Author: Anja Henner, PhD, Principal Lecturer (emerita)
Today’s young radiographers and radiologists are only familiar with image data in electronic form and the processing, transferring, and viewing of this data. Images are exactly where they are needed, they can be viewed in many places at the same time, and they can be processed. The darkroom, the film processor, the daylight film processor and the light box are historical relics. However, older people who are still working – and have worked – in that world, they remember. Has there been a quantum leap into direct digital medical imaging or is it a marathon?
Background
Scintigraphy or gamma scans were introduced for nuclear medicine using gamma cameras in the 1970s, and in the early days the size of the image matrix was 64 x 64. Computed tomography (CT) came to Finland in the 1970s to 1980s and required its own computer room to processing the image. The device produced only one two-dimensional image per rotation. In the late 1980s, the slip ring technique allowed for a spiral motion while the patient slid through the gantry. The imaging time was considerably shortened, which made it possible to get an image of an area of the body. I remember how in 1999 at the ECR (European Congress of Radiology) all CT equipment suppliers had a 4-slice device, and some had a transparent gantry, i.e., the movement of the X-ray tube was visible during imaging.
In angiography, film subtractions were used, especially in cerebral vascular imaging, to fade out bone structures and improve the visibility of small blood vessels, but at the end of the millennium, DSA (digital subtraction angiography) evolved and transformed angiography. Image processing and the correction of motion artefacts, for instance, provided new possibilities, at first clumsily and slowly.
There had already been big changes in general x-ray imaging before the so-called daylight film processors were introduced. X-ray film was originally developed by holding it in processing tanks of developer and fixer and rinsing with water in between. This was followed by a film processor device, into which the film was placed from a cassette opened in the dark. There were films of different sizes and qualities in the film bin: slow, gradual, steep, single and double emulsion. Similarly, there were also different kinds of intensifying screens, the properties of which were described by colour sensitivity (blue or green sensitive) and speed class (50, 100, 200, 400, 800). The X-ray film and the intensifying screen were selected according to the desired image quality, taking into account the indication for the x-ray and the required radiation dose. At the paediatric radiology department, I remember having at best 16 different combinations for patients of different ages and for different indications for examinations. The correct imaging values also had to be known for each combination: after the exposure, the image could no longer be processed. Overexposure or underexposure was immediately apparent and could not be corrected afterwards.
From film intensifying screen technology to digital, also in plain imaging
The first imaging plates (CRs) were introduced in general x-ray imaging in the 1990s. Imaging plate technology was a two-step process: X-ray radiation causes electrons to be trapped in a plate containing barium fluorohalides. The energy states are released when a laser light scans the image plate on the imaging plate dot by dot and line by line, producing fluorescent light. When digitised, they form the values of the elements of the image matrix. This process is an indirect acquisition of an image: the imaging plate is not digital. The advantage is that the exposure latitude of an imaging plate is great (overexposure / underexposure does not occur as easily), image data can be processed and archived. Initially, for several years, the image was printed on laser film, which required device compatibility. The DICOM standard enabled the compatibility of hardware from different equipment manufacturers and surely facilitated the work of both users and equipment suppliers.
Fuji had the first patent for an imaging plate, but other equipment vendors quickly launched their own imaging plates. In the early 2000s, there were several imaging plate systems operating under the same principles, but the imaging processing software and the names and parameters describing imaging plate “properties” were totally different. In principle, it was easy to start using imaging plates: the cassette sizes were mostly the same sizes as the ones used in film imaging, and the film processor was replaced by an image reader. But adjusting the X-ray generator’s automatic exposure control and optimising the image quality have probably been a nightmare for quite a few people. When imaging on film, the relative velocity indicated the radiation dose required: if the relative velocity doubled, the amount of radiation required was halved. Imaging plates came with dose indicators: S and L value, LgM value, EI and REX value. Each one had its own logic and they were very different and behaved very differently in terms of their numerical values. The IEC standard (IEC 62494-1) was developed in 2008 to harmonise these indicators, but its implementation has taken a long time. It should be noted that the exposure index (EI) indicates how well the expected dose of the image receptor has been achieved, but not the dose received by the patient. The DIMOND III project (2004) produced recommendations on image quality and speed class of the image receptor, but their implementation was cumbersome and slow due, at least in part, to the confusing jumble of dose indicators.
As the use of imaging plates became more widespread, a new device for capturing images appeared on the market, but a proper name could not seem to be found for it. The Radiation and Nuclear Safety Authority in Finland introduced the term flat panel detector, but the name has not become established with professionals. Therefore, it is more commonly referred to as a direct digital image receptor – or currently detector (DR). Here, too, different techniques are used: in some cases fluorescence occurs, which is converted into an electronic signal (indirect), in some cases fluorescence conversion is not required (direct), instead radiation causes an electrical signal directly in a semiconductor. What is essential in both techniques is that they use a detector on which everything can be imaged. There is also a portable detector for imaging in a ward, for instance, and a smaller detector for imaging children or limbs.
Mammography was the last to go digital
Imaging of the breast with a mammography device began in the 1970s, when a balloon placed in a cone was used for compression. Accurate resolution is required of a mammography image in order to detect small calcifications and new or nascent changes sufficiently early. The start of mammography screening in 1987 posed its own challenges. Thus, an attempt was made to improve the image quality with a slow film reinforcement plate combination. The cassette had a reinforcement plate on one side only and a single emulsion film was placed against it. The development process was optimised by slowing the throughput time of the film from the normal 90 seconds to three minutes and raising the temperature of the developer solution. With the advent of imaging plate technology, two-sided imaging plate scanning and in-house image processing programs for mammography images were developed. Efforts were also made to reduce the pixel size. For a long time, the Radiation and Nuclear Safety Authority assessed whether the quality of an imaging plate is sufficient for screening. The fading of image data and the cleanliness of imaging plates were also deemed problematic. Thus, in mammography, the transition to digital imaging occurred later. Today, most mammography equipment in Finland uses detector technology. According to the statistics of the Radiation and Nuclear Safety Authority, a total of 159 mammography devices have been registered in Finland, of which 98 are direct digital devices. In relative terms, a lot of imaging is done with imaging plate CR technology in Finland compared to other Nordic countries and Europe. In Sweden, for example, mammography is performed with fully direct digital equipment. The farther east and south you go in Europe, the more imaging plate CR technology is used.
Image monitors and information systems
The requirements for image monitors and software have changed dramatically in the past quarter of a century. Cathode Ray Tube (CRT) monitors are a thing of the past and Liquid Crystal Displays (LCDs) are bigger and more accurate, evolving all the time. The conditions for viewing the image have also changed: radiologists used to have to look at the images on the image board when examining the image, today the radiologist sits in a dark room somewhere – he or she could be at home or skiing. In the mid-1990s, there was in-service training for radiologists, where they envisioned how in the future a radiologist could communicate his or her findings of x-rays while relaxing at home – now 25 years later, that vision is part of clinical practice. It has been made possible by high-speed telecommunications connections that could not have been dreamed of at the end of the last millennium. It could take half an hour to transfer a digitally scanned lung image from general x-ray device to the PACS in other location. Of course, the processing of image data and patient data requires a good and secure connection. At the same time, it speeds up patient care regardless of location – hopefully.
A few memories along the way
During the era of film intensifying screens, aluminium wedges were often used to even out the exposure. In digital imaging, these are no longer needed because the reading of an X-ray can be optimized and the image can be processed. There is no overexposure or underexposure in digital imaging due to its wide dynamic range. This results in the same amount of radiation being used in all patients, regardless of patient size and the indication for the X-ray. Also, changing the imaging values and the effect of the imaging values overall on radiation dose and image quality are also less clear. Of course, with the use of flat panel detector technology in particular, patient doses have been reduced considerably. The Radiation and Nuclear Safety Authority has also taken this into account when setting reference doses. An achievable level has been set for flat panel detector for a few x-ray examinations.
Many things have changed over the years. Because X-ray film was expensive, it had to be conserved, therefore multiple projections were shot on the same film. In digital imaging, there is no need to worry about this. Another interesting change is the cropping of an image. There were numerous different sizes of film cassettes and X-ray films, and the cassette was selected according to the size of the subject. There were fewer sizes in the imaging plates and only one or two sizes in the detectors. Therefore, the cropping of the image may be too extensive, or the image may be cropped on the monitor after exposure. However, the original area exposed to radiation should always be seen in the image; this can be very significant information, for example, after scanning a pregnant person. Especially in the early days when imaging plates were introduced, the loosening of the collimation was noticeable, because the equipment supplier’s instruction was: direct radiation must be applied to the imaging plate.
There has also been a change in the staff of radiology departments: the number of physicists has increased considerably and the profession of developers, for example, has disappeared. A new position is person in charge of PACS, who makes corrections if, for example, images have been archived under the wrong patient’s name or the image data contains incorrect information. New methods have been introduced for quality assurance and dose monitoring has already been partially automated so that dose data and image quality can be monitored separately for each x-ray examination. Maybe in the near future you will even get a “dose passport” or at least some kind of dosage information that is specific for each patient.
Looking back, just over three decades of digital imaging feels like a short time. Looking ahead – no one can predict. Has there been a quantum leap – at least a digital jump. In health care overall, progress has been much slower than in business or banking services, for instance – you could say the speed has been at snail’s pace. It took a couple of decades to get an electronic prescription, and the My Kanta heath data archive was made and developed in the same manner. Imaging has traditionally been considered to be of a high technical standard and a pioneer. In this regard, one could say that perhaps a small leap has been taken or only a half marathon has been completed.
Anja Henner 1.11.2021