Smart chest X-ray worklist prioritization using artificial intelligence: a clinical workflow simulation

Abstract

Objective The aim is to evaluate whether smart worklist prioritization by artificial intelligence (AI) can optimize the radiology workflow and reduce report turnaround times (RTATs) for critical findings in chest radiographs (CXRs). Furthermore, we investigate a method to counteract the effect of false negative predictions by AI—resulting in an extremely and dangerously long RTAT, as CXRs are sorted to the end of the worklist. Methods We developed a simulation framework that models the current workflow at a university hospital by incorporating hospital-specific CXR generation rates and reporting rates and pathology distribution. Using this, we simulated the standard worklist processing “first-in, first-out” (FIFO) and compared it with a worklist prioritization based on urgency. Examination prioritization was performed by the AI, classifying eight different pathological findings ranked in descending order of urgency: pneumothorax, pleural effusion, infiltrate, congestion, atelectasis, cardiomegaly, mass, and foreign object. Furthermore, we introduced an upper limit for the maximum waiting time, after which the highest urgency is assigned to the examination. Results The average RTAT for all critical findings was significantly reduced in all prioritization simulations compared to the FIFO simulation (e.g., pneumothorax: 35.6 min vs. 80.1 min; p < 0.0001), while the maximum RTAT for most findings increased at the same time (e.g., pneumothorax: 1293 min vs 890 min; p < 0.0001). Our “upper limit” substantially reduced the maximum RTAT in all classes (e.g., pneumothorax: 979 min vs. 1293 min/1178 min; p < 0.0001). Conclusion Our simulations demonstrate that smart worklist prioritization by AI can reduce the average RTAT for critical findings in CXRs while maintaining a small maximum RTAT as FIFO. Key Points • Development of a realistic clinical workflow simulator based on empirical data from a hospital allowed precise assessment of smart worklist prioritization using artificial intelligence. • Employing a smart worklist prioritization without a threshold for maximum waiting time runs the risk of false negative predictions of the artificial intelligence greatly increasing the report turnaround time. • Use of a state-of-the-art convolution neural network can reduce the average report turnaround time almost to the upper limit of a perfect classification algorithm (e.g., pneumothorax: 35.6 min vs. 30.4 min).

Future Medical Artificial Intelligence Application Requirements and Expectations of Physicians in German University Hospitals: Web-Based Survey

“…This can be partly explained by the fact that it is a "high-level" term, often not mentioning concrete ML algorithms or models in a clear context. Examples of AI in medicine are AI applications to support diagnostic procedures, predict the course of the disease [13][14][15][16][17], enhance the potential of clinical decision support [18], and support the management of hospital workflows [19,20]. Thereby, AI offers the possibility to support physicians in delivering high-quality medicine and increasing medical care efficiency.…”

How does artificial intelligence in radiology improve efficiency and health outcomes?

“…An alternative method to reduce report turnaround times and promote early detection of critical findings includes worklist prioritization based on urgent findings detected by AI [26]. A study from a German university hospital simulated this concept on retrospective chest radiographs and found that turnaround times for reporting critical findings reduced from 80 min to 35-50 min [27]. In the United States, a commercial algorithm to prioritize intracranial hemorrhage resulted in reduced waiting time from 16 min to 12 min per positive case [28].…”

Chest radiographs and machine learning – Past, present and future

“…In addition to improving radiologist accuracy, machine learning models have the capacity to integrate with workflow systems to triage studies, identifying and serving high priority, time‐sensitive findings for faster reporting. Studies suggest that these systems reduce reporting time and alleviate radiologist workloads 28,34,35 . Triage functionality is likely to become more effective as machine learning solutions become more comprehensive; a model that only looks for a single finding (such as pneumothorax) can up‐triage cases where that pathology is identified, however, in doing so it will necessarily down‐triage cases with other serious or urgent problems (such as free sub‐diaphragmatic gas).…”

“…Studies suggest that these systems reduce reporting time and alleviate radiologist workloads. 28 , 34 , 35 Triage functionality is likely to become more effective as machine learning solutions become more comprehensive; a model that only looks for a single finding (such as pneumothorax) can up‐triage cases where that pathology is identified, however, in doing so it will necessarily down‐triage cases with other serious or urgent problems (such as free sub‐diaphragmatic gas).…”