Exam- and Sequence-Level Energy Quantification In MRI - Revealing Opportunities for Energy Savings
Abstract
Purpose
MRI scanners contribute substantially to healthcare energy use, yet quantitative benchmarks at the exam- and pulse sequence-level are limited. The purpose of this work was to develop an automated framework for MRI energy quantification at the exam and sequence level for baseline characterization, and to identify high-yield opportunities for prioritizing future energy-savings studies and interventions.
Methods
The main electrical load of a 3T MRI scanner in an outpatient imaging facility was continuously metered for 7 months. Scanner log files were automatically extracted, parsed to extract comprehensive scanner-reported metadata required for exam and sequence level energy allocation, and temporally aligned with the power data. Exams were further categorized by application (Neuro, Breast, Body, and MSK) based on protocol naming conventions. Exam and sequence level energy use was then summarized using the mean, standard deviation (SD), and coefficient of variation (COV), with median and interquartile range (IQR) used for comparisons across application groups.
Results
Across 2,524 exams and 173 unique protocols, protocol-level mean (SD) energy-per-exam ranged from 4.26 kWh (0.63) for the “Breast Marker Localization” protocol to 42.13 kWh (9.25) for “Whole Spine”. Neuro was the most utilized application (N=1041 exams) with a median (IQR) energy-per-exam of 18.30 kWh (12.44) compared to 18.31 kWh (5.42; N=703), 11.79 kWh (5.84; N=537), and 11.55 kWh (2.92; N=153) for Body, MSK, and Breast respectively. The highest single energy-per-exam was 87.40 kWh for a “Brain Tumor” protocol. A total of 56 exams utilized the Brain Tumor protocol (27.31 kWh mean energy/exam) with overall high variability (COV = 49.08%).
Conclusion
An automated framework for exam- and sequence-level energy quantification revealed large protocol-dependent differences and substantial variability in MRI energy consumption, highlighting actionable opportunities for future studies focused on energy savings while preserving imaging performance.