Comparison of image quality among three X-ray systems for chest radiography: first step in optimisation

Physical Image Quality Metrics for the Characterization of X-ray Systems Used in Fluoroscopy-Guided Pediatric Cardiac Interventional Procedures: A Systematic Review

Pediatric interventional cardiology procedures are essential in diagnosing and treating congenital heart disease in children; however, they raise concerns about potential radiation exposure. Managing radiation doses and assessing image quality in angiographs becomes imperative for safe and effective interventions. This systematic review aims to comprehensively analyze the current understanding of physical image quality metrics relevant for characterizing X-ray systems used in fluoroscopy-guided pediatric cardiac interventional procedures, considering the main factors reported in the literature that influence this outcome. A search in Scopus and Web of Science, using relevant keywords and inclusion/exclusion criteria, yielded 14 relevant articles published between 2000 and 2022. The physical image quality metrics reported were noise, signal-to-noise ratio, contrast, contrast-to-noise ratio, and high-contrast spatial resolution. Various factors influencing image quality were investigated, such as polymethyl methacrylate thickness (often used to simulate water equivalent tissue thickness), operation mode, anti-scatter grid presence, and tube voltage. Objective evaluations using these metrics ensured impartial assessments for main factors affecting image quality, improving the characterization of fluoroscopic X-ray systems, and aiding informed decision making to safeguard pediatric patients during procedures.

Radiation dose reduction and image quality evaluation for lateral lumbar spine projection

Purpose

Optimization studies in digital radiology help to reduce the radiological risk to patients and maximize the benefits associated with their clinical purpose. The aim of this study was to assess the optimization of lateral lumbar spine projection via a combination of exposure parameters adjustments and additional filtration using a sectional anthropomorphic phantom.

Materials and methods

We evaluated the effects of peak voltage, tube loading, and low-cost filters made of copper, titanium, brass, and nickel on both the perceived and physical quality of 125 radiographs obtained in a computer radiography system. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) with their Figure of Merit (FOM), based on the entrance surface air kerma with backscatter (ESAK), was used to assess physical image quality.

Results

The standard image had a perceived image quality, SNR, FOMSNR, CNR, FOMCNR and ESAK of 3.4, 22.3, 386.4, 23.6, 433.7 and 1.28 mGy, respectively. Copper (90.3% purity) and titanium (95.0% purity) filters reduced ESAK by an average of 60% without compromising diagnostic quality, while brass and nickel filters increased dose under the conditions of the study.

Conclusions

Our findings show that optimizing lumbar spine projection can reduce radiation dose without compromising image quality. Low-cost copper and titanium filters can be valuable in resource-limited settings. Further research can explore additional strategies for radiological optimization.

Effect of varying X-ray tube voltage and additional filtration on image quality and patient dose in digital radiography system

This study investigated the effect of varying x-ray tube voltage and additional filtration thicknesses on radiation dose and image quality in digital radiography system. The polymethylmethacrylate (PMMA) phantoms of different thicknesses simulating both the adult chest and abdomen and the pediatric patient's chest examinations were used. X-ray tube voltage range of 70-125 kVp was used for adult patient chest radiography, 70-100 kVp for adult patient abdominal radiography, and 50-70 kVp for pediatric 1-year-old chest examination. 0.1-0.3 mm Cu and 1.0 mm Al filters were used as additional filters. Patient doses were measured with an ionization chamber, considering the irradiation parameters recommended for radiographic examinations performed in radiology clinics in the EUR 16260 protocol. The Entrance Skin Dose (ESD) was calculated from the air kerma value measured at the entrance surface of the PMMA phantoms. Effective dose values were calculated by employing PCXMC 2.0 program. For image quality evaluations, CDRAD, LCD-4, Beam stop and Huttner test object used together with PMMA phantoms and Alderson RS-330 Lung/Chest phantom were used. Figure of Merit (FOM), which allows quantitative assessment in terms of image quality and patient dose, has been calculated. Based on the calculated FOM values were evaluated at the tube voltages and additional filter thicknesses recommended in the EUR 16260 protocol. Entrance Skin Dose and Inverse Image Quality Figure (IQF_inv) value obtained from contrast detail analysis decreased with increasing filter thickness and tube voltage. Decrease in ESD and IQF_inv with increasing tube voltage without additional filter was 56% and 21% for adult chest radiography, 69% and 39% for adult abdominal radiography and 34% and 6% for 1-year-old pediatric chest radiography. When calculated FOM values are examined, it can be recommended to use a 0.1 mm Cu filter at 90 kVp and a 0.1 mm Cu + 1.0 mm Al filter at 125 kVp for adult chest radiography. For adult abdominal radiography, 0.2 mm Cu filter at 70 and 80 kVp and 0.1 mm Cu filter at 90 and 100 kVp were found to be appropriate. It was determined that the appropriate additional filter at 70 kVp for 1-year-old chest radiography was 1.0 mm Al+0.1 mm Cu.

Effect of Contrast Level and Image Format on a Deep Learning Algorithm for the Detection of Pneumothorax with Chest Radiography

Under the black-box nature in the deep learning model, it is uncertain how the change in contrast level and format affects the performance. We aimed to investigate the effect of contrast level and image format on the effectiveness of deep learning for diagnosing pneumothorax on chest radiographs. We collected 3316 images (1016 pneumothorax and 2300 normal images), and all images were set to the standard contrast level (100%) and stored in the Digital Imaging and Communication in Medicine and Joint Photographic Experts Group (JPEG) formats. Data were randomly separated into 80% of training and 20% of test sets, and the contrast of images in the test set was changed to 5 levels (50%, 75%, 100%, 125%, and 150%). We trained the model to detect pneumothorax using ResNet-50 with 100% level images and tested with 5-level images in the two formats. While comparing the overall performance between each contrast level in the two formats, the area under the receiver-operating characteristic curve (AUC) was significantly different (all p < 0.001) except between 125 and 150% in JPEG format (p = 0.382). When comparing the two formats at same contrast levels, AUC was significantly different (all p < 0.001) except 50% and 100% (p = 0.079 and p = 0.082, respectively). The contrast level and format of medical images could influence the performance of the deep learning model. It is required to train with various contrast levels and formats of image, and further image processing for improvement and maintenance of the performance.