Health Informatics: Computer Architectures, Standards, EHRs, and Artificial Intelligence

Comprehensive practice flashcards detailing key terminology across computer and network architectures, healthcare standards, EHR implementations, and medical data science/AI.

Moore's Law

The observation that computer processing speed, capacity, and cost efficiency double approximately every 18 months.

Central Processing Unit (CPU)

The primary component of a computer that executes program instructions and processes data, with speed measured in clock cycles per second (Hertz) and floating-point instructions per second (FLOPS).

Random Access Memory (RAM)

The temporary, working memory of a computer system that holds active programs and data but is lost when power is turned off.

Application Programming Interface (API)

A standardized set of specifications and protocols that enables different software applications to communicate and exchange data across networks.

Database Management System (DBMS)

A specialized software application that houses, organizes, extracts, and manipulates data within a database.

Open Source Software

Software provided with freely accessible, shareable, and modifiable source code, such as Linux, Apache, Mozilla, and OpenOffice.

IPv4

An Internet Protocol addressing standard utilizing a 32-bit address format expressed as four decimal numbers separated by dots (e.g., 66.94.29.13).

IPv6

An Internet Protocol addressing standard utilizing a 128-bit hexadecimal address format developed to replace the depleted address pool of IPv4.

Transmission Control Protocol (TCP)

A core, connection-oriented network protocol that ensures packets reach their destination in sequence, resending dropped or damaged packets.

User Datagram Protocol (UDP)

A lightweight, connectionless transport protocol that does not guarantee delivery or packet order, commonly used for time-sensitive transmissions such as live streaming video.

Network Switch

A networking device that directs incoming data packets specifically to intended recipient devices at full bandwidth, operating like a traffic controller rather than broadcasting to all ports.

Wide Area Network (WAN)

A telecommunications network that spans a large geographical region to interconnect distant computers and local networks.

Personal Area Network (PAN)

A short-range network typically centered around a single person to link devices such as smartphones, computers, and wearables (commonly using Bluetooth).

Virtual Private Network (VPN)

A technology that employs encrypted tunneling protocols over public networks like the Internet to provide secure remote access to a private organizational network.

Standard (ISO Definition)

A documented agreement established by consensus and approved by a recognized body that provides rules, guidelines, or characteristics for activities or their results for repeated and common use.

Interoperability (IEEE Definition)

The ability of a system or product to work with other systems or products without special effort on the part of the user, enabled by common standards.

Semantic Interoperability (Level 4)

The highest level of interoperability, where messages contain standardized, coded data that can be automatically interpreted and processed computationally by receiving systems.

Duplicate Record

A health record integrity error that occurs when more than one distinct medical record is created and maintained for a single patient.

Overlaid Record

A health record integrity error that occurs when data from more than one individual is mistakenly combined into a single patient record.

National Provider Identifier (NPI)

A unique identification number assigned to healthcare providers in the United States, required by CMS to process Medicare and Medicaid claims.

ASC X12N

The standard electronic data interchange (EDI) format mandated under HIPAA administrative simplification rules for healthcare business and claims transactions.

Health Level Seven (HL7) International

A standards development organization dedicated to clinical and administrative data transmission, named after the seventh layer (Application Layer) of the OSI model.

Clinical Document Architecture (CDA)

An HL7 document markup standard that uses XML to define the structure and metadata of clinical documents so they are both human-readable and machine-computable.

Extensible Markup Language (XML)

A standardized data packaging and content language that wraps data in nested, descriptive tags and elements to facilitate cross-platform data exchange.

Fast Healthcare Interoperability Resources (FHIR)

A modern HL7 messaging and interoperability standard that offers richer semantics than HL7 V2 with less complexity than V3, built upon modern web technologies and RESTful APIs.

Digital Imaging and Communications in Medicine (DICOM)

A communications standard developed by ACR and NEMA designed for handling, storing, printing, and transmitting medical imaging information.

Medical Coding

The process of assigning standardized alphanumeric codes to diagnoses, encounter reasons, and healthcare services to facilitate billing, reimbursement, and clinical research.

Superbill

A clinical fee ticket or document listing the most frequently used diagnosis and procedure codes in a specific medical practice, used to initiate claims.

International Classification of Diseases (ICD)

A diagnostic coding and classification standard published by the World Health Organization (WHO) to track mortality and morbidity, clinically modified in the U.S. for reimbursement.

NCPDP SCRIPT

A standard created by the National Council for Prescription Drug Programs that defines the electronic transmission of prescription information between prescribers and pharmacies.

Logical Observation Identifiers Names and Codes (LOINC)

A universal terminology standard developed to identify and code laboratory test results and clinical observations.

Electronic Health Record (EHR)

An electronic record of patient health information conforming to recognized interoperability standards that can be created, managed, and consulted across multiple healthcare organizations.

Clinical Datum

A single clinical observation defined by four fundamental elements: the patient, the attribute, the value of the attribute, and the observation time.

Social Determinants of Health (SDOH)

Environmental, social, economic, and educational conditions—such as housing stability, literacy, and food access—that substantially influence health outcomes and life expectancy.

Single Sign-On (SSO)

An authentication process that permits a healthcare worker to log in once and gain authorized access to multiple connected clinical applications and EHR modules.

Computerized Provider Order Entry (CPOE)

An EHR application module enabling licensed practitioners to electronically submit clinical orders for medications, laboratory tests, consults, and diagnostic imaging.

Clinical Decision Support System (CDSS)

Software embedded within an EHR that assists healthcare providers in decision-making through rule-based medication alerts, reminders, order sets, and clinical calculators.

Electronic Prescribing (eRx)

The direct electronic transmission of a prescription from an EHR system to a retail or mail-order pharmacy, reducing illegibility errors and turnaround times.

Personal Health Record (PHR)

A health record controlled directly by the individual that contains data from multiple providers and patient self-reporting, separate from provider-owned legal health records.

Tethered PHR

A patient portal linked directly to a specific healthcare institution's EHR system, allowing patients to view parts of their medical record and communicate with providers.

Medical Scribe

A trained documentation specialist who accompanies clinicians during encounters to record clinical data directly into the EHR in real time.

Data Science

The interdisciplinary field studying evidence-based principles, algorithms, and processes for extracting nonobvious, actionable patterns and knowledge from large datasets.

Data Analytics

The extensive, systematic application of statistical models, quantitative methods, and explanatory frameworks to data to drive business and clinical decisions.

Machine Learning (ML)

A subfield of computer science (defined by Arthur Samuel in 1959) that provides computers with the ability to optimize performance and learn from data without explicit programming.

Deep Learning

A specialized branch of machine learning utilizing artificial neural networks with multiple stacked layers requiring intensive computational power, especially prominent in medical imaging.

Artificial Intelligence (AI)

The branch of computer science focused on developing systems and algorithms capable of performing complex tasks normally requiring human intelligence.

Big Data 4 Vs

The four defining characteristics of big data: Volume (scale of data), Velocity (generation speed), Variety (diversity of data types), and Variability (data source trust and consistency).

Data Provenance

The documented lineage, origin, ownership trail, and verified trustworthiness of a piece of data across its lifecycle.

Metadata

Structured descriptive information about data that provides context regarding its content, format, origin, or usage.

Data Wrangling

The process of cleaning, transforming, and structuring raw or disparate data into an appropriate format for visualization, analytics, or machine learning.

Supervised Learning

A machine learning paradigm in which an algorithm trains on labeled inputs mapped to known outputs to build predictive models, tested on separate validation data.

Unsupervised Learning

A machine learning approach in which algorithms discover inherent groupings, clusters, or patterns within unlabeled, unstructured data without predefined targets.

Reinforcement Learning

A machine learning method where an algorithm continuously learns optimal behaviors or decisions through trial, error, and feedback within a clinical or operational environment.

Sensitivity

The statistical measure of a model or diagnostic test to correctly identify true positive cases; values drop below $100\%$ in the presence of false negatives.

Specificity

The statistical measure of a model or diagnostic test to correctly identify true negative cases; values drop below $100\%$ in the presence of false positives.

Algorithmic Bias

Systematic and repeatable discrepancies in computer predictions, exemplified when cost-predicting algorithms underallocated healthcare resources to sicker minority patients due to past spending inequalities.

Black Box Medicine

The clinical use of opaque, highly complex artificial intelligence models whose decision-making logic and internal weights cannot be readily explained or interpreted by clinicians.