# Image Reject Patterns in Computed Radiography: Insights From a Ghanaian Radiology Department

## TL;DR

This study analyzed image rejection patterns in a Ghanaian hospital's radiography system, finding a high rejection rate due to positioning and anatomical errors, highlighting the need for better training and workflow improvements.

## Contribution

The study provides insights into image rejection patterns specific to a Ghanaian radiology department, emphasizing practical QA interventions for resource-limited settings.

## Key findings

- The overall image rejection rate was 16.5%, exceeding recommended thresholds.
- Anatomical cut-off and positioning errors were the most common causes of rejection.
- High rejection rates were observed in skull/sinus, pelvic, and abdomen examinations.

## Abstract

Image reject analysis is a critical quality assurance (QA) tool in diagnostic imaging, helping to minimise unnecessary radiation exposure and improve imaging efficiency. This study evaluates image rejection patterns in a computed radiography (CR) system at a major tertiary teaching hospital in Ghana, identifying key sources of errors and their implications for radiology practice.

A retrospective review of radiographic images acquired between April and June 2023 was conducted. Images, including those flagged as rejects were retrieved from the CR system and analysed for rejection rates, trends by anatomical region, and key error sources.

Of the 5889 images reviewed, 974 were rejected, resulting in an overall rejection rate of 16.5%. Rejection rates varied considerably across anatomical regions. High rejection rates were observed in skull/sinus (34.9%, n = 90/258), pelvic (29.9%, n = 88/294) and abdomen (26.9%, n = 84/312) examinations. Low rejects were recorded for ankle (1.8%, n = 2/110), humerus (2.4%, n = 2/82), forearm (6.7%, n = 6/90), elbow (9.7%, n = 6/62), and lower leg (7.5%, n = 16/214). Across all examinations, the three leading causes of image rejection were anatomical cut‐off (40.5%, n = 394), positioning errors (27.5%, n = 268), and beam centering errors (18.5%, n = 180). Less frequent causes included exposure‐related issues (6.6%, n = 64), patient movement (2.9%, n = 28), and artefacts or ghosting (4.1%, n = 40).

This study reinforces the role of image reject analysis as a valuable QA measure in CR systems. The high rejection rates observed highlight the need for targeted interventions in positioning, workflow optimization, and radiographer training, particularly in resource‐constrained settings to enhance diagnostic quality and patient safety.

This study examined image rejection patterns in a computed radiography system at a tertiary hospital in Ghana. An overall reject rate of 16.5% was observed, exceeding recommended thresholds, with anatomical cut‐off and positioning errors being the most common causes. The findings underscore the importance of targeted quality assurance strategies to improve imaging outcomes.