Erratum: On the role of artificial intelligence in medical imaging of COVID-19.

[This corrects the article DOI: 10.1016/j.patter.2021.100269.].

The Role of Imaging Informatics in Disaster Preparedness During the COVID-19 Pandemic.

Disaster preparedness is a major but necessary undertaking for every health care facility. The 2019 coronavirus (SARS-CoV-2) provided an unforeseen opportunity to compare the response of two radiology departments, University Health System A (UHSA) and University Health System B (UHSAB). Preparing for this disaster was unique since though unexpected, was thought to be detected early enough to allow for sufficient preparation. Unlike many other disasters which are short-term, single events, this has been an on-going event. Changes at both health systems included workflow alterations for exposure reduction to faculty, trainees, and staff. UHSA was able to quickly divert workflow to previously deployed home workstations, while University of Utah Health Sciences Center required 2 to 3 weeks to procure and initialize enough remote workstations to significantly affect departmental operations. Other measures such as universal masking, temperature screening at facility entrances, virtual appointments, and physical barriers were used by both systems to limit patient-to-patient, patient-to-staff, staff-to-patient, and staff-staff physical interaction to help decrease exposure risk. The goal of these preparations is to allow each department to fulfill imaging needs in support of the organizational clinical mission with the flexibility to adapt to the unique and dynamic nature of this disaster.

On the role of artificial intelligence in medical imaging of COVID-19.

Although a plethora of research articles on AI methods on COVID-19 medical imaging are published, their clinical value remains unclear. We conducted the largest systematic review of the literature addressing the utility of AI in imaging for COVID-19 patient care. By keyword searches on PubMed and preprint servers throughout 2020, we identified 463 manuscripts and performed a systematic meta-analysis to assess their technical merit and clinical relevance. Our analysis evidences a significant disparity between clinical and AI communities, in the focus on both imaging modalities (AI experts neglected CT and ultrasound, favoring X-ray) and performed tasks (71.9% of AI papers centered on diagnosis). The vast majority of manuscripts were found to be deficient regarding potential use in clinical practice, but 2.7% (n = 12) publications were assigned a high maturity level and are summarized in greater detail. We provide an itemized discussion of the challenges in developing clinically relevant AI solutions with recommendations and remedies.

Crowdsourcing pneumothorax annotations using machine learning annotations on the NIH chest X-ray dataset.

Pneumothorax is a potentially life-threatening condition that requires prompt recognition and often urgent intervention. In the ICU setting, large numbers of chest radiographs are performed and must be interpreted on a daily basis which may delay diagnosis of this entity. Development of artificial intelligence (AI) techniques to detect pneumothorax could help expedite detection as well as localize and potentially quantify pneumothorax. Open image analysis competitions are useful in advancing state-of-the art AI algorithms but generally require large expert annotated datasets. We have annotated and adjudicated a large dataset of chest radiographs to be made public with the goal of sparking innovation in this space. Because of the cumbersome and time-consuming nature of image labeling, we explored the value of using AI models to generate annotations for review. Utilization of this machine learning annotation (MLA) technique appeared to expedite our annotation process with relatively high sensitivity at the expense of specificity. Further research is required to confirm and better characterize the value of MLAs. Our adjudicated dataset is now available for public consumption in the form of a challenge.

Fast anatomy segmentation by combining coarse scale multi-atlas label fusion with fine scale corrective learning.

Deformable registration based multi-atlas segmentation has been successfully applied in a broad range of anatomy segmentation applications. However, the excellent performance comes with a high computational burden due to the requirement for deformable image registration and voxel-wise label fusion. To address this problem, we investigate the role of corrective learning (Wang et al., 2011) in speeding up multi-atlas segmentation. We propose to combine multi-atlas segmentation with corrective learning in a multi-scale analysis fashion for faster speeds. First, multi-atlas segmentation is applied in a low spatial resolution. After resampling the segmentation result back to the native image space, learning-based error correction is applied to correct systematic errors due to performing multi-atlas segmentation in a low spatial resolution. In cardiac CT and brain MR segmentation experiments, we show that applying multi-atlas segmentation in a coarse scale followed by learning-based error correction in the native space can substantially reduce the overall computational cost, with only modest or no sacrificing segmentation accuracy.

Automatic segmentation of the left ventricle into 17 anatomical regions in cardiac MR imaging.

Analysis and characterization of anatomical segments in the left ventricle (LV) of the heart in cardiac MRI convey clinical significance. Based on the standard defined by the American Heart Association (AHA), the LV is divided into 17 anatomical segments. In this paper, we propose a novel method to automatically partition the LV into 17 segments, which allows automated analysis of these segments. Our method starts with assigning each slice with a section tag by using the papillary muscles and the LV cavity as references: basal, mid-cavity, apical and apex. It then partitions each slice into 4 or 6 segments by extracting the relevant points on the outer circle of a fitted cylinder and identifying the image orientation by using the lung as a reference. We evaluate our method on 45 patients with different cardiac conditions. The partition of mid-cavity has the best agreement with the ground truth, followed by basal and then apical sections for all groups of patients.

Greening radiology.

PURPOSE: Reducing energy consumption has increased in importance with rising energy prices and funding cutbacks. With the introduction of electronic medical records on the rise in all fields of medicine, there will be a large jump in the number of computers in health care. Radiologist have the unique opportunity, as technological leaders, to direct energy efficiency measures as a means of cost savings and the reduction of airborne by-products from energy production to improve patients' lives. The aim of this study was to assess the many workstations and monitors throughout the authors' department to determine their electrical consumption and cost. METHODS: Equipment was monitored using an electricity meter during both active and standby states. Cost per kilowatt-hour was calculated at $0.11, not including taxes and fees. RESULTS: Any given monitor left on 24/7 would annually consume between 49.5 and 1,399.84 kWh, costing from $5.45 to $153.98. A single workstation left on 24/7 would use 455.65 to 2,358.72 kWh, costing from $59.91 to $259.46. In aggregate, all workstations and monitors would use approximately 137,759.54 kWh, costing $15,153.55. If all equipment were shut down after an 8-hour workday, the department would consume about 32,633.64 kWh, costing $3,589.70 thereby saving 83,866.6 kWh and $9,225.33. Although computers in the remainder of the hospital may use less energy than workstations, this serves as a predictive model for potential energy consumption and cost. CONCLUSIONS: With the increasing necessity of cost savings and energy reduction, this small and simple step, implemented hospital-wide, will lead to much larger cost savings across institutions.

Online social networking: a primer for radiology.

Online social networking is an immature, but rapidly evolving industry of web-based technologies that allow individuals to develop online relationships. News stories populate the headlines about various websites which can facilitate patient and doctor interaction. There remain questions about protecting patient confidentiality and defining etiquette in order to preserve the doctor/patient relationship and protect physicians. How much social networking-based communication or other forms of E-communication is effective? What are the potential benefits and pitfalls of this form of communication? Physicians are exploring how social networking might provide a forum for interacting with their patients, and advance collaborative patient care. Several organizations and institutions have set forth policies to address these questions and more. Though still in its infancy, this form of media has the power to revolutionize the way physicians interact with their patients and fellow health care workers. In the end, physicians must ask what value is added by engaging patients or other health care providers in a social networking format. Social networks may flourish in health care as a means of distributing information to patients or serve mainly as support groups among patients. Physicians must tread a narrow path to bring value to interactions in these networks while limiting their exposure to unwanted liability.