Finding MRI features to obviate the need of repeat spinal biopsies in clinically suspected persistent or recurrent spinal osteomyelitis.

Purpose: The aim of this study was to determine magnetic resonance imaging (MRI) features that could help differen-tiate the bone destruction due to persistent/recurrent spine infection from worsening bone destruction due to mechanical factors, which could help obviate the need for repeat spine biopsy. Material and methods: A retrospective study was performed on selected subjects who were more than 18 years of age, were diagnosed with infectious spondylodiscitis, underwent at least 2 spinal interventions for the diagnosis at the same level, and had MRI prior to each image-guided intervention. Both MRI studies were analysed for vertebral body changes, paravertebral collections, epidural thickening and collections, bone marrow signal changes, loss of vertebral body height, abnormal signal in intervertebral disc, and loss of disc height. Results: We observed that worsening of changes in paravertebral and epidural soft tissue were statistically more significant predictors of recurrent/persistent spine infection (p< 0.05). However, worsening destruction of vertebral body and intervertebral disc, abnormal vertebral marrow signal changes, and abnormal signal in intervertebral disc did not necessarily indicate worsening infection or recurrence. Conclusions: In patients of infectious spondylitis with suspected recurrence, the most common and pronounced MRI findings of worsening osseous changes can be deceiving and can result in negative repeat spinal biopsy. Changes in paraspinal and epidural soft tissues are more helpful in identifying the cause of worsening bone destruction. Correlation with clinical examination, inflammatory markers, and observing soft tissue changes on follow-up MRI is a more reliable way to identify patients who may benefit from repeat spine biopsy.

3D printed testing aids for radiographic quality control.

Quality control testing of radiographic and fluoroscopic imaging systems requires positioning of test objects in the x-ray beam in a precise and repeatable fashion. In this work we present several three-dimensional (3D) printed testing aids that improve efficiency, accuracy, and repeatability of quality control testing. We also present a new device for determining the location of the perpendicular ray in radiographic systems. These devices were designed in an open source software program (OpenScad, http://www.openscad.org) and 3D models were saved in .stl format for printing. The models were printed on either a MakerBot Replicator 2 or Replicator Z18 printer (MakerBot Industries, LLC, Brooklyn, NY). The testing aids were printed using polylactic acid (PLA) filament. To investigate the radiographic characteristics of the PLA used, test articles were printed and used to measure the half-value layer (HVL) thicknesses in mm of PLA and half-value densities (HVD) in g/cm2 of PLA for two different colors and over a wide range of radiographic beam qualities, using a portable fluoroscopic c-arm system. HVL thicknesses of clear PLA ranged from approximately 20 mm at 50 kV nominal tube voltage to 27 mm at 120 kV nominal tube voltage. For green PLA, the HVL thickness was 19 mm at 50 kV tube voltage and 25.7 mm at 120 kV tube voltage. The HVD of clear PLA ranged from 2.37 g/cm2 at 50 kV nominal tube voltage to 3.19 g/cm2 at 120 kV nominal tube voltage. For green PLA, the HVD was 2.35 g/cm2 at 50 kV tube voltage and 3.17 g/cm2 at 120 kV tube voltage. The cost of the devices range from under $2 to approximately $20 in materials. The files used to create the models are freely available at https://github.com/Upstate3DLab/3D-Printed-Radiographic-Test-Tools.

Evaluation of commercial extension plates for the ACR CT accreditation phantom.

The American College of Radiology (ACR) Computed Tomography (CT) Accreditation Program requires submission of phantom scans acquired with the ACR accreditation phantom. There is a known issue with some wide-beam scanners in which the Hounsfield unit (HU) value of water may be correct when using the scanner manufacturer's phantom, but will be out of range in some scan modes when scanning the accreditation phantom. The phantom manufacturer has developed a product known as Extension Plates to eliminate the water HU value issue. The purpose of this technical note is to evaluate the effectiveness of the Extension Plates in alleviating the water HU issue. The ACR phantom was scanned on nine different CT scanners representing four CT manufacturers at eight different facilities. Scanner models included 16- and 64-channel geometries from each manufacturer. All scanners passed routine daily water HU testing per the manufacturer's instructions. The accreditation phantom was scanned in helical and axial modes both with and without the Extension Plates present. Regions of interest were placed on the linearity test objects as well as the water HU test object in Module 1 of the phantom. Mean values were recorded and compared with the acceptable ranges specified by the ACR accreditation phantom testing instructions. Water HU values failed for one scanner model when scanned in helical mode using the widest collimation available and the Extension Plates were not present. All other scanner models passed the water HU linearity test with or without the Extension Plates in both axial and helical scan modes. Three of the four manufacturers tested failed the linearity test for different materials. The presence of the Extension Plates only affected the HU measurement for the water test object.

Factors Affecting Dimensional Accuracy of 3-D Printed Anatomical Structures Derived from CT Data.

Additive manufacturing and bio-printing, with the potential for direct fabrication of complex patient-specific anatomies derived from medical scan data, are having an ever-increasing impact on the practice of medicine. Anatomic structures are typically derived from CT or MRI scans, and there are multiple steps in the model derivation process that influence the geometric accuracy of the printed constructs. In this work, we compare the dimensional accuracy of 3-D printed constructs of an L1 vertebra derived from CT data for an ex vivo cadaver T-L spine with the original vertebra. Processing of segmented structures using binary median filters and various surface extraction algorithms is evaluated for the effect on model dimensions. We investigate the effects of changing CT reconstruction kernels by scanning simple geometric objects and measuring the impact on the derived model dimensions. We also investigate if there are significant differences between physical and virtual model measurements. The 3-D models were printed using a commercial 3-D printer, the Replicator 2 (MakerBot, Brooklyn, NY) using polylactic acid (PLA) filament. We found that changing parameters during the scan reconstruction, segmentation, filtering, and surface extraction steps will have an effect on the dimensions of the final model. These effects need to be quantified for specific situations that rely on the accuracy of 3-D printed models used in medicine or tissue engineering applications.

KERMA ratios in pediatric CT dosimetry.

BACKGROUND: Patient organ doses may be estimated from CTDI values. More accurate estimates may be obtained by measuring KERMA (Kinetic Energy Released in Matter) in anthropomorphic phantoms and referencing these values to free-in-air X-ray intensity. OBJECTIVE: To measure KERMA ratios (R(K)) in pediatric phantoms at CT. MATERIALS & METHODS: CT scans produce an air KERMA K in a phantom and an air KERMA K(CT) at isocenter. KERMA ratios (R(K)) are defined as (K/K(CT)), measured using TLD chips in phantoms representing newborns to 10-year-olds. RESULTS: R(K) in the newborn is approximately constant. For the other phantoms, there is a peak R(K) value in the neck. The median R(K) values for the GE scanner at 120 kV were 0.92, 0.83, 0.77 and 0.76 for newborns, 1-year-olds, 5-year-olds and 10-year-olds, respectively. Organ R(K) values were 0.91 ± 0.04, 0.84 ± 0.07, 0.74 ± 0.09 and 0.72 ± 0.10 in newborns, 1-year-olds, 5-year-olds and 10-year-olds, respectively. At 120 kV, a Siemens Sensation 16 scanner had R(K) values 5% higher than those of the GE LightSpeed Ultra. CONCLUSION: KERMA ratios may be combined with air KERMA measurements at the isocenter to estimate organ doses in pediatric CT patients.

In-patient to isocenter KERMA ratios in CT.

PURPOSE: To estimate in-patient KERMA for specific organs in computed tomography (CT) scanning using ratios to isocenter free-in-air KERMA obtained using a Rando phantom. METHOD: A CT scan of an anthropomorphic phantom results in an air KERMA K at a selected phantom location and air kerma K(CT) at the CT scanner isocenter when the scan is repeated in the absence of the phantom. The authors define the KERMA ratio (R(K)) as K∕ K(CT), which were experimentally determined in a Male Rando Phantom using lithium fluoride chips (TLD-100). R(K) values were obtained for a total of 400 individual point locations, as well as for 25 individual organs of interest in CT dosimetry. CT examinations of Rando were performed on a GE LightSpeed Ultra scanner operated at 80 kV, 120 kV, and 140 kV, as well as a Siemens Sensation 16 operated at 120 kV. RESULTS: At 120 kV, median R(K) values for the GE and Siemens scanners were 0.60 and 0.64, respectively. The 10th percentile R(K) values ranged from 0.34 at 80 kV to 0.54 at 140 kV, and the 90th percentile R(K) values ranged from 0.64 at 80 kV to 0.78 at 140 kV. The average R(K) for the 25 Rando organs at 120 kV was 0.61 ± 0.08. Average R(K) values in the head, chest, and abdomen showed little variation. Relative to R(K) values in the head, chest, and abdomen obtained at 120 kV, R(K) values were about 12% lower in the pelvis and about 58% higher in the cervical spine region. Average R(K) values were about 6% higher on the Siemens Sensation 16 scanner than the GE LightSpeed Ultra. Reducing the x-ray tube voltage from 120 kV to 80 kV resulted in an average reduction in R(K) value of 34%, whereas increasing the x-ray tube voltage to 140 kV increased the average R(K) value by 9%. CONCLUSIONS: In-patient to isocenter relative KERMA values in Rando phantom can be used to estimate organ doses in similar sized adults undergoing CT examinations from easily measured air KERMA values at the isocenter (free in air). Conversion from in-patient air KERMA values to tissue dose would require the use of energy-appropriate conversion factors.

Optimizing dose in computed tomographic guided procedures.

In the recent past, computed tomography (CT) use has grown by approximately 10% per year, with 62 million examinations performed in the USA in 2006. While these studies make up only 15% of the total number of radiologic examinations, they contribute approximately half of the public exposure due to medical radiation. CT-guided procedures comprise a small, but important minority of the total number of CT scans performed each year, as a significant dose can be delivered in this setting. This article reviews techniques to optimize patient exposure during CT-guided interventions.

An exposure indicator for digital radiography: AAPM Task Group 116 (executive summary).

Digital radiographic imaging systems, such as those using photostimulable storage phosphor, amorphous selenium, amorphous silicon, CCD, and MOSFET technology, can produce adequate image quality over a much broader range of exposure levels than that of screen/film imaging systems. In screen/film imaging, the final image brightness and contrast are indicative of over- and underexposure. In digital imaging, brightness and contrast are often determined entirely by digital postprocessing of the acquired image data. Overexposure and underexposures are not readily recognizable. As a result, patient dose has a tendency to gradually increase over time after a department converts from screen/film-based imaging to digital radiographic imaging. The purpose of this report is to recommend a standard indicator which reflects the radiation exposure that is incident on a detector after every exposure event and that reflects the noise levels present in the image data. The intent is to facilitate the production of consistent, high quality digital radiographic images at acceptable patient doses. This should be based not on image optical density or brightness but on feedback regarding the detector exposure provided and actively monitored by the imaging system. A standard beam calibration condition is recommended that is based on RQA5 but uses filtration materials that are commonly available and simple to use. Recommendations on clinical implementation of the indices to control image quality and patient dose are derived from historical tolerance limits and presented as guidelines.

Converting dose-length product to effective dose at CT.

PURPOSE: To determine effective dose (ED) per unit dose-length product (DLP) conversion factors for computed tomographic (CT) dosimetry. MATERIALS AND METHODS: A CT dosimetry spreadsheet was used to compute patient ED values and corresponding DLP values. The ratio of ED to DLP was determined with 16-section CT scanners from four vendors, as well as with five models from one manufacturer that spanned more than 25 years. ED-to-DLP ratios were determined for 2-cm scan lengths along the patient axis, as well as for typical scan lengths encountered at head and body CT examinations. The dependence of the ratio of ED to DLP on x-ray tube voltage (in kilovolts) was investigated, and the values obtained with the spreadsheet were compared with those obtained by using two other commercially available CT dosimetry software packages. RESULTS: For 2-cm scan lengths, changes in the scan region resulted in differences to ED of a factor of 30, but much lower variation was obtained for typical scan lengths at clinical head and body imaging. Inter- and intramanufacturer differences for ED/DLP were generally small. Representative values of ED/DLP at 120 kV were 2.2 microSv/mGy x cm (head scans), 5.4 microSv/mGy x cm (cervical spine scans), and 18 microSv/mGy x cm (body scans). For head scans, ED/DLP was approximately independent of x-ray tube voltage, but for body scans, the increase from 80 to 140 kV increased the ratio of ED to DLP by approximately 25%. Agreement in ED/DLP data for all three software packages was generally very good, except for cervical spine examinations where one software package determined an ED/DLP ratio that was approximately double that of the other two. CONCLUSION: This article describes a method of providing CT users with a practical and reliable estimate of adult patient EDs by using the DLP displayed on the CT console at the end of any given examination.

A method to obtain mean organ doses in a RANDO phantom.

A quantitative method of obtaining average organ dose from point measurements made in the male RANDO phantom is described for 24 compact organs of interest in patient dosimetry. A three-dimensional Cartesian coordinate system was created by considering each of the 36 RANDO phantom sections as the z coordinate, and using a rectangular frame to assign x and y coordinates relative to the center of each section. Anatomical atlases and clinical experience were used to identify the location and extent of each organ and tissue in the RANDO phantom. This proposed scheme is comparable to one used in a commercial phantom and offers investigators a comprehensive protocol for obtaining mean organ doses in the RANDO phantom.

Computing effective doses to pediatric patients undergoing body CT examinations.

BACKGROUND: The computation of patient effective doses to children is of particular interest given the relatively high doses received from this imaging modality, as well as the increased utilization of CT in all areas of medicine. Current methods for computing effective doses to children are relatively complex, and it would be useful to develop a simple method of computing pediatric effective doses for clinical purposes that could be used by radiologists and technologists. OBJECTIVE: To obtain pediatric effective doses for body CT examinations by the use of adult effective doses obtained from effective dose (E) per unit dose length product (DLP) coefficients, and energy imparted to a child relative to an adult. MATERIALS AND METHODS: Adult E/DLP coefficients were obtained at 120 kV using the ImPACT CT dosimetry spreadsheet. Patients were modeled as cylinders of water, and values of energy imparted to cylinders of varying radii were generated using Monte Carlo modeling. The amounts of energy imparted to the chest and abdomen of children relative to adults (R(en)) were obtained. Pediatric effective doses were obtained using scaling factors that accounted for scan length, mAs, patient weight, and relative energy imparted (R(en)). RESULTS: E/DLP values were about 16 microSv/mGy cm for males and about 19 microSv/mGy cm for females. R(en) at 120 kV for newborns was 0.35 for the chest and 0.49 for the abdomen. At constant mAs, the effective dose to 6-month-old patients undergoing chest CT examinations was found to be about 50% higher than that to adults, and for abdominal examinations about 100% higher. CONCLUSION: Adult effective doses can be obtained using DLP data and can be scaled to provide corresponding pediatric effective doses from body examinations on the same CT scanner.

Comparison of head and body organ doses in CT.

We compare head and body organ doses received by adult patients undergoing whole body scans operated using the same technique factors. Dosimetry data were obtained for six CT scanners (16 and 64 slice) from four vendors. Organ doses were obtained using the ImPACT CT dose software package for an adult male, together with the corresponding head and body CTDI(w). Our data provide a link between the CTDI doses generated on most commercial scanners for each clinical CT examination and doses to organs and tissues within the directly irradiated region of an adult patient. The average numerical ratio of the brain dose to the head phantom CTDI(w) is 0.84+/-0.05, the average ratio of the lung dose and liver dose to the body phantom CTDI(w) is 1.65+/-0.05 and 1.48+/-0.05, respectively. When scanned under identical conditions, lung doses are similar to brain doses, and liver doses are only approximately 10% lower. By comparison, the average body to head CTDI(w) ratio was 0.49+/-0.06, which erroneously implies that doses to organs in the head are twice those of doses to organs in the body at the same techniques. Two CT dosimetry phantom sizes are therefore not required, and our findings support the need to reassess the role, if any, of current cylindrical acrylic dosimetry phantoms.

How do lesion size and random noise affect detection performance in digital mammography?

RATIONALE AND OBJECTIVES: We investigated the effect of random noise and lesion size on detection performance in mammography. MATERIALS AND METHODS: Digital mammograms were obtained of an anthropomorphic breast phantom with and without simulated mass lesions. Digital versions of the mass lesions, ranging in size from 0.8 to 12 mm, were added back to the breast phantom image. Four alternate forced choice experiments were performed to determine the lesion contrast required to achieve a 92% correct lesion detection rate, denoted I92. Experiments were performed using identical phantom images and different versions of phantom images obtained using the same techniques but with different random noise patterns. RESULTS: For lesions larger than 1 mm, the slope of the contrast detail curves was always positive. This behavior contrasts with conventional contrast-detail curves in uniform backgrounds in which the slope is approximately -0.5. There was no difference between twinned experiments and those obtained using different patterns of random noise for lesions greater than 1 mm. When the lesion size was reduced below 1 mm, the detection threshold increased indicating a deterioration of lesion detectability, and detection performance was significantly lower when random noise patterns were used. CONCLUSION: Our results suggest that lesion detection is dominated by anatomical structure for lesions with a size >1 mm, but by random noise for submillimeter sized lesions.

Performance of a single lookup table (LUT) for displaying chest CT images.

RATIONALE AND OBJECTIVES: To evaluate the ability to view mediastinal and lung information on one window setting by processing images with a bilinear lookup table (LUT). MATERIALS AND METHODS: Chest computed tomography (CT) studies were obtained from 32 consecutive adult patient studies, which included 7 iodine contrast studies. From each CT examination, four sections were selected containing the suprasternal notch, carina, right inferior pulmonary vein, and the dome of the lower hemidiaphragm. Each image was processed with a bilinear LUT in addition to the lung setting (window width 1500 and window level -500) and mediastinal setting (window width 450 and window level 50) normally used. Seven radiologists compared the quality of the bilinear LUT with the corresponding lung and mediastinum display settings. A five-point scale was used to assess image quality, with a score of 5 being equivalent to the corresponding lung (or mediastinum) display, 3 being acceptable, and 1 being unacceptable. RESULTS: The average score of the bilinear LUT for all images and readers was 3.90 +/- 0.93 for lung information and 3.17 +/- 1.00 for mediastinum information. Use of the bilinear LUT resulted in unacceptable images in 0.3% cases for lung information and 5.9% for mediastinum information. Chest CT images that contained iodinated contrast resulted in a higher score than those obtained without contrast, but these differences did not achieve statistical significance. CONCLUSION: Use of a bilinear LUT has the potential to significantly improve operational efficiency with acceptable image quality for most chest CT images.

Patient size and x-ray transmission in body CT.

Physical characteristics were obtained for 196 patients undergoing chest and abdomen computed tomography (CT) examinations. Computed tomography sections for these patients having no evident pathology were analyzed to determine patient dimensions (AP and lateral), together with the average attenuation coefficient. Patient weights ranged from approximately 3 kg to about 120 kg. For chest CT, the mean Hounsfield unit (HU) fell from about -120 HU for newborns to about -300 HU for adults. For abdominal CT, the mean HU for children and normal-sized adults was about 20 HU, but decreased to below -50 HU for adults weighing more than 100 kg. The effective photon energy and percent energy fluence transmitted through a given patient size and composition was calculated for representative x-ray spectra at 80, 100, 120, and 140 kV tube potentials. A 70-kg adult scanned at 120 kVp transmits 2.6% of the energy fluence for chest and 0.7% for abdomen CT examinations. Reducing the patient size to 10 kg increases transmission by an order of magnitude. For 70 kg patients, effective energies in body CT range from approximately 50 keV at 80 kVp to approximately 67 keV at 140 kVp; increasing patient size from 10 to 120 kg resulted in an increase in effective photon energy of approximately 4 keV. The x-ray transmission data and effective photon energy data can be used to determine CT image noise and image contrast, respectively, and information on patient size and composition can be used to determine patient doses.

Use of the VIP-Man model to calculate energy imparted and effective dose for x-ray examinations.

A male human tomographic model was used to calculate values of energy imparted (epsilon) and effective dose (E) for monoenergetic photons (30-150 keV) in radiographic examinations. Energy deposition in the organs and tissues of the human phantom were obtained using Monte Carlo simulations. Values of E/epsilon were obtained for three common projections [anterior-posterior (AP), posterior-anterior (PA), and lateral (LAT)] of the head, cervical spine, chest, and abdomen, respectively. For head radiographs, all three projections yielded similar E/epsilon values. At 30 keV, the value of E/epsilon was approximately 1.6 mSv J(-1), which is increased to approximately 7 mSv J(-1) for 150 keV photons. The AP cervical spine was the only projection investigated where the value of E/epsilon decreased with increasing photon energy. Above 70 keV, cervical spine E/epsilon values showed little energy dependence and ranged between approximately 8.5 mSv J(-1) for PA projections and approximately 17 mSv J(-1) for AP projections. The values of E/epsilon for AP chest examinations showed very little variation with photon energy, and had values of approximately 23 mSv J(-1). Values of E/epsilon for PA and LAT chest projections were substantially lower than the AP projections and increased with increasing photon energy. For abdominal radiographs, differences between the PA and LAT projections were very small. All abdomen projections showed an increase in the E/epsilon ratio with increasing photon energy, and reached a maximum value of approximately 13.5 mSv J(-1) for AP projections, and approximately 9.5 mSv J(-1) for PA/lateral projections. These monoenergetic E/epsilon values can generate values of E/epsilon for any x-ray spectrum, and can be used to convert values of energy imparted into effective dose for patients undergoing common head and body radiological examinations.

Experimental investigation of the dose and image quality characteristics of a digital mammography imaging system.

Our purpose in this study was to investigate the image quality and absorbed dose characteristics of a digital mammography imaging system with a CsI scintillator, and to identify an optimal x-ray tube voltage for imaging simulated masses in an average size breast with 50% glandularity. Images were taken of an ACR accreditation phantom using a LORAD digital mammography system with a Mo target and a Mo filter. In one experiment, exposures were performed at 80 mAs with x-ray tube voltages varying between 24 and 34 kVp. In a second experiment, the x-ray tube voltage was kept constant at 28 kVp and the technique factor was varied between 5 and 500 mAs. The average glandular dose at each x-ray tube voltage was determined from measurements of entrance skin exposure and x-ray beam half-value layer. Image contrast was measured as the fractional digital signal intensity difference for the image of a 4 mm thick acrylic disk. Image noise was obtained from the standard deviation in a uniformly exposed region of interest expressed as a fraction of the background intensity. The measured digital signal intensity was proportional to the mAs and to the kVp5.8. Image contrast was independent of mAs, and dropped by 21% when the x-ray tube voltage increased from 24 to 34 kVp. At a constant x-ray tube voltage, image noise was shown to be approximately proportional to (mAs)(-05), which permits the image contrast to noise ratio (CNR) to be modified by changing the mAs. At 80 mAs, increasing the x-ray tube voltage from 24 to 34 kVp increased the CNR by 78%, and increased the average glandular dose by 285%. At a constant lesion CNR, the lowest average glandular dose value occurred at 27.3 kVp. Increasing or decreasing the x-ray tube voltage by 2.3 kVp from the optimum kVp increased the average glandular dose values by 5%. These results show that imaging simulated masses in a 4.2 cm compressed breast at approximately 27 kVp with a Mo/Mo target/filter results in the lowest average glandular dose.

Radiographic techniques in screen-film mammography.

The objectives of this study were to document imaging physics parameters associated with mammography physics surveys, and investigate how the choice of tube potential affects average glandular dose (AGD) and x-ray exposure time. Data from 60 mammography units were obtained pertaining to representative values of mAs, exposure time, half value layer, AGD and film density when acquiring phantom images. The survey of clinical systems showed that for a normal sized breast as represented by the mammography accreditation phantom, 60% of these units were operated at 25 kVp, and 33% at 26 kVp. Median exposure times were 1.14 s at 25 kVp and 0.73 s at 26 kVp. The median AGD was 1.62 mGy at 25 kVp and 1.51 mGy at 26 kVp. As expected, the choice of x-ray tube potential did not significantly affect the median film density value of 1.5. Five clinical systems, all from different vendors, had measurements performed of the AGD and x-ray exposure time as a function of x-ray tube potential at a constant film density. For a typical clinical x-ray unit, increasing the x-ray tube potential from 25 to 28 kVp reduced the exposure time by 50%, and reduced the AGD by 26%.

How good is the ACR accreditation phantom for assessing image quality in digital mammography?

RATIONALE AND OBJECTIVES: The purpose of this study was to evaluate the American College of Radiology (ACR) accreditation phantom for assessing image quality in digital mammography. MATERIALS AND METHODS: Digital images were obtained of an ACR accreditation phantom at varying mAs (constant kVp) and varying kVp (constant mAs). The average glandular dose for a breast with 50% glandularity was determined for each technique factor. Images were displayed on a 5 mega-pixel monitor, with the window width and level settings individually optimized for viewing the fibers, specks, and masses in the ACR phantom. Digital images of the ACR phantom were presented in a random manner to eight observers, each of whom indicated the number of objects visible in each image. RESULTS: Intraobserver variability was greater than interobserver variability for the detection of fibers and specks, but the reverse was true for the detection of masses. As the mAs increased, the number of fibers visible increased from less than one at 5 mAs to all six being visible at 80 mAs. The corresponding number of visible specks increased from 12 to 24, and the number of visible masses increased from 1.25 to about four. Above 26 kVp, object visibility was constant with increasing x-ray tube voltage. Reducing the x-ray tube voltage to 24 kVp, however, reduced the number of visible fibers from six to five, the number of visible specks from 24 to 21.1, and the number of visible masses from four to 3.1. Observer performance was approximately constant for average glandular doses greater than 1.6 mGy, so that the range of lesion detectability in the ACR phantom occurs at doses lower than those normally encountered in clinical practice. CONCLUSION: The current design of the ACR phantom is unsatisfactory for assessing image quality in digital mammography.