Overcoming the Challenges in the Development and Implementation of Artificial Intelligence in Radiology:A Comprehensive Review of Solutions Beyond Supervised Learning

Artificial intelligence (AI) in radiology is a rapidly developing field with several prospective clinical studies demonstrating its benefits in clinical practice. In 2022, the Korean Society of Radiology held a forum to discuss the challenges and drawbacks in AI development and implementation. Various barriers hinder the successful application and widespread adoption of AI in radiology, such as limited annotated data, data privacy and security, data heterogeneity, imbalanced data, model interpretability, overfitting, and integration with clinical workflows. In this review, some of the various possible solutions to these challenges are presented and discussed; these include training with longitudinal and multimodal datasets, dense training with multitask learning and multimodal learning, self-supervised contrastive learning, various image modifications and syntheses using generative models, explainable AI, causal learning, federated learning with large data models, and digital twins.

Clinical Impact of a Quality Improvement Program Including Dedicated Emergency Radiology Personnel on Emergency Surgical Management: A Propensity Score-Matching Study

Objective@#To investigate the clinical impact of a quality improvement program including dedicated emergency radiology personnel (QIP-DERP) on the management of emergency surgical patients in the emergency department (ED). @*Materials and Methods@#This retrospective study identified all adult patients (n = 3667) who underwent preoperative body CT, for which written radiology reports were generated, and who subsequently underwent non-elective surgery between 2007 and 2018 in the ED of a single urban academic tertiary medical institution. The study cohort was divided into periods before and after the initiation of QIP-DERP. We matched the control group patients (i.e., before QIP-DERP) to the QIP-DERP group patients using propensity score (PS), with a 1:2 matching ratio for the main analysis and a 1:1 ratio for sub-analyses separately for daytime (8:00 AM to 5:00 PM on weekdays) and after-hours. The primary outcome was timing of emergency surgery (TES), which was defined as the time from ED arrival to surgical intervention. The secondary outcomes included ED length of stay (LOS) and intensive care unit (ICU) admission rate. @*Results@#According to the PS-matched analysis, compared with the control group, QIP-DERP significantly decreased the median TES from 16.7 hours (interquartile range, 9.4–27.5 hours) to 11.6 hours (6.6–21.9 hours) (p < 0.001) and the ICU admission rate from 33.3% (205/616) to 23.9% (295/1232) (p < 0.001). During after-hours, the QIP-DERP significantly reduced median TES from 19.9 hours (12.5–30.1 hours) to 9.6 hours (5.7–19.1 hours) (p < 0.001), median ED LOS from 9.1 hours (5.6–16.5 hours) to 6.7 hours (4.9–11.3 hours) (p < 0.001), and ICU admission rate from 35.5% (108/304) to 22.0% (67/304) (p < 0.001). @*Conclusion@#QIP-DERP implementation improved the quality of emergency surgical management in the ED by reducing TES, ED LOS, and ICU admission rate, particularly during after-hours.

A Rare Case of Granulomatosis with Polyangiitis-Related Periaortitis at the Ascending Aorta

Acute aortic syndrome including intramural hematoma is a life-threatening disease encountered in the emergency department with high in-hospital mortality even when a timely diagnosis is made. Therefore, accurate differential diagnosis of acute aortic syndrome and its mimics is important to determine the patient's treatment plan. Until now, a few pathologic diseases mimicking aortic intramural hematoma such as periaortic lymphoma and immunoglobulin G4-elated periaoritis have been reported. Here, we report a rare case of granulomatosis with polyangiitis-related periaortitis misdiagnosed as aortic intramural hematoma presenting to the emergency department with chest pain and similar radiologic findings.

Complication Following Ultrasound-Guided High-Intensity Focused Ultrasound for the Treatment of Uterine Adenomyosis: Case Report of CT Imaging Features

High intensity focused ultrasound (HIFU) is a non-surgical and non-invasive treatment option in patients with uterine myoma and adenomyosis. As the use of HIFU increases in the clinical practice, it is important to be aware of imaging findings related to ultrasound (US)-guided HIFU ablation and its potential complications. However, there are few reports on the imaging findings regarding complications of US-guided HIFU ablation. Here, we report a case of acute complication after US-guided HIFU ablation, surgically confirmed as thermal injury with necrosis of skin, subcutaneous tissue, anterior abdominal wall muscles, peritoneum and uterus.

Added Value of Bone Suppression Image in the Detection of Subtle Lung Lesions on Chest Radiographs with Regard to Reader's Expertise

BACKGROUND: Chest radiographs (CXR) are the most commonly used imaging techniques by various clinicians and radiologists. However, detecting lung lesions on CXR depends largely on the reader's experience level, so there have been several trials to overcome this problem using post-processing of CXR. We investigated the added value of bone suppression image (BSI) in detecting various subtle lung lesions on CXR with regard to reader's expertise. METHODS: We applied a software program to generate BSI in 1,600 patients in the emergency department. Of them, 80 patients with subtle lung lesions and 80 patients with negative finding on CXR were retrospectively selected based on the subtlety scores on CXR and CT findings. Ten readers independently rated their confidence in deciding the presence or absence of a lung lesion at each of 960 lung regions on the two separated imaging sessions: CXR alone vs. CXR with BSI. RESULTS: The additional use of BSI for all readers significantly increased the mean area under the curve (AUC) in detecting subtle lung lesions (0.663 vs. 0.706; P < 0.001). The less experienced readers were, the more AUC differences increased: 0.067 (P < 0.001) for junior radiology residents; 0.064 (P < 0.001) for non-radiology clinicians; 0.044 (P < 0.001) for senior radiology residents; and 0.019 (P = 0.041) for chest radiologists. The additional use of BSI significantly increased the mean confidence regarding the presence or absence of lung lesions for 213 positive lung regions (2.083 vs. 2.357; P < 0.001) and for 747 negative regions (1.217 vs. 1.195; P = 0.008). CONCLUSION: The use of BSI increases diagnostic performance and confidence, regardless of reader's expertise, reduces the impact of reader's expertise and can be helpful for less experienced clinicians and residents in the detection of subtle lung lesions.

Focal type of peliosis hepatis

No abstract available.

Sonography of hepatic hemangioma accompanied by arterioportal shunt

No abstract available.