What Is Medical Imaging Informatics Infrastructure (MIII) in PACS Systems?

The  medical imaging informatics infrastructure (MIII) is an organized method for selecting tools and synthesizing data from HIS, RIS, PACS, and ePR systems with the goal of developing an image-based diagnostic or treatment system. All of this is part of the discipline known as medical informatics.

The medical imaging informatics infrastructure (MIII) is a server designed to leverage existing PACS resources, images, and related data.

MIII has applications in education, research, and large-scale horizontal and longitudinal clinical services. 

MIII is the ideal vehicle for facilitating this use of PACS, in addition to its daily clinical service. Below, we will learn more about its functionalities.

Architecture and Components of an MIII System

An MIII system comprises the following components: 

1. Medical images and associated data (including the PACS database).
2. Tools for image processing. 
3. Display, graphical user interface.
4. Communication networks.
5. Database management and application-oriented software. 

PACS and Related Data

The data source in the MIII consists of the PACS database and other databases related to the health information system that contain: 

• Patient demographic data. 
• Case histories.
• Medical images.
• The corresponding diagnostic reports. 
• Laboratory test results. 

All this data is organized and archived according to standardized protocols and data formats, such as DICOM for images and HL7 for text. 

In addition, a controlled health vocabulary can be used as a standard for identifiers, codes, and medical messages proposed by the American Medical Informatics Association.

Image Processing

The image processing software allows configuring the image content indexing and retrieval mechanism. Its functions include segmentation, region of interest determination, texture analysis, content analysis, morphological operations, image registration, and image matching. 

These can be performed automatically or interactively via an input gateway or a PACS system server. The result of image processing can be a new image or a description of certain features. 

The data extracted by the image processing functions can be added to the image data archive of the MIII system.

Database and Knowledge Management

The software of the database and knowledge base management component has several functions:

1. Integrates and organizes PACS images along with all related data, extracts image features and keywords from image processing, and derives medical heuristic rules and guidelines into a coherent multimedia data model.
   
2. Supports online database management, content-based indexing and retrieval, formatting, and distribution for display and manipulation.
   
3. Data extraction tools. With the knowledge base as a guide, these can be used as a method for extracting relevant information from the databases.

Display and Graphical User Interface

The interface display and the graphical user interface are output components. Both components are related to workstation design.

Display includes 3D rendering, image data fusion, and visualization of static and dynamic images. It will also use data extracted from image processing (that is, segmentation, enhancement, and shading) for output rendering.

Display can be performed on a standard workstation (WS) or with high-performance graphics engines. For a low-performance workstation, the final display can be precomputed and packaged onto the WS; whereas with high-performance graphics engines, rendering can be done in real time on the WS.

Communication Networks

Communication networks include network hardware and communication protocols necessary to connect the MIII components to one another, as well as to the PACS system.

The MIII communication network can have two architectures: 

• Have its own dedicated network with a connection to the PACS networks. This must be transparent to users and provide the high-speed performance needed for MIII to request PACS images and related data, distributing the results to users' workstations.
  
• It can share communication networks with the PACS. Here, the MIII must have a logical segment isolated from the PACS networks so that it does not interfere with the PACS's daily clinical functions. 

System Security and Integration

Once images have been processed in the PACS system, certain validation mechanisms are needed to ensure their integrity and accuracy. When these components are implemented in the MIII, application-oriented software can be designed and developed to integrate the components needed for a specific clinical, research, or educational application.

System integration includes the system interface, shared data, and workspace software. 

The system's interface software uses existing protocols and communication networks to connect all the infrastructure components into an integrated imaging information system. 

The workspace and shared data software allocates and distributes resources, including data, storage space, and workstations, to online users.

PACS-Based Medical Imaging Information

Originally, PACS systems arose as an image management system to improve the efficiency of radiological practice. This has evolved so much that it has become a hospital-wide integrated system that handles information media in many forms, including: voice, text, medical records, waveform images, and video recordings. 

To integrate these various types of information, multimedia technology is required, which includes: hardware platforms, information and database systems, communication protocols, display technology, and system interconnection and integration. 

As a PACS grows in size and functionality, so does the content of its database. The wealth of information within the PACS offers the opportunity for an entirely new approach to medical research and practice through the discipline of medical informatics.

Below, we present some MIII resources as an illustration of its connectivity and use. 

Some of the resources used by the PACS-based medical imaging informatics infrastructure (MIII) are:

Indexed Image-Based Content

PACS systems use an enhanced MIII database that contains the  keywords of diagnostic reports, patient history, and image sequences, as well as certain image features for content-based indexing of underlying modality images in the PACS database.

The use of artificial indexing such as patient name, ID, age group, disease category, etc., through one-dimensional keyword searches, is a fairly simple procedure. 

On the other hand, indexing through image content is complicated, because the query first has to understand the content of the image, which can include abstract terms (for example, regions of interest), derived quantitative data (for example, area and volume of the object of interest), and texture information (for example, interstitial disease).

Three-Dimensional Rendering and Display

The 3D PACS rendering is designed as a data management system. It lacks the computational capacity for image content analysis at the imaging workstation or the PACS controller. 

For this reason, it is necessary to allocate 3D computing and rendering resources in the MIII infrastructure for high-performance computations in case MIII workstations request such a task. 

Once a given 3D computation is complete, the results, along with display capability, can be distributed to other imaging workstations over high-speed networks. 

Distributed Computing

The basic idea behind distributed computing is that if several computers are networked together, the workload can be divided into smaller parts so that each computer works faster and the total time decreases. 

Request a Demo

References

• PACS and imaging informatics. Basic Principles And Applications. H. K. Huang, D.Sc., FRCR (Hon.) 2004.
• Medical Imaging Informatics. Radiology Key.