Information Technology

On this page

1. Introduction
2. Imaging informatics and DICOM standard
3. PACS-RIS systems
4. Teleradiology
5. Artificial intelligence

Introduction

Digital imaging offers many advantages over conventional, film-based imaging. The most compelling is the ability to store, retrieve, distribute and review images at any time and in any location which is appropriately networked.  Digital imaging can produce more consistent images than traditional methods and enables teleradiology, but it also poses increased technical challenges. Medical imaging informatics involves all relevant aspects of imaging, such as image generation, processing, management, transfer, storage, display, privacy and security. It is part of the broader filed referred as medical informatics. 

Medical imaging informatics is a relatively new area of medical physics.  It includes the digital aspects of the medical imaging department, all aspects of digital imaging, network architecture and topology, network infrastructure, integration, connectivity, and security.  The medical physicist must understand digital imaging and communications in medicine (DICOM) and DICOM conformance, Health Level 7, and integrating the healthcare enterprise (IHE). The medical physicist must be part of any institutional decision regarding transmission and display of medical images and must participate in decisions regarding the application of hospital and radiology information systems in the medical imaging department.

Imaging informatics and DICOM standard

Imaging informatics involves the computerized management of imaging data and related information, extending from patient demographic information and referral data to reports, statistics and image diagnosis, including distribution through teleradiology platforms. Digital medical images are typically managed in a dedicated image network and database known as a picture archiving and communications system (PACS), whereas all other patient information is typically recorded in a radiology information system (RIS).

In general, all modern imaging equipment will capture images in a format compatible with the digital imaging and communications in medicine (DICOM) standard published and updated by the Medical Imaging Technology Alliance, a division of the Association of Electrical and Medical Imaging Equipment Manufacturers (NEMA), on whose web site the standard is openly available.

DICOM is a set of standard-based protocols for medical device intercommunication and for the storage and transmission of medical images. The standard preserves fidelity and metadata on each image and examination, and in particular, specific patient identification and the details of image acquisition. A primary purpose of DICOM is the interchange of images and their accompanying information. The standard describes information object definitions (IODs), each of which is specific to a type of image produced by a particular modality, but which shares a common structure.

The DICOM standard can properly preserve cross-sectional imaging data from CT, MRI, positron emission tomography (PET) and single photon emission computed tomography (SPECT). It allows manipulation and reconstruction of the images using various software tools, which can enhance visualization beyond that possible with the originally captured images. DICOM formats retain the full fidelity of the acquired image and also contain metadata that identify the patientand equipment used, describe the examination type and date, retain the order and collation of images into series within an examination, describe the acquisition protocol (e.g. pulse sequence parameters, X ray exposure parameters), describe the image characteristics (e.g. slice thickness, location, acquisition matrix), store the desired image display parameters specific to that examination and store the presentation modes specific to that type of examination.

It is important to note that not all devices support all types of images from different modalities. Accordingly, DICOM defines the combination of a service class and an IOD as a service–object pair (SOP) class. For example, the combination of the storage service class and the CT image IOD is the CT image storage SOP class. The purpose of defining SOP classes is to allow the sender and receiver to negotiate their mutual capabilities when establishing a connection on the network, or what DICOM refers to as an ‘association’. In order to ensure compatibility between IT and image acquisition systems, and display and storage equipment from different vendors, it is important that equipment comply with the appropriate subsets of the DICOM standard and relevant Integrating the Healthcare Enterprise (IHE) protocols. Each manufacturer is required by the standard to document their product’s capabilities in a conformance statement, including tabulation of SOP classes and transfer syntaxes supported.

When stored or transmitted in uncompressed form, digital images occupy an amount of space proportional to the matrix size of the image, i.e., each pixel occupies a fixed number of bytes. However, images typically contain a significant amount of redundant information that can be represented more compactly. Therefore, the DICOM images may be compressed for the purposes of storage and transmission, typically being reduced to half or one third of the original image data size. Some forms of compression allow complete and exact recovery of the original data from the compressed data, and these are referred to as lossless or ‘reversible’ compression schemes. Lossy compression is when some of the original image data are irretrievably discarded. For some referral, consultation and review purposes, images can be transmitted in various lossy compression schemes to optimize the speed of transfer. However, the acceptable levels of compression for different tasks, needs to be defined.  

PACS-RIS systems

A digital image, which may be replicated as many times as necessary, requires a means of distribution, electronic archival and an electronic method of display. Individual devices can be constructed to perform each of these tasks. Grouping such devices together to form a system that manages these tasks produces a picture archiving and communications system (PACS). PACS is therefore responsible for the storage and dynamic access and transfer of clinical images from a variety of imaging modalities.

A PACS may be a small system that addresses the needs of one department, perhaps involving a single modality, with a modest number of workstations for a limited number of specialized users. More commonly, the PACS serves an entire enterprise, such as a hospital. The PACS is responsible not only for acquisition, archival and distribution, but also management of the workflow and integration with other departmental or enterprise-wide information systems such as the radiology information system (RIS), cardiology information system, laboratory information system or hospital information system (HIS). The PACS may have both local users, accessing information from within the organization, and remote users, accessing information from off-site.

The four major attributes of PACS are imaging modalities, secured network for the transmission of patient information, workstations for interpreting and reviewing images and archives for the storage and retrieval of images and reports. Important features of the PACS are the timely and efficient access to images, as well as interpretations and study of patient data.

As noted above, digital medical images are typically managed in PACS, whereas all other patient information is typically recorded in a RIS. A RIS is an information system where patients are registered, examinations are scheduled and radiologists’ reports are recorded, stored and distributed. The RIS can also provide management information and may hold information that is important for revenue generation. In principle, RIS represents a database used by radiology departments with function to store, manage and distribute patient radiological data and images. The main features of the system are patient tracking and scheduling, reporting and image tracking.

PACS and a RIS communicate with each other via Digital Imaging and Communications in Medicine (DICOM).  They need to work seamlessly together. Proper integration of the RIS and PACS can provide productivity improvements in radiology departments resulting in faster study turnaround times for patients and clinicians.

Teleradiology

Digital imaging enables teleradiology. It is the transmission of a complete, full resolution, unaltered images to a hospital distant from where the images were acquired, with a purpose of primary diagnostic interpretation, expert secondary consultation or preoperative surgical planning. Coupled with a picture archiving and communications system (PACS), teleradiology is a powerful tool for both a host radiological facility and for an extended set of clinical users.  Teleradiology enables remote review, consultation and interpretation of medical images. It is a practical and effective method to address the uneven geographical distribution and local shortages of imaging specialists. The teleradiology should be clearly distinguished from the situation in which a small number of low quality images are sent primarily for the purposes of discussion or demonstration.

Teleradiology can be performed locally, between buildings in the same hospital, or in collaboration with other health facilities anywhere in the world. It offers alternatives to traditional image interpretation approaches which require on-site staff capable of radiological interpretation.  Furthermore, teleradiology can improve access to expert medical opinion, either for primary or secondary interpretation, provide access to medical image reporting for underserviced centres, support patient consultations, provide access to image interpretation for remote regions or facilitate reporting in shifts.

Teleradiology adopted modern technology and different use cases. In the presence of a PACS that is remotely accessible, or a centralized archive, teleradiology is no longer distinct from any other form of remote access and the internet-based client.  Although the transfer of images across the world is no longer a demanding task, it is important to note that effective telemedicine solutions require a proper workflow to efficiently handle  large numbers of cases. Furthermore, several important elements must be considered to ensure sustainable and effective teleradiology services, such as image transmission technologies, archive and distribution systems, quality of medical image displays at the site of interpretation, medical registration, credential checking for reporting radiologists and compatibility of information systems. With teleradiology, it is now possible to do interpretation work from home or balance load the interpretation work between different sites.  However, extension of local systems to provide remote access may be limited by network performance issues as well as security issues. Teleradiology equipment, as well as image acquisition equipment, should be DICOM (Digital Imaging and Communication in Medicine) compliant for communication with workstations, telecommunication devices and image storage.

A common PACS/RIS workstation configuration consists of several displays (e.g., one RIS monitor and two reading monitors) as well as workstation software connected to both RIS and PACS.  The reading displays should fulfil the DICOM Greyscale Display Function Standard if used for the primary interpretation of examinations. The role of quality control of monitors, as well as control of ambient light, cannot be overemphasized. A poorly performing monitor, or ambient light levels which are too high, will lead to missed or wrong diagnosis, directly affecting patient care.

There has been growing interest in internet-based applications healthcare, many of them with applications in medical imaging. As a result of increased utilization of wearable devices and smartphones, there are open discussions on their potential role in routine imaging workflow, research, education and patient engagement in radiology.  The ongoing discussions are coupled with technical and ethical challenges, such as assuring the necessary level of image quality and resolution, data security or lack of ethical guidance.

Artificial intelligence

Artificial intelligence (AI) applications are used in medical imaging to facilitate different clinical tasks such as image processing, computer-assisted diagnosis (CAD) for detection of pathologies, image co-registration, patient-specific dosimetry, or prediction of clinical outcomes. Furthermore, AI is used to improve radiology workflow and for optimization studies, e.g., in the form of virtual imaging trials for optimization of design and use of imaging equipment.

AI has tremendous prognostic capacity when combined with radiomics. However, relevant clinical decision-making support is based on multiple reliable imaging biomarkers. To achieve the desired outcome, standardized protocols for image acquisition, feature extraction, and analysis must be considered when utilizing these AI-based tools.

Multidisciplinary collaboration is required in continued development of AI-based medical imaging applications. Targeted areas of quality assurance and safety include standardized image reconstruction protocols, patient-specific optimisation of radiation doses, and validation of AI tools. To ensure quality and safety in medical imaging, the current quality assurance programmes need to be extended to the relevant aspects of AI tools used in clinical practice. The quality assurance should address the performance and safety of AI tools and the involvement of imaging medical physicists is of paramount importance. Medical physicist is a key professional in this process.