Study Objectives and Methods

‍

This section details the objectives and methods for both the initial one-month study and the extended six-month follow-up study, outlining the approach taken to assess the effectiveness of AZtrauma in improving radiological diagnostics.

‍

Initial One-Month Study

Objective:

To evaluate the immediate impact of AI-based radiologist worklist prioritization on report TAT for X-ray examinations positive for fractures, aiming to quantify operational efficiency and diagnostic accuracy enhancements shortly after AZtrauma's implementation.

Methods:

• Study Design: Utilized a retrospective, multi-center approach comparing TAT before and after AZtrauma's integration.
• Data Collection Period: Between February 15 and March 14, for 2022 (23,088 examinations without AI) and 2023 (27,594 examinations with AI), providing a direct comparison of operational metrics pre and post-AI implementation.
• Inclusion Criteria: Included all adult X-ray examinations (CR, RF, or DX) performed at SimonMed facilities within the specified periods, which had a completed radiology report.
• Exclusion Criteria: Excluded examinations lacking a final radiology report, not analyzed by AI (for the with AI period), and pediatric cases (under the age of 22).
• Analysis: PACS data was employed to calculate TAT, offering a measure of AZtrauma effect on improving diagnostic turnaround times.

Initial One-Month Study Results:

• Overall TAT Reduction: The integration of AZtrauma led to patients being diagnosed approximately 6 times faster on average, a profound decrease in TAT for fracture detection.
• Priority-Based TAT: The Al's prioritization system effectively distinguished between high and low urgency cases, optimizing TAT across different priority levels.
• Prevalence of Diagnosed Fractures: The prevalence of acute fractures based on radiology reports saw an increase from 10.6% in the period without Al (February 15 to March 14, 2022) to 11.5% in the period with Al (February 15 to March 14, 2023).

‍

Extended Six-Month Study

Objective:

Following the initial findings, the study was extended to six months to validate the sustainability and consistency of improvements observed with AZtrauma over a longer timeframe.

Methods:

• Study Period: Expanded to include data from January 1 to June 30, contrasting 2022 (159,601 examinations without AI) with 2023 (170,703 examinations with AI).
• Examination Volume and Prevalence: Detailed examination volumes and the prevalence of diagnosed fractures were analyzed to assess AZtrauma's broader operational and clinical impact.
• Data Analysis: Again, PACS data was employed for an extended comparison of TAT changes, ensuring a focus on AZtrauma's ability to maintain diagnostic accuracy and efficiency improvements over time.

Extended Six-Month Study Results:

• Sustained TAT Reduction: Over the six-month period, the consistent application of AZtrauma continued to yield a significant decrease in TAT, with patients being diagnosed much faster compared to the pre-Al period.
• Long-Term Diagnostic Consistency: The extended study validated AZtrauma’s high level of diagnostic performance over a longer timeframe, maintaining its sensitivity, specificity, and predictive values. This consistency attests to AZtrauma's reliability and the potential for broader applications within radiology.
• Prevalence Over Six Months: The extended study revealed a slight increase in the prevalence of diagnosed acute fractures from 10.4% in the period without AI (January 1 to June 30, 2022) to 11.8% with AI (January 1 to June 30, 2023).

‍

Alert Proximity Impact

Analysis of alert timings revealed that exams prioritized closer to alert issuance experienced faster TATs, emphasizing the importance of frequent updates for optimal efficiency. The extended study reaffirmed the role of alert frequency in maximizing operational efficiency.

‍

‍The Role of AI in Fracture Detection

‍

AI has transformed the field of radiology, presenting unprecedented opportunities to improve workflow efficiency, diagnostic accuracy, and patient care. It stands at the intersection of advanced computing, machine learning, and vast datasets, reshaping the operational dynamics and clinical outcomes of radiological practices. The following section will look into AI's multifaceted role in radiology, highlighting its effects and its broader implications for healthcare delivery.

‍

Improved Patient Care

The integration of AZtrauma has markedly improved patient care by significantly reducing the turnaround time for the diagnosis of fractures. With AI, on average, patients now receive their results up to 6 times faster than without, a transformation that not only accelerates treatment initiation but also positively impacts patient experience and outcomes.

‍

Decreased Likelihood of Missed Fractures

The high NPV and sensitivity of AZtrauma decrease the likelihood of missed fractures. This improvement in diagnostic accuracy ensures that fractures are identified more reliably, reducing the risk of complications associated with missed diagnoses.

‍

Reducing Radiologist Burnout

With the assistance of AI, radiologists experience decreased burnout and enjoy greater confidence in their diagnostic reads. The reduction in routine and manual image review tasks alleviates the cognitive load on radiologists, making their work more focused and efficient.

‍

Increased Efficiency and Productivity

By automating the prioritization of worklists and assisting in the rapid identification of fractures, AZtrauma allows radiologists to focus their expertise where it is most needed, enhancing the overall workflow and throughput of radiological assessments.

‍

‍

‍

Conclusion

The implementation of AZtrauma by SimonMed signifies an important turn in radiology, marking a significant improvement in healthcare delivery. This innovation in Artificial Intelligence has streamlined the diagnostic process for fractures, achieving a dramatic reduction in TAT and significantly elevating patient care quality by enabling diagnoses to be made 6 times faster on average. This initiative serves as a milestone in the evolution toward a more efficient, precise, and patient-focused healthcare system. The evident benefits realized from AZtrauma's deployment—ranging from operational efficiency improvements to diagnostic accuracy—illustrate the potential of AI to reshape radiology.

‍

In sum, the path to fully unleashing the capabilities of AI in healthcare is paved with continued innovation, in-depth research, and collaborative efforts spanning radiology, technology, and ethical considerations. The journey forward is geared toward not only expanding the application of AI within radiology but also exploring new horizons in patient care and treatment outcomes.