8.2 Image Display Fundamentals — Pixels, Matrix & Field of View

Why Image Display Fundamentals Matter

Before an image can be windowed, measured, or reformatted, it must first exist as a displayed digital image built from a defined grid and field of view. ARRT's outline places these building blocks under Image Evaluation and Archiving → A. Image Display: pixel/voxel, matrix, magnification (pan/zoom), and display field of view (DFOV). These four terms sound basic, but they are the vocabulary that spatial-resolution, artifact, and reconstruction questions elsewhere on the exam all assume you already know — and the exam frequently tests the arithmetic relationship between them, not just definitions.

Pixel, Voxel, and Matrix

A pixel ("picture element") is the smallest 2D unit of a displayed digital image; each pixel is assigned a single CT number (in Hounsfield units) that is displayed as one shade of gray. Because CT data is inherently three-dimensional, every pixel you see on the monitor actually represents a voxel ("volume element") — the pixel's surface area multiplied by the reconstructed slice thickness. When the voxel's height (slice thickness) equals its width and depth (pixel dimensions), the voxel is called isotropic; isotropic voxels are what make smooth, non-degraded multiplanar reformats possible (see Section 8.1).

The matrix is the grid of rows and columns of pixels that make up the image — most CT scanners reconstruct a standard 512 × 512 matrix, though many newer scanners now offer higher-resolution 768 × 768 or 1024 × 1024 matrices for applications like temporal bone or extremity imaging where fine spatial detail matters most.

The Pixel-Size Formula

The single most testable relationship in this section is:

Pixel size = Display Field of View (DFOV) ÷ Matrix size

For example, a 512 × 512 matrix reconstructed over a 320 mm DFOV yields a pixel size of 320 ÷ 512 = 0.625 mm. Shrink the DFOV to 160 mm on the same 512 matrix and the pixel size drops to 0.3125 mm — smaller pixels, finer spatial detail. This is why reducing DFOV at reconstruction (not simply zooming the display afterward) genuinely improves spatial resolution: the same matrix now represents a smaller physical area, so each pixel covers less real anatomy.

Scan Field of View vs. Display Field of View

These two "field of view" terms are frequently confused on the exam because they sound similar but govern completely different stages of the imaging chain.

If a technologist selects too small an SFOV before scanning and anatomy falls outside it, that anatomy is permanently lost from the raw data and cannot be recovered by any DFOV adjustment afterward — a rescan is the only fix.

Matrix Size Trade-offs

Choosing a matrix is not simply "bigger is always better." A larger matrix (say, 1024 x 1024 instead of the standard 512 x 512) reconstructed over the same DFOV produces smaller pixels and finer spatial detail, but it also increases reconstruction computation time and the storage size of every image, which matters for PACS capacity and network transfer speed (Section 8.4 covers PACS and archiving in full). Most routine body CT is reconstructed at 512 x 512 because it already balances spatial detail against file size and reconstruction speed; higher matrices are reserved for applications — temporal bone, extremity, or dedicated inner-ear protocols — where the added spatial detail changes the clinical answer.

Magnification: Display Zoom/Pan vs. Reconstruction Zoom

The exam distinguishes two categories of "magnification" that are commonly confused in practice:

1. Post-acquisition zoom/pan (display magnification): simply enlarges pixels that have already been reconstructed. No new information is added, and spatial resolution does not improve — in fact, zooming too far can make the image look blocky or "pixelated" because you are stretching existing data rather than acquiring finer detail.
2. Target (reconstruction) zoom: reducing the DFOV at the time of image reconstruction, recalculating a smaller physical area across the full matrix from the same raw dataset. Because this genuinely shrinks pixel size (per the formula above), it does improve effective spatial resolution — this is why a "targeted" reconstruction of a kidney or a single vertebral body looks sharper than simply zooming into the full-abdomen image on the viewer afterward.

Exam Scenario

A technologist scans the abdomen and pelvis using a 500 mm SFOV to be sure the entire patient width is captured, including a large body habitus patient's lateral soft tissue. After the scan, the radiologist requests a dedicated series focused on the right kidney. The technologist reconstructs a second series from the same raw data using a 200 mm DFOV centered over the kidney. The SFOV itself never changes (the raw data extent was fixed at scan time), but the smaller DFOV, spread across the same 512 matrix, produces smaller pixels and a sharper, more detailed kidney image than simply zooming into the original full-abdomen series on the workstation.