Peer Review Tool for General Radiography Technologists Improves Image Quality.

PURPOSE: Quality improvement is vital to ensure health-care providers meet optimal patient care standards. Within our jurisdiction, accreditation requires image peer review as part of the quality assurance program. We propose a method to improve quality assurance in radiography by implementing a novel software-based peer review system for radiography technologists. METHODS: This is a retrospective study. A peer review tool was developed in Microsoft Excel and Visual Basic. The tool has 14 image quality criteria, which were selected based on national and international criteria, each containing standardized answers ensuring a common scoring regime. The tool provides data analysis and storage of all peer reviews performed. Radiography supervisors utilized the tool to evaluate image quality of various body parts at 28 hospitals. The tool enabled each Medical Imaging Department to objectively score images at their own hospital. Approximately 2% of all radiographs were randomly chosen for peer review. Additionally, the tool allowed for regional analysis based on hospital, body part, and quality criterion. RESULTS: Initial findings exposed equipment-related issues such as worn imaging plates, artifacts, and poor exposures, which prompted increased preventative maintenance. Other documented issues included foreign objects, inadequate collimation and centering, and inconsistent usage of lead markers. After identifying quality assurance-related issues, hospitals implemented education, resulting in improved overall image quality scores in subsequent audits. CONCLUSION: The peer review tool helped identify and correct various issues affecting image quality and ensures our program meets required accreditation standards. Furthermore, staff found utilizing the tool to identify areas for improvement improved collaboration, ongoing education, and support between staff.

Scanner and kVp dependence of measured CT numbers in the ACR CT phantom.

Quality control testing of CT scanners in our region includes a measurement of CT numbers in the American College of Radiology (ACR) CT phantom using a standardized protocol. CT number values are clinically relevant in determining the composition of various tissues in the body. Accuracy is important in the characterization of tumors, assessment of coronary calcium, and identification of urinary stone composition. Effective quality control requires that tolerance ranges of CT number values be defined: a measured value outside the range indicates the need for further investigation and possible recalibration of the scanner. This paper presents the results of CT number measurements on 36 scanners (25 GE, 10 Siemens and 1 Toshiba) at each available kVp. Among the five materials (solid water, air, polyethylene, acrylic, bone-equivalent) the measured CT numbers exhibit manufacturer and kVp dependence, which should be taken into account when defining tolerances. With this scan protocol, air and solid water values are significantly higher on GE scanners than on Siemens scanners (p-value < 0.01 at each kVp). The CT numbers of polyethylene and acrylic increase with kVp, while the bone-equivalent CT number decreases. These results are used to define manufacturer- and kVp-specific tolerance ranges for the CT numbers of each material in this phantom, which will be used in our quality control program.

Measurement of joint kinematics using a conventional clinical single-perspective flat-panel radiography system.

PURPOSE: The ability to accurately measure joint kinematics is an important tool in studying both normal joint function and pathologies associated with injury and disease. The purpose of this study is to evaluate the efficacy, accuracy, precision, and clinical safety of measuring 3D joint motion using a conventional flat-panel radiography system prior to its application in an in vivo study. METHODS: An automated, image-based tracking algorithm was implemented to measure the three-dimensional pose of a sparse object from a two-dimensional radiographic projection. The algorithm was tested to determine its efficiency and failure rate, defined as the number of image frames where automated tracking failed, or required user intervention. The accuracy and precision of measuring three-dimensional motion were assessed using a robotic controlled, tibiofemoral knee phantom programmed to mimic a subject with a total knee replacement performing a stair ascent activity. Accuracy was assessed by comparing the measurements of the single-plane radiographic tracking technique to those of an optical tracking system, and quantified by the measurement discrepancy between the two systems using the Bland-Altman technique. Precision was assessed through a series of repeated measurements of the tibiofemoral kinematics, and was quantified using the across-trial deviations of the repeated kinematic measurements. The safety of the imaging procedure was assessed by measuring the effective dose of ionizing radiation associated with the x-ray exposures, and analyzing its relative risk to a human subject. RESULTS: The automated tracking algorithm displayed a failure rate of 2% and achieved an average computational throughput of 8 image frames/s. Mean differences between the radiographic and optical measurements for translations and rotations were less than 0.08 mm and 0.07° in-plane, and 0.24 mm and 0.6° out-of-plane. The repeatability of kinematics measurements performed using the radiographic tracking technique was better than ±0.09 mm and 0.12° in-plane, and ±0.70 mm and ±0.07° out-of-plane. The effective dose associated with the imaging protocol used was 15 µSv for 10 s of radiographic cine acquisition. CONCLUSIONS: This study demonstrates the ability to accurately measure knee-joint kinematics using a single-plane radiographic measurement technique. The measurement technique can be easily implemented at most clinical centers equipped with a modern-day radiographic x-ray system. The dose of ionizing radiation associated with the image acquisition represents a minimal risk to any subjects undergoing the examination.

A Radiographic Diagnostic Reference Level Survey Using Patient and Phantom Data.

A diagnostic reference level (DRL) survey was conducted for seven common radiographic projections across an integrated health region, covering 27 hospitals and clinics. The projections surveyed were Chest Posterior-Anterior (PA) and Lateral, Abdomen Supine and Upright, L-spine Anterior-Posterior (AP) and Lateral, and Pelvis. Dose area product (DAP) values were collected from patient examinations in 43 digital radiography (DR) rooms and in 18 conventional rooms which use computed radiography (CR). In each room, data were collected for between 10 and 20 patient exposures for each surveyed projection. In addition, for each projection and room, a DAP value was measured for the exposure of a uniform acrylic phantom of standardized thickness. DAP values in DR rooms were found to be significantly lower than in CR rooms (p < 0.05). Therefore, DR and CR rooms were analyzed separately. Based on the survey results, separate DRLs are presented for DR and CR rooms, for both patient and phantom DAP values. DRLs have been set at the 75th percentile values; 25th percentile and median values are also presented to characterize the range of observed values. When identifying which rooms were above the DRLs, the patient and phantom data identified a partially different set of rooms. Possible reasons for these differences, including uncertainties in the patient data, are discussed.

Quantification of in vivo implant wear in total knee replacement from dynamic single plane radiography.

An in vivo method to measure wear in total knee replacements was developed using dynamic single-plane fluoroscopy. A dynamic, anthropomorphic total knee replacement phantom with interchangeable, custom-fabricated components of known wear volume was created, and dynamic imaging was performed. For each frame of the fluoroscopy data, the relative location of the femoral and tibial components were determined, and the apparent intersection of the femoral component with the tibial insert was used to calculate wear volume, wear depth, and frequency of intersection. No difference was found between the measured and true wear volumes. The precision of the measurements was ±39.7 mm(3) for volume and ±0.126 mm for wear depth. The results suggest the system is capable of tracking wear volume changes across multiple time points in patients. As a dynamic technique, this method can provide both kinematic and wear measurements that may be useful for evaluating new implant designs for total knee replacements.

US-fluoroscopy registration for transcatheter aortic valve implantation.

Transcatheter aortic valve implantation is a minimally invasive alternative to open-heart surgery for aortic stenosis in which a stent-based bioprosthetic valve is delivered into the heart on a catheter. Limited visualization during this procedure can lead to severe complications. Improved visualization can be provided by live registration of transesophageal echo (TEE) and fluoroscopy images intraoperatively. Since the TEE probe is always visible in the fluoroscopy image, it is possible to track it using fiducial-based single-perspective pose estimation. In this study, inherent probe tracking performance was assessed, and TEE to fluoroscopy registration accuracy and robustness were evaluated. Results demonstrated probe tracking errors of below 0.6 mm and 0.2°, a 2-D RMS registration error of 1.5 mm, and a tracking failure rate of below 1%. In addition to providing live registration and better accuracy and robustness compared to existing TEE probe tracking methods, this system is designed to be suitable for clinical use. It is fully automatic, requires no additional operating room hardware, does not require intraoperative calibration, maintains existing procedure and imaging workflow without modification, and can be implemented in all cardiac centers at extremely low cost.