لخّصلي

Chest radiography is the most commonly performed diagnostic X-ray examination. The radiation dose to the patient for this examination is relatively low but because of its frequent use, the contribution to the collective dose is considerable. Consequently, optimization of dose and image quality offers a challenging area of research. In this article studies on dose reduction, different detector technologies, optimization of image acquisition and new technical developments in image acquisition and post-processing will be reviewed.
Studies indicate that dose reduction in PA chest images to at least 50% of commonly applied dose levels does not affect diagnosis in the lung fields; however, dose reduction in the mediastinum, upper abdomen, and retrocardiac areas appears to directly deteriorate diagnosis. In addition to patient dose, the design of the various digital detectors seems to have an effect on image quality. With respect to image acquisition, studies showed that using a lower tube voltage improves the visibility of anatomical structures and lesions in digital chest radiographs but also increases the disturbing appearance of ribs.
New techniques that are currently being evaluated are dual-energy, tomosynthesis, temporal subtraction, and rib suppression. These technologies may improve diagnostic chest X-rays further. They may for example reduce the negative influence of over projection of ribs, referred to as anatomic noise. In chest X-rays, this type of noise may be the dominating factor in the detection of nodules. In conclusion, optimization and new developments will enlarge the value of chest X-rays as a mainstay in the diagnosis of chest diseases.
Chest radiography is the most frequently performed diagnostic X-ray examination; it is of value for solving a wide range of clinical problems. X-ray images of the chest provide important information for deciding upon further steps in the establishment of a diagnosis, treatment, and follow-up procedure. Chest radiography remains the mainstay for diagnosis of many pulmonary diseases, even despite recent developments in cross-sectional imaging of the thorax, particularly computed tomography (CT). The advantages of chest radiography over cross-sectional imaging are lower cost, lower dose, and speed of acquisition and diagnosis.
In a European Directive, the need for optimization of acquisition techniques for X-ray imaging and limitation of the patient dose is established. The effective dose related to a posterior–anterior (PA) radiographic chest image is about 0.02 mSv.
For comparison, this is about 0.5% of a CT scan of the chest. The effective dose related to the lateral chest image is approximately two times higher compared to the dose of a PA projection. Although individual patient dose in chest radiography is relatively low, its contribution to the collective dose is significant due to the frequent use of this examination. In the Netherlands, about a third of all diagnostic X-ray examinations are a chest X-ray.
The associated estimated contribution to the collective dose is about 18%. Similar figures are reported in other Western countries. Chest radiography may be implemented also in screening programs in some countries, this would have a substantial impact on the collective dose from chest X-rays. The frequent use and diagnostic importance of chest X-rays make the optimization of image quality and patient dose an important area of research. Therefore, the relation between these aspects is herewith discussed. Studies that investigated dose reduction and perceived image quality will also be discussed. The digital chest X-ray acquisition technique will be reviewed and recent chest radiography technologies will be assessed, both with respect to diagnostic accuracy and radiation dose to the patient.
Dose in digital chest radiography mainly affects the noise in the images. Noise in radiography can be defined as uncertainty or imprecision of the recording of an image, i.e. unwanted stochastic fluctuations in the image. The most disturbing effect on image quality (and thereby on diagnosis) is that noise can cover or reduce the visibility of certain structures. The loss of visibility is especially significant for low-contrast objects.
An important source of noise in X-ray images is related to the random manner in which the photons are distributed within the image. In a uniform digital image, pixel values (that are associated with the individual detector elements) will vary around their expected value. This type of noise is known as quantum noise. The degree of the fluctuations is related to the exposure level: quantum noise in a detector element is proportional to the square root of the exposure level (noise expressed by the standard deviation of the signal). Here it should be noted that the signal in the detector element is proportional to the number of photons imparting on it. Therefore signal-to-noise ratio (SNR) and thereby image quality will improve with higher exposure levels. For example, improving the SNR by a factor of 2 can only be obtained by increasing the dose by a factor of 4 (assuming that quantum noise is the predominant source of noise). The resulting improvement in image quality will obviously have to be assessed clinically against the increased dose absorbed by the patient. Apart from quantum noise other additional noise sources must be considered in digital radiography, i.e. detector noise (for instance electronic noise) and anatomical noise. Detector noise becomes more significant at low exposure levels whereas for higher exposure levels quantum noise and anatomic noise will dominate in medical radiographs.
Anatomic noise can be referred to as overlaying anatomic features such as ribs, lung vessels, heart, mediastinum, and diaphragm in a chest radiograph. This occurs since chest radiography involves the projection of a three-dimensional structure onto a two-dimensional image. It has been shown that anatomic noise can have an important negative effect on observer performance in detecting abnormalities, especially in chest radiography.
With the former film-screen systems, the range of patient dose in clinical practice was inherently limited by its sensitivity (speed class). Because of the small dynamic range, film-screen radiography images appear underexposed at low doses and overexposed at higher doses. With digital radiography underexposure or overexposure is less likely to occur. This can be explained by the wide dynamic range of these detectors. However, as explained above, the selected dose level used will influence the quantum noise level in the image and thereby the diagnostic potential. Low exposures will still create images with a clear appearance of gross anatomical structures but increased quantum noise will possibly hamper visualization of subtle anatomic and pathologic structures. This aspect of digital radiographic systems forms an extra challenge and opportunity for optimizing patient dose and (perceived) image quality.
An additional aspect of digital radiography is that differences in digital detector technology lead to differences in image quality and dose. In the past decade in most Western European hospitals radiological film-screen (FS) radiography has been replaced by digital radiography systems. Already in the early eighties, computed radiography (CR) with phosphor plate systems was introduced. The radiation level received at each point of X-ray photons reaching the storage phosphor screen is stored in the local electron configuration by elevating the energy level of electrons (excitation) in the phosphor. After exposure, the photon energy stored in the phosphor plate is read out by a laser scanner, and a digital image is obtained.
At first, the quality of these systems was moderate and the dose needed for recording chest images was higher compared to the FS systems. Over the last 20 years, the CR systems improved in both dose requirements in image quality. Recent advances in CR are more efficient collection of light by reading both sides of the screen (dual-sided read CR) which results in an increased signal-to-noise ratio, line scan read out yields improved speed and the use of needle-like phosphor allows for improved X-ray absorption efficiency (a thicker phosphor) without loss of spatial resolution.
In the nineties, so-called direct radiography (DR) digital systems with a flat-panel detector (FPD) became available for chest radiography. In a short period, different digital radiography chest systems were introduced for clinical use. In FPD the conversion of the latent X-ray image into a measurable signal most often occurs in a layer of Cesium Iodide (CsI-FPD), Gadolinium Oxisulphide (GOSFPD) or Selenium (Se-FPD) in combination with a matrix of thin film transistors (TFTs) from which the light (CsI, GOS) or electrical charge (Se), provoked in the layer, is read out and transformed into a digital image. Detectors that use a scintillator that converts X-ray into light, are called indirect conversion systems. The light is actually transformed by photodiodes in the TFT layer into an electrical signal. The detector material in direct conversion systems directly converts X-ray photons into an electrical charge. A selenium layer is used for this purpose. A different technique uses charge-coupled devices (CCD) in combination with a scintillation layer. The chest is much larger than currently available CCD chips. Solutions for protecting the chest on a CCD chip are the use of lenses or tapered optical fibers, at the cost of reduced dose efficiency and degraded image quality. A better solution is the use of the so-called slot-scan technique. This technique uses a linear array of small CCD detectors in combination with a narrow X-ray beam that scans the chest.
Several studies have shown the advantages of digital systems compared to FS, for instance, the improved visibility of the mediastinal areas in the image. One of these studies used an anthropomorphic phantom with simulated lesions to investigate the detection of lesions in the chest for a digital CCD slot-scan system. It was found that with the digital system, the number of lesions observed in the mediastinum was almost twice the number found with the FS system. For the lung lesions, no significant difference was found. The dose levels were roughly comparable between the digital and the FS system (speed class 400).
In another study, the diagnostic performance was compared for eight different digital radiography chest systems. The systems were assessed for the detection of simulated chest disease under clinical conditions. The following systems were regarded: four different flat-panel detector systems, two different charge-coupled device systems, one selenium-coated drum, and one storage phosphor system. Differences in diagnostic performance were found among the eight different digital chest systems in the configurations under which they were routinely applied in clinical practice. Interestingly, differences in detection rates could not be explained by dose. The differences in detector design were given as an explanation here. The DR systems significantly outperformed the single-sided read CR system with respect to image quality whereas the dose levels used with the DR systems were lower. Furthermore, the results suggested that the scanning CCD or slot-scan technique gave the best detection results. Due to the superior anti-scatter properties of the small scanning detector (most scattered photons will go along it), a grid can be omitted, therefore this technique is associated with excellent image quality at relatively low doses.
Several other researchers investigated or compared different digital systems. Better performance at low spatial frequencies for CsI-FPD systems compared to Se-FPD systems is found by studies that use physical parameters to investigate image quality as a function of spatial frequency. This is explained by the high atomic number and high density of CsI resulting in a good capture of the latent X-ray image. On the contrary, at the higher spatial frequencies, the Se-based systems show better performance since these systems show less blurring of the image signal.
For indirect conversion detectors, apart from the high atomic number and high density, the needle-like structure of CsI reduces the spreading of light in the scintillator and allows the use of a thicker layer with higher efficiency compared with unstructured scintillators like GOS and regular CR systems.
An extensive overview of studies that investigated dose requirements and image quality of various digital systems for chest radiography is given in a recent publication. Common systems are given as a function of dose and image quality in clinical practice. The circles represent the uncertainty in the results: the acquisition technique may vary in practice, different research methods are used in literature (each giving inherently limited results), differences found between systems are not necessarily unequivocal, and finally, differences between manufacturers may exist per detector type.
Quantum noise and detector noise hamper the detection of objects with low contrast. Detectability of objects on a uniform background is a function of object contrast and object size (detail). Detectability of objects can be improved by increasing radiation dose. It was demonstrated however that in radiographic chest images for the detection of lung nodules with sizes in the order of 10 mm, the projected anatomy (anatomic noise) is far more disturbing than quantum noise and system noise.
Anatomic noise can hinder detection through two processes: local influence, or “camouflaging,” and global influence, or “confusion”. Local influence is for instance a rib that is projected over a small nodule while global influence describes lesions that fit very well in the global pattern of certain anatomical structures. Confusion is for instance a small nodule that may be interpreted as a tangentially projected vessel. Anatomical noise by the projection of ribs is of particular concern for the detection of lung nodules (i.e., potential malignancies) because the ribs overlay about 75% of the area of the lungs. It was estimated that 50% of nodules measuring 6–10 mm are missed owing to the superimposition of anatomy like ribs, heart, and mediastinal structures. Interestingly, purely on the basis of inherent contrast, a nodule as small as 3 mm in diameter should be visible on chest radiographs when anatomic noise is not taken into account. In general, it appears that many overlooked lung nodules in chest radiographs are visible in retrospect.
A number of methods for investigating the relationship between dose and image quality have been developed. Objective measurements of physical characteristics, such as modulation-transfer function, detective quantum efficiency or contrast-to-noise-ratio, and contrast-detail studies are often used. Alternatively, anthropomorphic phantom studies and clinical studies can be used for subjective (quality rating) and objective (lesion detection) observer performance studies.
The detective Quantum Efficiency (DQE) is a measure of the combined effect of the noise and contrast performance of an imaging system, it is expressed as a function of object detail. Noise can be expressed by the signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), or the noise power spectrum (NPS). An imaging system’s ability to render the contrast of an object as a function of object detail is traditionally expressed as its modulation-transfer function (MTF). The combination of the functions NPS and MTF determines the above-mentioned DQE. These objectives' physical measurements describe the system's technical imaging performance but it is still difficult to translate the outcome to the clinical situation that is far more complex than these measurements can describe.
Contrast-detail studies can be used to determine dose levels that are related to a desired contrast-detail performance or can be used to compare different systems or acquisition techniques. An advantage of contrast-detail studies over objective physical measurements is that contrast-detail performance includes the performance of human observers. A limitation with both contrast detail performance and physical measurements is that the anatomic background is not taken into account.
Anthropomorphic phantoms better approximate the clinical reality with respect to anatomic background. Some studies use these phantoms in combination with artificial lesions that provide a standard of reference in evaluating observer detection performance. Unfortunately, it is impossible to simulate the appearance of the whole spectrum of diseases with artificial lesions.
An alternative is performing clinical dose experiments through computer simulation of the effect of dose reduction on image quality (note that most often it is assumed unethical to perform clinical low-dose studies with real patients). A number of researchers reported on methods for simulating reduced-dose images. Such techniques have been applied in digital radiography and computed tomography.
A well-established method for reduced dose simulation that has been previously described uses DQE and NPS at the original and simulated dose levels to create an image containing filtered noise. The method provides for simulated images containing noise that, in terms of frequency content, agree very well with original images at the same dose levels.
Another study described and validated a pragmatic technique for simulating reduced-dose digital chest images, similar to noise simulation techniques used for CT. This model includes using the raw pixel standard deviation as a measure of noise. Gaussian noise with a certain standard deviation is added to the original image to obtain a simulated reduced dose image with the desired pixel standard deviation for each pixel. After the addition of noise, the raw image was post-processed in the standard way. Nodular lesions were simulated in this study in combination with the low-dose simulation. Then, the detection performance of simulated lesions in simulated and real low-dose images was investigated using this method by performing an observer model study.
A number of studies have been performed to investigate the possible clinical effects of dose reduction in chest radiography using the methods described above.
The purpose of one low-dose simulation study was to investigate whether radiologists can rank the image quality of digital radiographic images that correspond to different dose levels. In this observer study, chest radiographs of 20 patients (both PA and lateral images) with a variety of chest pathologies were used to simulate reduced dose levels corresponding to 50%, 25%, and 12% of the original dose. For each patient, the images were displayed as hard copies on the film-viewing box in random order. Four radiologists ranked the quality of the corresponding images and rated diagnostic quality. The 100% reference dose was not recognized as the best-quality image in nearly half of the cases. Larger dose reductions to 25% and 12% were usually recognized as third and fourth best quality. The preliminary subjective findings suggested therefore that a 50% dose reduction seems feasible in a variety of chest pathologies. However, further dose reduction (more than 50%) clearly reduced the perceived diagnostic quality.
In addition to their subjective study on dose reduction, the same authors also objectively investigated to what extent dose reduction resulted in decreased detection of simulated nodules in the lungs. Raw data from 20 normal clinical digital PA chest images were used in this study. Low-dose images were simulated and nodules were attached digitally. Hard copies were printed representing 100% dose and simulated patient doses of 50, 25, and 12%. These hard-copy images were reviewed by four radiologists. It was found that the decrease in radiation dose from 100% to 50, 25, or 12% had no effect on lesion detection in the lungs. However, the decrease in radiation dose had a prominent effect on lesion detection in the mediastinum. The detection performance of the four radiologists in the mediastinum deteriorated with each dose reduction step.
A different study by other authors evaluated the influence of different doses on the patient and peak kilovoltage settings on diagnostic performance with respect to PA radiographic chest images performed with an FPD. An anthropomorphic chest phantom was used in combination with simulated lesions attached to the phantom. The acquisition technique was varied when making chest radiographs of the phantom. Four radiologists assessed all of the images. For the lung fields only, no significant loss in diagnostic performance could be demonstrated, even after a 50% reduction in radiation dose. This however did not count for the mediastinum where the diagnostic performance deteriorated at this level of dose reduction. This finding is in accordance with the conclusions of the study reported earlier.
Diagnostic X-rays used for radiographs represent a certain energy range: the X-ray spectrum. The contrast in a radiograph is influenced by the X-ray spectrum used. In general radiography, a common X-ray spectrum (of intermediate energies) usually provides the best compromise between sufficient image quality and acceptable patient radiation dose. However, in chest radiography, such an intermediate X-ray spectrum would result in a very pronounced appearance of the ribs in the radiographs. Bone particularly effectively attenuates X-rays at intermediate energies and the resulting bright appearance of the ribs in the radiographs would seriously hinder the evaluation of chest pathology. Historically, to overcome this problem, a compromise had to be made: chest radiography is routinely performed using a relatively high-energy X-ray spectrum. This results in relatively ‘translucent’ represented ribs but also results in less contrast in the soft tissues and particularly soft tissue lesions.
Some studies indicate that CR compared to FS can be used properly in high-energy radiography of the chest providing equal image quality at a comparable effective dose to the patient. Other studies find improved SNRs or visibility scores when using lower tube voltages under constant effective dose levels for CR and DR. Some recent studies however indicate that lower tube voltages in chest radiography have a negative side effect on prominent imaging of the ribs which may disturb the radiologist in his or her detection. Therefore in another study, it is concluded that a compromise has to be made in tube voltage setting. It was demonstrated that an optimum X-ray spectrum for chest radiography with cesium iodide–amorphous silicon flat-panel detectors is obtained with 120 kVp and 0.2 mm of copper filtration. In addition, it was concluded from an anthropomorphic phantom study with human observers that standard PA chest radiographs should be acquired with 120 kVp (also routinely used) in terms of diagnostic performance and effective dose.
The strong current interest in, dual-energy, tomosynthesis, and temporal subtraction has been made possible by the introduction of digital radiography.
An effective method to improve radiologic detection performance seems reduction or elimination of ribs, which has been shown to be the main factor limiting the detection of subtle lung nodules as an early sign of lung cancer. This can be achieved by dual-energy chest radiography.
Dual-energy comprehends weighted subtraction of low and high-energy images and results in images representing bone structures or images representing soft tissue. It was found that dual-energy added to standard PA chest radiography significantly improves the detection of small non-calcified pulmonary nodules and the detection of calcified chest lesions.
However, the technique has some disadvantages as well: (1) the radiation dose is considerably higher compared to standard digital PA images which can be in the order of factor 2–3, (2) Reduced signal-to-noise ratio in result images since the noise of two separated images adds up into the soft tissue or bone image and contrast in the resulting images is reduced, (3) Additional hardware is needed for making images at different energies in a short time interval; e.g. fast switching filters and fast detector read out, and (4) extra images that may increase the workload.
Temporal subtraction comprehends the subtraction of a current image and a prior image of the same patient. Standard PA images are used, so that no extra dose to the patient is needed. The technique facilitates detecting pathologic changes over time and could provide diagnostic advantages. As a result, images arising from nodules hidden by bones or the heart are potentially less likely to be missed. Furthermore, infiltrative shadows should be better distinguished from breast tissue and pectoral muscles in the lung area.
Subtraction of current and prior images is not straightforward because it is not likely that the two chest radiographs will have exactly matching projections. Temporal subtraction uses sophisticated warping algorithms to eliminate the effect of these differences in projection.
The technique is currently only clinically used in Japan. A number of studies showed significant improvements in diagnostic accuracy with respect to nodule detection.
Eliminating ribs has already been shown to be effective with dual-energy suppression radiography despite the disadvantages related to this technique that are hampering implementation in routine clinical practice. An alternative rib suppression method is based on image processing techniques and has potential advantages over dual-energy radiography. Major advantages of such a post-processing approach compared to a dual-energy subtraction technique are: (1) the technique requires no additional radiation dose to patients but uses only chest radiographs acquired with a standard digital radiography system, (2) no specialized equipment for generating dual-energy X-ray exposures is required, and (3) noise levels are not necessarily increased. Digital chest radiography provides a sufficiently large dynamic range and high bit depth to register accurately the entire range of radiation intensities in the latent PA chest image. This implies that the visibility of lesions in the lungs should improve after the subtraction of intensities corresponding to ribs.
One paper describes a framework for rib suppression based on regression and applied on a pixel level. Radiographs with known soft tissue and bone images obtained by dual-energy imaging were used for training the framework. Other authors presented a comparable approach They employed a massive training artificial neural network to create a rib-suppressed image also using dual energy training data. A reduction of the contrast of ribs in chest radiographs is reported in both studies. A drawback of the two methods, likewise the dual energy suppression technique, is that the resulting images suffer from additional noise.
In this technique ribs and lung fields are segmented automatically using image processing techniques. The rib signal is estimated from image and model information and subtracted from the rib (for each rib). By scaling the difference value, the degree of suppression can be varied. Gives a result of the method used and shows the effect of rib suppression.
Digital tomosynthesis is a method of producing coronal cross-section images using a digital detector and a chest X-ray system with a moving X-ray tube. In tomosynthesis, a series of low-dose exposures are made by moving the X-ray tube within a limited angular range. The acquired images are processed into slices that show anatomical structures at different depths and angles. These coronal cross-section images have a relatively high spatial resolution in the image plane, but less resolution in the depth direction. Tomosynthesis can improve the visibility of pulmonary nodules by producing cross-section images without overprotection of ribs and overlying vasculature. The actual effectiveness of tomosynthesis is currently under study in a US national screening trial.
In tomosynthesis, the initial projection images show the anatomy from different orientations due to parallax. They are shifted and added together to render structures in one plane in sharp focus. As a result, objects in other planes are blurred out in the resulting image. Deblurring algorithms are applied to eliminate the blurring of structures that are out-of-plane. A recent paper reports that a typical set of acquisition parameters for chest imaging is 71 projection images, 20◦ total tube movement, and 5 mm spacing of reconstructed sections. The overall radiation exposure to the patient has been reported to be comparable to a former FS lateral chest image. The entire acquisition sequence is completed in 11 s (which is within a single breath hold for most patients).
Digital chest radiography has many advantages over film-screen radiography, including improved diagnostic quality, especially in areas of high attenuation (mediastinum, upper abdomen, retrocardiac) and lower dose to the patient. There are a wide variety of technically different digital systems, flat-panel detectors, and slot-scanning systems related to excellent image quality. The introduction of new techniques (tomosynthesis, dual-energy, temporal subtraction, and rib suppression) enables further improved diagnostic accuracy. These techniques however must be further developed and validated on a larger scale. A drawback may be the extra images that are added to the workload of the radiologist. In conclusion, digital chest radiography is a time-efficient and inexpensive investigation of low doses and good diagnostic quality. Optimizations and innovations may further improve diagnostic accuracy.