Reject analysis was performed on 288,000 computed radiography (CR) image records collected from a university hospital (UH) and a large community hospital (CH). Each record contains image information, such as body part and view position, exposure level, technologist identifier, and--if the image was rejected--the reason for rejection. Extensive database filtering was required to ensure the integrity of the reject-rate calculations. The reject rate for CR across all departments and across all exam types was 4.4% at UH and 4.9% at CH. The most frequently occurring exam types with reject rates of 8% or greater were found to be common to both institutions (skull/facial bones, shoulder, hip, spines, in-department chest, pelvis). Positioning errors and anatomy cutoff were the most frequently occurring reasons for rejection, accounting for 45% of rejects at CH and 56% at UH. Improper exposure was the next most frequently occurring reject reason (14% of rejects at CH and 13% at UH), followed by patient motion (11% of rejects at CH and 7% at UH). Chest exams were the most frequently performed exam at both institutions (26% at UH and 45% at CH) with half captured in-department and half captured using portable x-ray equipment. A ninefold greater reject rate was found for in-department (9%) versus portable chest exams (1%). Problems identified with the integrity of the data used for reject analysis can be mitigated in the future by objectifying quality assurance (QA) procedures and by standardizing the nomenclature and definitions for QA deficiencies.

The measured physical quantities provide a robust reflection of perceptual image quality in clinical images. The methodology can be readily applied for automated evaluation of perceptual image quality in clinical chest radiographs.

A conventional approach to computed tomography (CT) or cone beam CT (CBCT) metal artifact reduction is to replace the X-ray projection data within the metal trace with synthesized data. However, existing projection or sinogram completion methods cannot always produce anatomically consistent information to fill the metal trace, and thus, when the metallic implant is large, significant secondary artifacts are often introduced. In this work, we propose to replace metal artifact affected regions with anatomically consistent content through joint projection-sinogram correction as well as adversarial learning. To handle the metallic implants of diverse shapes and large sizes, we also propose a novel mask pyramid network that enforces the mask information across the network's encoding layers and a mask fusion loss that reduces early saturation of adversarial training. Our experimental results show that the proposed projection-sinogram correction designs are effective and our method recovers information from the metal traces better than the state-of-the-art methods.

A large database of digital chest radiographs was developed over a 14-month period. Ten radiographic technologists and five radiologists independently evaluated a stratified subset of images from the database for quality deficiencies and decided whether each image should be rejected. The evaluation results showed that the radiographic technologists and radiologists agreed only moderately in their assessments. When compared against each other, radiologist and technologist reader groups were found to have even less agreement than the inter-reader agreement within each group. Radiologists were found to be more accepting of limitedquality studies than technologists. Evidence from the study suggests that the technologists weighted their reject decisions more heavily on objective technical attributes, while the radiologists weighted their decisions more heavily on diagnostic interpretability relative to the image indication. A suite of reject-detection algorithms was independently run on the images in the database. The algorithms detected 4 % of postero-anterior chest exams that were accepted by the technologist who originally captured the image but which would have been rejected by the technologist peer group. When algorithm results were made available to the technologists during the study, there was no improvement in inter-reader agreement in deciding whether to reject an image. The algorithm results do, however, provide new quality information that could be captured within a site-wide, reject-tracking database and leveraged as part of a site-wide QA program.