Background: Awareness of energy efficiency has been rising in the industrial and residential sectors but only recently in the health care sector.Purpose: To measure the energy consumption of modern CT and MRI scanners in a university hospital radiology department and to estimate energy-and cost-saving potential during clinical operation. Materials and Methods:Three CT scanners, four MRI scanners, and cooling systems were equipped with kilowatt-hour energy measurement sensors (2-Hz sampling rate). Energy measurements, the scanners' log files, and the radiology information system from the entire year 2015 were analyzed and segmented into scan modes, as follows: net scan (actual imaging), active (room time), idle, and system-on and system-off states (no standby mode was available). Per-examination and peak energy consumption were calculated. Results:The aggregated energy consumption imaging 40 276 patients amounted to 614 825 kWh, dedicated cooling systems to 492 624 kWh, representing 44.5% of the combined consumption of 1 107 450 kWh (at a cost of U.S. $199 341). This is equivalent to the usage in a town of 852 people and constituted 4.0% of the total yearly energy consumption at the authors' hospital. Mean consumption per CT examination over 1 year was 1.2 kWh, with a mean energy cost (6standard deviation) of $0.22 6 0.13. The total energy consumption of one CT scanner for 1 year was 26 226 kWh ($4721 in energy cost). The net consumption per CT examination over 1 year was 3580 kWh, which is comparable to the usage of a two-person household in Switzerland; however, idle state consumption was fourfold that of net consumption (14 289 kWh). Mean MRI consumption over 1 year was 19.9 kWh per examination, with a mean energy cost of $3.57 6 0.96. The mean consumption for a year in the system-on state was 82 174 kWh per MRI examination and 134 037 kWh for total consumption, for an energy cost of $24 127. Conclusion:CT and MRI energy consumption is substantial. Considerable energy-and cost-saving potential is present during nonproductive idle and system-off modes, and this realization could decrease total cost of ownership while increasing energy efficiency.

ICU Cockpit: a secure, fast, and scalable platform for collecting multimodal waveform data, online and historical data visualization, and online validation of algorithms in the intensive care unit. We present a network of software services that continuously stream waveforms from ICU beds to databases and a web-based user interface. Machine learning algorithms process the data streams and send outputs to the user interface. The architecture and capabilities of the platform are described. Since 2016, the platform has processed over 89 billion data points (N = 979 patients) from 200 signals (0.5–500 Hz) and laboratory analyses (once a day). We present an infrastructure-based framework for deploying and validating algorithms for critical care. The ICU Cockpit is a Big Data platform for critical care medicine, especially for multimodal waveform data. Uniquely, it allows algorithms to seamlessly integrate into the live data stream to produce clinical decision support and predictions in clinical practice.

Explainable AI (XAI) is considered the number one solution for overcoming implementation hurdles of AI/ML in clinical practice. However, it is still unclear how clinicians and developers interpret XAI (differently) and whether building such systems is achievable or even desirable. This longitudinal multi-method study queries (n=112) clinicians and developers as they co-developed the DCIP – an ML-based prediction system for Delayed Cerebral Ischemia. The resulting framework reveals that ambidexterity between exploration and exploitation can help bridge opposing goals and requirements to improve the design and implementation of AI/ML in healthcare.

Explainable artificial intelligence (XAI) has emerged as a promising solution for addressing the implementation challenges of AI/ML in healthcare. However, little is known about how developers and clinicians interpret XAI and what conflicting goals and requirements they may have. This paper presents the findings of a longitudinal multi-method study involving 112 developers and clinicians co-designing an XAI solution for a clinical decision support system. Our study identifies three key differences between developer and clinician mental models of XAI, including opposing goals (model interpretability vs. clinical plausibility), different sources of truth (data vs. patient), and the role of exploring new vs. exploiting old knowledge. Based on our findings, we propose design solutions that can help address the XAI conundrum in healthcare, including the use of causal inference models, personalized explanations, and ambidexterity between exploration and exploitation mindsets. Our study highlights the importance of considering the perspectives of both developers and clinicians in the design of XAI systems and provides practical recommendations for improving the effectiveness and usability of XAI in healthcare.