The practice of diagnostic radiology has undergone immense transformation due to technological and scientific progress. However, these advancements have not translated into a reduced workload for radiologists. An updated analysis of the 2024 medical imaging literature reveals that, rather than easing the burden, recent developments—especially in artificial intelligence —are contributing to a sustained and significant increase in workload. This trend holds true across both academic tertiary care centres and non-academic general teaching hospitals, with AI studies notably associated with workload intensification. Understanding these dynamics is critical for shaping workforce strategies, policy decisions and the future integration of emerging technologies in clinical settings.
Persistent Workload Growth Across Practice Settings
The review of 416 articles published in 2024 across major imaging and clinical journals indicated that over half had the potential to directly impact patient care. In both academic and non-academic settings, 56.5% of the studies reviewed were considered clinically relevant. Among these, nearly half, 48.9%, were expected to increase radiologists’ workload, while only 0.4% were projected to reduce it. This data mirrors findings from a similar 2019 analysis, highlighting a consistent trend of rising demands.
Must Read: Radiology Workforce: A New EU Approach
Most studies enhancing existing imaging applications did not reduce the time or complexity involved in image acquisition, processing or interpretation. In fact, a significant portion required longer post-processing and interpretation times, directly contributing to the workload. Such incremental changes, though often subtle, collectively burden radiology departments already under pressure from growing imaging volumes and resource limitations.
Artificial Intelligence: More Burden Than Relief
According to the study, AI has not yet delivered on its promise to alleviate workload in diagnostic radiology. On the contrary, studies with AI as their primary focus were found to be significantly more likely to increase workload. The odds ratio for workload increase associated with AI studies was 14.3 in academic centres and 13.7 in non-academic settings, both with high statistical significance. The integration of AI tools often necessitates additional steps in the workflow, including image post-processing and result verification, thereby increasing time demands. For example, deep learning models designed for predicting hematoma expansion may require additional post-processing pipelines and manual interpretation, extending the time radiologists must spend per case.
These outcomes suggest that current AI implementations may be more supportive than autonomous, functioning as supplementary layers rather than standalone solutions. In many cases, AI introduces new complexities without offsetting existing tasks, underscoring the need for more targeted development that addresses practical clinical burdens.
Strategic Implications for Health Systems
The continued escalation in diagnostic radiologists’ workload has critical implications for healthcare systems. Without corresponding increases in staffing or efficiency-enhancing tools, radiologists are at risk of burnout and diagnostic errors. This challenge is compounded by the nature of radiologists' work, of which image interpretation comprises only a part. Other responsibilities—such as protocol development, trainee supervision and interdisciplinary communication—consume considerable time and contribute to the overall clinical load. As the number of significant studies continues to increase, and considering that positive findings are more likely to be published and implemented in practice, the growth of new or expanded imaging tasks may unintentionally lead to ongoing workload pressures for radiologists. Health administrators must consider these realities in workforce planning and AI investment strategies. Importantly, future research and innovation should prioritise solutions that demonstrably reduce operational burdens rather than add new layers of complexity.
Recent scientific developments in diagnostic imaging, especially those involving artificial intelligence, are contributing to a continued rise in radiologists’ workload. Despite hopes that AI might alleviate pressures, most implementations so far have had the opposite effect. This upward trend in workload is observable across diverse clinical settings and is primarily driven by extended post-processing and interpretation requirements. Without proactive staffing and strategic planning, healthcare systems risk overburdening radiologists, thereby affecting care quality and provider well-being. To harness the true potential of innovation, future developments must focus not only on clinical effectiveness but also on tangible workflow optimisation.
Source: European Journal of Radiology
Image Credit: Freepik
