Fundamentals of Medical Imaging Informatics

Information Science and Healthcare Information Technology

  • Information Science:

    • Defined by Techtarget as the utilization of computers, storage, networking, physical devices, infrastructure, and operational processes to create, process, store, secure, and exchange all forms of electronic data.
  • Healthcare Information Technology (Health IT):

    • The domain of information technology centered on the design, development, creation, implementation, and maintenance of information systems customized for the healthcare industry.
  • Informatics:

    • Refers to the foundational process of transforming raw data into meaningful, actionable information.
    • Computers and communication technologies serve as indispensable, central elements within this transformation process.
  • Imaging Informatics:

    • Defined by the Society for Imaging Informatics in Medicine (SIIM) as the comprehensive study and application of information and communications technology processes dedicated to the acquisition, manipulation, analysis, and distribution of medical image data.

Digital Image Characteristics

  • Matrix:

    • A digital image is structurally composed of a two-dimensional array of numbers, designated as a matrix.
    • The matrix is arranged in columns (MM) and rows (NN) that delineate small square regional divisions termed picture elements or pixels.
    • Matrix Size and Image Quality: Increasing the number of pixels within the image matrix enhances image quality and improves the visual perception of structural detail (Luiten, A.L., 1995, Medicamundi, 40, 95-100).
  • Field of View (FOV):

    • Represents the physical dimension or total anatomical area of the patient included within the image.
    • A larger FOV encompasses a broader anatomical region.
  • Pixel (Picture Element):

    • The foundational, smallest individual element comprising a digital image matrix.
    • Pixel Size: The dimensions of a pixel directly determine the spatial resolution and level of detail attainable in the digital image.
  • Pixel Bit Depth:

    • Refers to the specific number of bits contained within each individual pixel.
    • Standard medical digital imaging systems feature bit depths ranging between 1010 and 1616 bits.
    • Higher bit depth enables the imaging system to display a broader spectrum of gray shades.

Medical Image Quality Parameters

  • Brightness:

    • Pertains to the physical visual presentation of the image on a display monitor.
    • Dependent on the monitor's physical capacity to emit light through its display surface.
  • Contrast Resolution:

    • The inherent ability of a digital system to distinguish and display subtle variations in shades of gray.
  • Spatial Resolution:

    • The capacity of an imaging system to accurately demonstrate minute physical details of an object.
  • Quantum Noise:

    • Represents unwanted image noise that provides zero diagnostic utility while actively degrading overall image quality.
    • Visually presents as fluctuations in brightness or optical density across the displayed image.
    • Governed by photon arrival: a lower quantity of x-ray photons reaching the image receptor (IR) results in greater visible quantum noise.
  • Dynamic Range:

    • The ability of an imaging system to respond accurately across varying levels of radiation exposure.
    • Enables a wider array of tissue densities to be visualized on the digital image, imparting the appearance of enhanced detail.
  • Exposure Latitude:

    • The functional range of diagnostic exposure values that an image detector can record.
    • Directly tied to detector design: detectors possessing a higher dynamic range capture a broader spectrum of exposure values.

Digital Image Processing Categories

  • Image Enhancement:

    • Processing operations executed to generate an image that is visually optimized and more pleasing to the observer.
  • Image Restoration:

    • Operations aimed at improving image quality by correcting inherent distortions, degradations, or systemic artifacts.
  • Image Analysis:

    • Computational operations involved in extracting quantitative metrics, geometric parameters, or analytical measurements directly from image data.
  • Image Compression:

    • Techniques designed to decrease digital file size, thereby expediting transmission times across networks and minimizing storage repository demands.
  • Image Synthesis:

    • Advanced processing operations that generate new synthetic images from secondary image sets or non-image data sources.

Spatial Frequency Filtering and Pre-Processing

  • Edge Enhancement (High-Pass Filtering):

    • Executed when specific spatial frequencies corresponding to regions of interest are predefined.
    • Amplifies target high frequencies while suppressing non-target frequencies.
    • Directly increases image contrast and enhances anatomical border sharpness.
  • Smoothing (Low-Pass Filtering):

    • Conducted by mathematically averaging each pixel's frequency with the values of adjacent surrounding pixels.
    • Suppresses high-frequency quantum noise at the expense of reducing overall contrast and detail sharpness.

Post-Processing Techniques

  • Annotation:

    • The administrative or clinical addition of text labels directly onto an image.
    • Common structural labels and orientation flags include: LL, RR, APAP, PAPA, LATERALLATERAL, SUPINESUPINE, PRONEPRONE, PORTABLEPORTABLE, CONTROLCONTROL, CROSS−TABLECROSS-TABLE, ASSIGN PATIENTASSIGN\,PATIENT, Unassigned ImageUnassigned\,Image, APPLY AND ACCEPTAPPLY\,AND\,ACCEPT, and EUXCEUXC.
  • Window Width and Window Level:

    • Window Level (WLWL): Controls overall image brightness; raising or lowering WLWL adjusts screen darkness.
    • Window Width (WWWW): Regulates display contrast by defining the proportional range between black and white values.
    • Operational parameter examples include: WW=603WW = 603, WW=249WW = 249, WW=499WW = 499, WW=95WW = 95, WW=399WW = 399 matched with WL=248WL = 248, WL=399WL = 399, WL=−10WL = -10, and WL=−1WL = -1
  • Background Removal or Shuttering:

    • Darkens unexposed white collimation margins bordering the anatomical region of interest.
    • Effectively eliminates monitor veil glare to prevent visual fatigue.
  • Image Orientation and Magnification:

    • Software tools allowing re-orientation (rotation/flipping) and spatial magnification of selected anatomical regions.
  • Image Inversion:

    • Reverses original grayscale display values, rendering bone as black and soft tissue as white.
  • Digital Subtraction:

    • A processing technique that subtracts superimposed anatomical structures to highlight specific features such as contrast-filled vasculature.
  • Image Stitching:

    • A specialized software method that connects multiple sequential images into a unified panoramic display when target anatomy exceeds single-cassette dimensions.

Computed Tomography (CT) Post-Processing Methods

  • Multiplanar Reconstruction (MPR):

    • A prevalent three-dimensional (3D3D) technique where volumetric data acquired as thin axial slices is reconstructed into orthogonal or oblique planes.
  • Maximum Intensity Projection (MIP) and Minimum Intensity Projection (MinIP):

    • MIP: Evaluates highest pixel densities along ray paths to depict vascular networks.
    • MinIP: Evaluates lowest pixel densities to depict air-filled structures.
  • Volume Rendering Technique (VRT):

    • Evaluates entire data volumes and assigns specific color palettes based on tissue density intensity.
    • Utilizes histogram graphs to differentiate bones, contrast media, and soft tissue organs using distinct color profiles.
  • Shaded Surface Display (SSD):

    • Utilizes a defined threshold value of pixel intensity.
    • Eliminates all data below the designated threshold and renders structures above it as solid 3D3D objects with colored surface properties.

Radiology Workflow and Information Systems

  • Workflow:

    • Refers to the step-by-step progression required to complete a given operational process.
    • In medical imaging, workflow encompasses the sequence from initial exam order entry to final report transcription.
    • Defined by Siegel and Reiner as the overall movement of patients, image data, and clinical information across imaging departments and healthcare enterprises.
  • Hospital Information System (HIS):

    • An enterprise-wide information system built to support clinical operations and administrative patient management.
    • Encompasses two foundational operational arms:
    • Clinical Information Systems (CISs)
    • Administrative Information Systems (AISs)
  • Radiology Information System (RIS):

    • A dedicated software network optimized for managing radiological image data and operational workflows.
    • Key capabilities include:
    • Order entry
    • Exam scheduling
    • Report generation
    • Billing preparation
    • Operational management functions: quality assurance (QA), inventory monitoring, and statistical analysis.