Artificial intelligence guided physician directive improves head and neck planning quality and practice Uniformity: A...

Area of Science:

• Medical Physics
• Radiation Oncology
• Artificial Intelligence in Medicine

Background:

• Physician practice patterns in radiation oncology can exhibit variability, impacting treatment plan consistency.
• Integrating advanced decision support tools is crucial for standardizing care, especially during practice expansion.
• A retrospective evaluation identified significant non-uniformity among physicians before AI implementation.

Purpose of the Study:

• To assess the impact of an artificial intelligence (AI) dose predictor as a physician decision support tool (DST) on practice uniformity.
• To evaluate the AI DST's effectiveness in standardizing treatment planning following the integration of new physicians.
• To quantify improvements in treatment plan adherence post-AI DST implementation.

Main Methods:

• An AI dose predictor model was developed based on the standard practice and integrated into the treatment planning system.
• The AI DST was implemented during a period of practice expansion involving three new physicians.
• A phase 1 retrospective evaluation compared treatment plan adherence before and after AI DST utilization.

Main Results:

• Prior to AI DST implementation, only 52.9% of treatment plans achieved physician directives due to practice non-uniformity among three physicians.
• Following the integration of the AI DST, a significant improvement in practice uniformity was observed.
• The frequency of clinical plans achieving AI-guided physician directives increased to 80.4% after AI DST implementation.

Conclusions:

• AI dose predictors, when integrated as physician decision support tools, can effectively enhance practice uniformity in radiation oncology.
• The AI DST demonstrated a substantial improvement in treatment plan consistency, achieving a higher rate of physician directive adherence.
• Implementing AI-driven decision support is a viable strategy to standardize care and improve quality in evolving clinical practices.