image’s bit depth.

7. Digital Image-Processing Methods

The video presents several ways to manipulate digital images mathematically while retaining the original image data:

• Contrast enhancement: Adjusts window width and level to emphasize the desired range of image values.
• Filtering and image enhancement:
  • Convolution applies a mathematical operation across neighboring pixels.
  • Low-pass filtering (smoothing) reduces high-frequency noise, often by averaging nearby pixel values. It can make an image appear smoother.
  • Band-pass filtering keeps a selected range of image frequencies or values while suppressing others. The lecture relates this to selecting particular CT display windows, such as bone or lung windows.
  • High-pass filtering (edge enhancement) emphasizes edges and can make boundaries, such as lesion margins, more visible.
  • Threshold-based edge filtering can increase contrast between regions and highlight small structures, though the appearance depends on how the threshold is set.
• Magnification (zooming): Enlarges a selected region by copying and enlarging its pixels. Excessive enlargement can make the matrix or individual pixels visible; smoothing may be used to soften the result.
• Volume rendering: Uses three-dimensional image data, such as CT data, to create a rendered view of structures including bones and blood vessels.

8. Digital Subtraction Imaging

Digital subtraction can emphasize contrast-filled vessels by subtracting one image from another. The lecture describes three approaches:

1. Masking

   • Acquire an image before contrast administration (the mask) and another after contrast administration.
   • Subtract the mask from the post-contrast image to suppress background anatomy and emphasize the vessels.
   • Patient movement between images can cause subtraction artifacts. Serial acquisition and carefully selected time intervals can help manage this problem.

2. Dual-energy technique

   • Acquire images using two different X-ray energy settings (high and low kVp).
   • Use the differences between the images to separate or suppress selected materials.
   • The lecture notes that this approach is less affected by movement between separately acquired mask and post-contrast images.

3. Hybrid technique

   • Combines masking and dual-energy methods to produce a more detailed depiction of contrast-filled vessels.

The lecture notes that digital subtraction can reveal small vessels and support medical assessment or treatment planning.

Speakers and Sources Featured

One presenter/narrator delivers the lecture. No other speakers or interviewees are featured in the subtitles. The video is from the Lana Prasetya channel.