Our glossary of scientific abbreviations on the Chimaera website

In medical imaging and the development of AI-based applications, we encounter many abbreviations and terms.
This glossary is intended to serve as a brief guide through this jungle of abbreviations, summarizing key terms in an understandable way to classify complex content and facilitate professional exchange.

ACP

Advance Care Planning (ACP) is a structured counseling approach in healthcare that helps people reflect on and clearly define their preferences for future medical, nursing, psychosocial, and spiritual care. Specially trained professional counselors guide the process, which is based on the patient's individual needs and values.
ACP is considered an ongoing, dynamic form of living will that strengthens autonomy and helps avoid mistreatment.

AI-B2

AI-B2 stands for "artificial intelligence for business-to-business". It describes the targeted application of AI technologies to improve collaboration and business relationships between companies. Examples include using machine learning to analyze supply chains, using AI-powered chatbots for corporate customer service, and automating sales processes. The goal is to increase efficiency, productivity, and data-driven decision-making in the B2B environment.

AI-B2 at Chimaera – Learn more

CAD

CAD is the abbreviation for computer-aided design. It is a technology that uses computer programs and software to support the creation, modification, analysis, and optimization of two- or three-dimensional designs. In medical technology, CAD is often used to design prostheses, implants, or other medical devices. The seamless integration of image data from medical imaging into CAD models makes it possible to develop precise, customized solutions for patients.

CBCT

CBCT (cone beam computed tomography) is a specialized imaging technique that produces detailed, three-dimensional images of hard tissues, such as bones and teeth.
Unlike a conventional CT scanner, which uses a thin, fan-shaped X-ray beam, CBCT uses a cone-shaped beam. This wide beam can cover the entire target area in a single rotation of the device. This results in lower radiation doses for patients and faster scans. CBCT is mainly used in dentistry, oral surgery, and ear, nose, and throat (ENT) medicine.

Data Quality Management and Automated Processing of CBCT Data – Learn more

CE

CE stands for "Conformité Européenne" (European Conformity) and is a legally required marking on products confirming compliance with all applicable EU directives and regulations on safety, health, and environmental protection. 
In the field of medical devices, the CE marking indicates that the product meets the EU's basic legal requirements and can be sold throughout the European Economic Area. It is not a quality feature, but rather a minimum standard and legal proof of conformity. Manufacturers affix the mark themselves, usually after successfully completing a conformity assessment procedure.

CNN

CNN (convolutional neural network) is a type of deep learning model particularly suited for processing image data. 
A CNN is modeled on the brain's visual perception process. It uses special filters, known as "convolution layers," to automatically recognize and extract important features in images, such as edges, textures, and shapes. This enables CNNs to perform tasks such as image classification, object recognition, and image segmentation with precision. Therefore, CNNs are one of the most important technologies in medical image analysis.

Automated Data Entry Validation with LLMs or CNNs – Learn more

CSV

Computerized Systems Validation (CSV) is a documented, structured process that ensures a computer system reliably performs its intended functions in a regulated environment, such as the medical device industry, pharmaceuticals, or biotechnology. 
CSV includes not only the final verification of the completed system but covers the entire system lifecycle – from requirements definition through development and commissioning to operation and decommissioning. As a result, successful CSV ensures, for example, that laboratory information systems or production software operate safely, reliably capture relevant data, and do not pose unnecessary risks to patients or users. 
Performing CSV is a regulatory requirement to demonstrate product and patient safety in critical areas, such as medical devices or pharmaceutical manufacturing.

CT

CT stands for computed tomography. It is an imaging technique that uses X-rays to produce detailed, cross-sectional images of the body. 
Unlike conventional X-rays, which provide a two-dimensional view, CT scans allow for three-dimensional viewing of bones, organs, and tissues. CT scans are often used in emergency medicine to diagnose injuries and diseases and plan surgeries and radiation therapies. It is also referred to as a PET-CT when combined with PET.

Hip and femur segmentation from CT – Learn more

DICOM

DICOM (Digital Imaging and Communications in Medicine) is an open standard for storing and exchanging medical image data. This information can include digital images as well as additional data, such as segmentations, surface definitions, and image registrations. The DICOM standardizes the format for storing data as well as the communication protocol for exchanging it.
Communication according to this protocol takes place between DICOM nodes, which are computers or devices capable of sending, receiving, or storing data in accordance with the DICOM standard. A DICOM node is a computer or device that is capable of sending, receiving, or storing data in accordance with the DICOM standard. Examples of DICOM nodes include imaging devices (e.g., CT scanners), archiving systems (PACS), and image viewers.

Your customized web solution – Learn more

DNN

A deep neural network (DNN) is a type of artificial neural network that uses multiple hidden layers between the input and output layers. This deep structure allows DNNs to identify intricate patterns and relationships in data, rendering them useful for various applications, including image recognition, speech processing, and recommendation systems.

DNN Model Compression – Learn more

DVT

DVT stands for digital volume tomography. It is used synonymously with cone beam computed tomography (CBCT).

DWG

DWG stands for "Drawing" and refers to a proprietary file format created by Autodesk. 
It is used as the native format for technical drawings and 2D and 3D models in computer-aided design (CAD) programs, particularly AutoCAD. This binary format precisely and efficiently stores geometries, layer information, text, blocks, and other CAD-specific data, enabling the creation, editing, and exchange of complex design data in architecture, engineering, and medical technology. Despite its proprietary nature, DWG files can be opened and edited in many common CAD applications, though compatibility may vary between program versions.

FDA

The U.S. Food and Drug Administration (FDA) is the United States' food and drug regulatory agency and is part of the U.S. Department of Health and Human Services. Its primary responsibility is to protect public health in the United States. The FDA monitors the safety and efficacy of human and veterinary drugs, biological products, medical devices, food, radiation-emitting devices, tobacco products, and cosmetics, including those manufactured and imported to the United States. The FDA also oversees the manufacturing, licensing, labeling, and distribution of these products in the U.S. market.

fMRI

fMRI stands for functional magnetic resonance imaging. It is an imaging technique that measures brain activity.
By detecting changes in blood flow and oxygen levels, fMRI reveals which areas of the brain are active during specific tasks, such as thinking, speaking, or perceiving. It is primarily used in neuroscientific research and for planning brain surgeries.

Automated organ segmentation in MRI – Learn more

ERP

The acronym ERP is an abbreviation for the term “Enterprise Resource Planning”. 
It refers to the business task of efficiently planning, controlling, and managing available resources, such as personnel, capital, operating resources, and materials. ERP systems support this process by digitally mapping and connecting all business processes via a central database, enabling cross-departmental, streamlined operations. 
Thus, ERP systems increase a company's productivity by automating processes and providing a uniform database that allows for transparent control of operations.

GenAI

In artificial intelligence (AI), generative refers to AI systems that use generative models and stored learning data to produce new results, such as texts, speech, sound recordings, images, or videos, based on input data. Generative AI (GenAI)models learn patterns and structures from training data and use them to generate new data with similar characteristics.

Generative AI – Learn more

 

GDPR

The General Data Protection Regulation (GDPR) is a European Union regulation that standardizes the processing of personal data by private and public entities across the EU. It is intended to protect personal data within the EU. 
For companies like Chimaera GmbH that work in healthcare and handle medical image data, complying with the GDPR is of the utmost importance because it ensures sensitive data is handled securely and legally.

Girder

Girder is commercial automation software in the System & Utilities category that enables centralized control and automation of a wide variety of devices and applications. The software offers a graphical user interface that can be used to create both simple and complex automation routines and control individual processes in the home, office, or industrial environment—for example, for lighting, security systems, audio/video devices, and PC functions. Girder supports numerous hardware standards and software protocols, enables drag-and-drop script creation, can be expanded with plug-ins, and offers remote access via web servers and mobile apps.

GPU

A GPU (Graphics Processing Unit) is a type of computer processor. Though originally developed for displaying graphics in video games, GPUs are now essential hardware components in the fields of machine learning and medical image processing. Unlike a CPU, which processes complex tasks sequentially, a GPU can perform thousands of calculations simultaneously. This makes it ideal for training AI models and accelerating the processing of large image datasets.

AI building block: GPU server – Learn more

LLM

LLM stands for Large Language Model. It is a sophisticated AI model that has been trained to understand and generate human language. 
LLMs learn to recognise grammar, syntax, and semantic relationships by processing huge amounts of text data. They can compose texts, answer questions, create translations and generate code based on instructions (or 'prompts'). LLMs form the technological basis of many GenAI applications, such as chatbots and text generators.

Tailor-made LLMs for healthcare – Learn more

MDR

MDR stands for Medical Device Regulation. It is a European regulation that governs the approval and monitoring of medical devices within the European Union. 
The MDR replaces the previous Medical Devices Directive and aims to increase patient safety by imposing stricter requirements on the manufacture and distribution of medical devices, including software as a medical device (SaMD). The MDR is of crucial importance for any company that places medical devices on the EU market.

MDR development and approval – Learn more

MONAI

Medical Open Network for AI (MONAI) is a framework for deep learning (DL) in medical imaging that is both open-source and community-supported. 
MONAI offers a suite of domain-specific implementations of various deep learning (DL) algorithms and utilities, meticulously engineered for medical imaging tasks. MONAI is utilized in both research and industry settings, facilitating the development of a variety of medical imaging applications, including image segmentation, image classification, image registration, and image generation.

AI Model Zoo – Learn more

MRI

MRI (MR) stands for magnetic resonance imaging (also known as nuclear magnetic resonance imaging). It is an imaging technique used in medical diagnostics that produces detailed cross-sectional images of the body.
Unlike computed tomography (CT), MRI does not use X-rays, but rather strong magnetic fields and radio waves to create images of soft tissues, organs, and structures such as joints, muscles, the brain, or the spinal cord. Since MRI does not use ionizing radiation, it is often used when radiation exposure should be avoided.

Compensation of Cardiac Motion in 3D MR Time Series – Learn more

NDT

Non-destructive testing (NDT) is a method of examining materials, components, or systems for defects, faults, or damage without damaging the test object.
Unlike destructive testing methods, which test an object until it fails, NDT uses techniques such as X-rays or ultrasound to analyze an object's internal structure. NDT plays an important role in quality control, particularly in industry and medical technology, as it ensures products meet required standards without rendering them unusable.

Non-destructive quality testing – Learn more

NIfTI

NIfTI stands for the Neuroimaging Informatics Technology Initiative. It is a widely used file format for storing and exchanging medical image data, particularly in neuroimaging.
The format was developed to improve interoperability by providing uniform, standardized data structures for images from procedures such as magnetic resonance imaging (MRI) and functional magnetic resonance imaging (fMRI). This allows scientists and software to more easily share and process data, greatly simplifying the analysis and exchange of research results.

Our Chimaera technology stack – Learn more

NLP

Natural language processing (NLP) is a subfield of artificial intelligence that deals with the automated analysis, processing, and interpretation of human language in text or speech form by computers. 
The goal of NLP is to extract structured information from unstructured linguistic data to make human language usable for applications such as text analysis, speech recognition, machine translation, and chatbots. 
In medicine, NLP is used to automatically evaluate doctors' notes, extract relevant information from patient records, and improve medical research data management.

Data Extraction and Validation – Learn more

PACS

PACS (Picture Archiving and Communication System) is a network of computers and servers that has been specially developed for storing, managing, distributing, and displaying medical image data. A PACS system replaces traditional imaging media such as X-ray films and enables easy access to digital images (e.g., from CT, MRI, or PET) from different departments or locations. It is a central component in modern radiology and a hub for the exchange of DICOM data.

PET

PET (positron emission tomography) is a nuclear medicine imaging technique that visualizes metabolic processes and functions in the body. 
Unlike imaging techniques such as MRIs or CT scans, which primarily show anatomical structures, PET scans show biological activity in tissue. To accomplish this, a small amount of a radioactive substance, or tracer, is administered to the patient. This substance accumulates in certain cells or organs. PET scans measure the radiation emitted by this substance to create detailed images of metabolic activity. This method is primarily used in oncology, neurology, and cardiology for diagnosis. PET is often combined with a CT scan, so it is also referred to as a PET-CT scan.

Registration of melanoma PET and CT scans – Learn more

ROI

The term "region of interest" (ROI) refers to an area of interest in an image. Typically, the area of interest is an organ being examined or a representative subcutaneous region.
 A comparison of an ROI with a measuring probe in the image area can be performed in a variety of ways with respect to size and shape. Count rates recorded over time, spectrally, or along a path are generally used as measurement curves.
In the field of computer-aided image processing and imaging techniques, it is employed as a two-dimensional or three-dimensional area.

SaMD

SaMD stands for "software as a medical device." It describes software that performs medical functions independently of hardware medical devices. Examples of SaMD include apps that analyze medical images for diagnosis and software that interprets patient data to support treatment decisions. The development of SaMD is subject to strict regulatory requirements (e.g., the EU's MDR) to ensure patient safety and effectiveness.

AI/ML software as a medical device – Learn more

SDK

SDK (software development kit) is a comprehensive collection of tools, libraries, documentation, and code examples that help software developers create applications. SDKs form the basis for developing software for specific platforms, operating systems, or technologies by providing the necessary compilers, utilities, and interfaces.
The Chimaera SDK is a toolbox that simplifies software development for medical image processing. It offers ready-made components for tasks such as image processing, rendering, and integrating AI algorithms. Using an SDK can significantly reduce development time from prototypes to finished medical devices. However, purchasing an SDK may be subject to certain rules, such as a non-disclosure agreement.

Our Chimaera SDK – Learn more

STL

STL stands for "stereolithography" and refers to an additive manufacturing process, or 3D printing, as well as the associated file format for 3D models.
In the STL process, a liquid photopolymer is cured layer by layer with light to create highly precise and detailed models, prototypes, or components. The STL format describes the surface of a digital 3D object using a collection of triangular surfaces. It is widely used in medical technology for customized implants, surgical planning, and prototype production, for example.

VOS

VOS stands for "video object segmentation." A subfield of image processing and computer vision, VOS aims to automatically separate a specific object in a video from the background and track it over time.
In medicine, VOS is used to segment and track moving structures in the human body in real time. This is particularly useful in echocardiography for analyzing heart wall motion and in laparoscopy for tracking surgical instruments. Unlike simple image segmentation, which analyzes each individual image (frame), VOS considers the movement and temporal continuity of the object. This enables more precise and consistent analysis.
Other areas of application for video object segmentation include surveillance, autonomous vehicles, and film editing.

VOS from key frames – Learn more

XNAT

XNAT is an open-source platform designed for managing, archiving, and processing medical image data, especially for research and multicenter studies. 
XNAT enables the upload, pseudonymization, and structured management of image data, such as MRI and DICOM files. It offers role-based user management and provides extensible workflows and APIs for automated image and metadata processing. 
XNAT supports the integration of external analysis tools and has plugins for viewers and annotations. It is supported by an active community and extensive documentation.