AI-Driven Advancements in Mammography: Assessing Efficiency and Burnout to Enhance Patient Care
Description: In California we often have six month wait times to schedule a mammogram along with a growing shortage of radiologists to interpret mammograms and rising radiologist burnout. Most mammograms are now three-dimentional (3D) tomography exams, which have more than 50 pictures of each breast instead of just a single two-dimentional picture. These 3D exams require longer interpretation times, leading to further exhaustion for radiologists, longer patient wait times, delays in breast cancer diagnoses, and poor quality of care. While artificial intelligence tools may improve accuracy, its impact on interpretation time and radiologists' well-being remains unknown. We will conduct a two-year randomized clinical trial on the impact of artificial intelligence on screening mammography accuracy involving over 90,000 exams at UCLA Health. We will assess whether artificial intelligence will catch cancers missed by radiologists and reduce false positives. During the trial radiologists will interpret half of the images with and half without artificial intelligence support. For this California Breast Cancer Research Program award, we propose adding new data collection during the trial to allow us to study the concurrent impact of artificial intelligence on efficiency and wait times. We aim to determine if artificial intelligence helps radiologists interpret mammograms faster without compromising accuracy and if it can reduce their fatigue and burnout. We will begin by consulting UCLA breast radiologists and surveying California practitioners to get their perspectives on artificial intelligence tools. Collaborating with patient advocates, we'll ensure the survey captures patient-centric concerns. We will also track interpretation time during the trial, allowing us to evaluate efficiency along with accuracy, while post-trial discussions will provide insight into radiologists' experiences. Our study will offer new insights into integrating artificial intelligence tools into breast cancer screening, informing real-world practice. By understanding artificial intelligence’s impact on efficiency and burnout, we aim to optimize its implementation statewide, empowering radiologists to focus on complex cases, patient care, and diverse communities. This pioneering research promises more effective and patient-centered breast cancer screening practices in California and beyond.
Amount: $927,000
Start date: 2024-08-01
Investigators: Joann Elmore, Joann Elmore
Institution: David Geffen School of Medicine at UCLA
Funded outputs count: 0
Funder award ID: g97c52
OpenAlex ID: G4368940187