1/39
Comprehensive vocabulary flashcards covering medical imaging informatics systems (PACS, RIS, DICOM, HL7, IHE), networking architecture, OSI layers, machine learning development, and digital radiographic exposure indices.
Cloud Computing
The delivery of information or computing services via the internet that provides off-site storage of imaging data and facilitates image exchange through web portals or applications.
Machine Learning (ML)
A branch of data science that provides computers with the capability to learn using algorithms based on continuous inputs and iterations without explicit rule programming.
Training Phase (Machine Learning)
The initial machine learning development phase in which the computer is supervised using labeled datasets to learn and accurately classify specific information.
Testing Phase (Machine Learning)
The development phase where the computer is unsupervised and tested to ensure independent performance, requiring a minimum of 85% sensitivity/specificity (90% target for FDA approval).
Validation Phase (Machine Learning)
The machine learning phase where algorithms run completely without human intervention and must achieve an accuracy of 90% or better.
Deep Learning (DL)
An end-to-end computer-driven learning method where human-guided preprocessing and feature extraction are bypassed, utilizing multi-layered artificial neural networks inspired by the human brain.
Overfitting
A machine learning limitation occurring when an algorithm is trained only to recognize very specific features rather than generalizing across broader datasets.
American Board of Imaging Informatics (ABII)
The credentialing body that certifies imaging informatics professionals and administers the Certified Imaging Informatics Professional (CIIP) designation.
Digital Imaging and Communication in Medicine (DICOM)
A universal standard developed in 1983 by the ACR and NEMA that defines file formats and network protocols for exchanging medical images across multi-vendor equipment.
Health Level Seven (HL7)
An ANSI-accredited standard established in 1987 designed to format and exchange text-based medical and administrative information, such as patient demographics, orders, and billing.
Integrating the Healthcare Enterprise (IHE)
An initiative that creates clinical implementation profiles and blueprints specifying how DICOM and HL7 standards must operate together to support specific clinical workflows.
DICOM Modality Worklist (DMWL)
A PACS/RIS integration service that queries the RIS to retrieve scheduled exam lists and automatically populates imaging modality consoles with patient demographic and procedure data.
Service Class User (SCU)
The DICOM role assigned to an imaging device or software application that acts as the client by utilizing or requesting a service (such as an acquisition modality pushing an image).
Service Class Provider (SCP)
The DICOM role assigned to a network device or archive that acts as a server by providing or fulfilling a requested service (such as a PACS archive storing incoming studies).
Service Object Pair (SOP) Class
The basic functional unit of DICOM, formed by pairing a DICOM Message Service Element (DIMSE) command (such as store or print) with an information object definition.
Pixel Pitch
The physical distance between the center of one pixel and the center of the adjacent pixel in a digital detector matrix, where smaller pitch produces higher spatial detail.

Computer-Aided Detection (CADe)
An algorithmic imaging tool designed to automatically analyze medical images to detect and localize suspicious findings or lesions.
Computer-Aided Diagnosis (CADx)
An advanced informatics tool that evaluates user-specified regions of interest within medical images and classifies abnormalities as benign or malignant.
Lossless Compression
Reversible image data compression (e.g., DICOM JPEG, JPEG 2000) that reduces file size while retaining exact original pixel values upon decompression.
Lossy Compression
Irreversible image compression that achieves high compression ratios by discarding image data, which can introduce artifacts and permanently compromise image fidelity.
Open Systems Interconnection (OSI) Model
A 7-layer architectural framework that standardizes telecommunication and computer network protocols from physical signal transmission up to end-user software interaction.
Bus Network Topology
A linear local area network layout where all nodes connect along a single shared transmission backbone ended by terminators, allowing only one node to transmit at a time.

Ring Network Topology
A closed-loop network configuration where each node connects to adjacent nodes, transmitting data sequentially in circular paths to provide resistance to cable breaks.

Star Network Topology
The most prevalent LAN design, where all peripheral devices connect individually to a central hub or switch, isolating local line failures from the rest of the network.

Mesh Network Topology
A complex network architecture where nodes possess multiple redundant point-to-point connections to each other, integrating characteristics of bus, star, and ring configurations.

Category 5 Unshielded Twisted Pair (Cat 5 UTP)
The minimum standard twisted-pair copper network cable required to support PACS image transmission, utilizing RJ-45 termination with a maximum run distance of 100m.
Fiber Optic Cable
A transmission medium that conveys data as light pulses through pure glass filaments via total internal reflection, providing complete immunity to electromagnetic interference.
Switch (Networking)
An intelligent networking device serving as the core of modern star topologies that establishes dedicated, simultaneous point-to-point data connections between individual nodes.
Router
A layer-3 network device that examines destination addresses within data packets to direct traffic efficiently across differing network protocols and subnets.
Storage Area Network (SAN)
A high-speed, dedicated sub-network of shared storage devices operating primarily on the Fibre Channel protocol to archive enterprise radiological imaging data.
Low Earth Orbit (LEO) Satellite
A satellite positioned less than 1,500km above Earth that completes an orbit every 100minutes, enabling high-speed transmission with minimal latency (20–40ms).
Latent Image
The invisible image recorded on a digital image receptor following exposure before any computer processing or rescaling algorithms are applied.
Manifest Image
The visible image displayed on a digital monitor following computer preprocessing, automated histogram analysis, and look-up table conversions.
Direct Digital Radiographic Receptor
An image receptor system that directly captures X-ray exposure across a detector matrix, converting analog radiation signals directly to digital values via an analog-to-digital converter.

Sensitivity Number (S#)
An exposure index utilized by Konica (mean) and Fuji (median) that is inversely proportional to exposure at the receptor, standardized so that S=200 corresponds to 1mR.
Carestream Exposure Index (EI)
A directly proportional logarithmic indicator of the mean air kerma reaching the detector, calculated using the formula EI=(1000×log10(mR))+2000.
Target Exposure Index (EI_T)
The manufacturer- or department-specified optimal exposure index calibrated for a specific anatomical projection to produce diagnostic image quality at minimal radiation dose.
Deviation Index (DI)
A standardized exposure indicator quantifying variance from the target exposure index, defined mathematically as DI=10×log10(EITEI).
Deviation Index Value of -3.0
A metric indicating a 50% underexposure to the digital image receptor, requiring the radiographer to double the mAs (+100%) for technical correction.
Deviation Index Value of +3.0
A metric indicating a 100% overexposure to the digital image receptor, requiring the radiographer to halve the mAs (−50%) to correct technical factors.