Charlene Liew

Background: Radiology departments frequently manage large, heterogeneous worklists using first-in, first-out (FIFO) reporting workflows. This approach does not account for clinical urgency and may contribute to prolonged reporting delays, particularly in high-volume settings. AI systems are increasingly being integrated into radiology workflows, not only for image analysis but also as tools for worklist prioritization and report generation. However, real-world evidence of their operational impact remains limited. Objective: This study aimed to evaluate the impact of an AI-triaged reading worklist combined with AI-assisted report generation on radiologist workflow efficiency, measured by report generation time (RGT) and overall turnaround time (TAT) for chest radiographs in a real-world hospital setting. Methods: We conducted a single-center prospective paired study using a single-sequence crossover design. Eight board-certified radiologists interpreted chest radiographs during 2 reporting sessions: an unaided session using standard FIFO worklists and an AI-assisted session using an AI-triaged worklist with integrated report generation tools, separated by a 4-week washout period. Chest radiographs acquired between November 2023 and January 2024 were included. RGT was defined as the time from opening a study to report finalization, and TAT was defined as the time from the start of a reporting session to report finalization. Statistical comparisons were performed using nonparametric tests. Results: A total of 1054 chest radiographs were included. Median RGT decreased from 2 (IQR 1-4) minutes in the unaided session to 0.53 (IQR 0.22-1.12) minutes in the AI-assisted session (P<.001), representing a 73.3% reduction. The largest reduction was observed in radiographs categorized as normal, with median RGT decreasing from 2 (IQR 1-3) minutes to 0.2 (IQR 0.13-0.35) minutes (a 90% reduction). Mean TAT decreased from 876.21 (SD 1014.20; 95% CI 816.06-940.55) minutes to 82.25 (SD 83.38; 95% CI 76.98-87.27) minutes with AI assistance, corresponding to a 90.6% reduction. Significant reductions in TAT were observed across all urgency categories, including critical studies (all P<.001). Conclusions: In a real-world clinical setting, the use of AI-triaged worklists and AI-assisted report generation was associated with substantial reductions in RGT and overall TAT for chest radiographs. These findings suggest that AI, when deployed as workflow infrastructure rather than a diagnostic replacement, may meaningfully improve radiology operational efficiency and reporting timeliness.

This commentary discusses the perspectives of two authors working in India and Singapore in medical imaging and highlights differences in their basic structural and regulatory frameworks.

Radiology is increasingly adopting AI systems, and to ensure their safety after the systems go live, postdeployment monitoring is essential as part of the total product life cycle framework.

BACKGROUND Radiology departments frequently manage large, heterogeneous worklists using first-in, first-out (FIFO) reporting workflows. This approach does not account for clinical urgency and may contribute to prolonged reporting delays, particularly in high-volume settings. AI systems are increasingly being integrated into radiology workflows, not only for image analysis but also as tools for worklist prioritization and report generation. However, real-world evidence of their operational impact remains limited. OBJECTIVE This study aimed to evaluate the impact of an AI-triaged reading worklist combined with AI-assisted report generation on radiologist workflow efficiency, measured by report generation time (RGT) and overall turnaround time (TAT) for chest radiographs in a real-world hospital setting. METHODS We conducted a single-center prospective paired study using a single-sequence crossover design. Eight board-certified radiologists interpreted chest radiographs during 2 reporting sessions: an unaided session using standard FIFO worklists and an AI-assisted session using an AI-triaged worklist with integrated report generation tools, separated by a 4-week washout period. Chest radiographs acquired between November 2023 and January 2024 were included. RGT was defined as the time from opening a study to report finalization, and TAT was defined as the time from the start of a reporting session to report finalization. Statistical comparisons were performed using nonparametric tests. RESULTS A total of 1054 chest radiographs were included. Median RGT decreased from 2 (IQR 1-4) minutes in the unaided session to 0.53 (IQR 0.22-1.12) minutes in the AI-assisted session (P<.001), representing a 73.3% reduction. The largest reduction was observed in radiographs categorized as normal, with median RGT decreasing from 2 (IQR 1-3) minutes to 0.2 (IQR 0.13-0.35) minutes (a 90% reduction). Mean TAT decreased from 876.21 (SD 1014.20; 95% CI 816.06-940.55) minutes to 82.25 (SD 83.38; 95% CI 76.98-87.27) minutes with AI assistance, corresponding to a 90.6% reduction. Significant reductions in TAT were observed across all urgency categories, including critical studies (all P<.001). CONCLUSIONS In a real-world clinical setting, the use of AI-triaged worklists and AI-assisted report generation was associated with substantial reductions in RGT and overall TAT for chest radiographs. These findings suggest that AI, when deployed as workflow infrastructure rather than a diagnostic replacement, may meaningfully improve radiology operational efficiency and reporting timeliness.

Amyloidosis is a rare disorder where abnormal protein aggregates are deposited in tissues forming amyloid fibrils, leading to progressive organ failure. Although any organ can be affected, cardiac involvement is the main cause of morbidity and mortality associated with amyloidosis as diagnosis is often delayed due to the indolent nature of the disease in some forms. An early diagnosis of disease and knowledge of the type/subtype of cardiac amyloidosis (CA) are essential for appropriate management and better outcome. Echocardiography is often the first line of investigation for patients suspected of CA and offers superior hemodynamic assessment. Although cardiovascular magnetic resonance (CMR) imaging is not diagnostic of CA, it provides vital clues to diagnosis and has a role in disease quantification and prognostication. Radiolabeled bone seeking tracers are the mainstay of diagnosis of CA and when combined with screening of monoclonal light chains, bone scintigraphy offers high sensitivity in diagnosing transthyretin type of CA. This review aims to describe the noninvasive imaging assessment and approach to diagnosis of patients with suspected CA. Imaging features of echocardiography, nuclear scintigraphy, and CMR are described with a brief mention on computed tomography. Article published online: 16 August 2022 © 2022. World Association of Radiopharmaceutical and Molecular Therapy (WARMTH). This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commercial purposes, or adapted, remixed, transformed or built upon. (https://creativecommons.org/licenses/by-nc-nd/4.0/) Thieme Medical and Scientific Publishers Pvt. Ltd. A-12, 2nd Floor, Sector 2, Noida-201301 UP, India

Since the onset of the COVID-19 pandemic in 2019, many clinical prognostic scoring tools have been proposed or developed to aid clinicians in the disposition and severity assessment of pneumonia. However, there is limited work that focuses on explaining techniques that are best suited for clinicians in their decision making. In this article, we present a new image explainability method named ensemble AI explainability (XAI), which is based on the SHAP and Grad-CAM++ methods. It provides a visual explanation for a deep learning prognostic model that predicts the mortality risk of community-acquired pneumonia and COVID-19 respiratory infected patients. In addition, we surveyed the existing literature and compiled prevailing quantitative and qualitative metrics to systematically review the efficacy of ensemble XAI, and to make comparisons with several state-of-the-art explainability methods (LIME, SHAP, saliency map, Grad-CAM, Grad-CAM++). Our quantitative experimental results have shown that ensemble XAI has a comparable absence impact (decision impact: 0.72, confident impact: 0.24). Our qualitative experiment, in which a panel of three radiologists were involved to evaluate the degree of concordance and trust in the algorithms, has showed that ensemble XAI has localization effectiveness (mean set accordance precision: 0.52, mean set accordance recall: 0.57, mean set ${{\bm{F}}_1}$ : 0.50, mean set IOU: 0.36) and is the most trusted method by the panel of radiologists (mean vote: 70.2%). Finally, the deep learning interpretation dashboard used for the radiologist panel voting will be made available to the community. Our code is available at https://github.com/IHIS-HealthInsights/Interpretation-Methods-Voting-dashboard .

Background: Chest radiography may be used together with deep-learning models to prognosticate COVID-19 patient outcomes Purpose: T o evaluate the performance of a deep-learning model for the prediction of severe patient outcomes from COVID-19 pneumonia on chest radiographs. Methods: A deep-learning model (CAPE: Covid-19 AI Predictive Engine) was trained on 2337 CXR images including 2103 used only for validation while training. The prospective test set consisted of CXR images (n=70) obtained from RT-PCR confirmed COVID-19 pneumonia patients between 1 January and 30 April 2020 in a single center. The radiographs were analyzed by the AI model. Model performance was obtained by receiver operating characteristic curve analysis. Results: In the prospective test set, the mean age of the patients was 46 (+/-16.2) years (84.2% male). The deep-learning model accurately predicted outcomes of ICU admission/mortality from COVID-19 pneumonia with an AUC of 0.79 (95% CI 0.79-0.96). Compared to traditional risk scoring systems for pneumonia based upon laboratory and clinical parameters, the model matched the EWS and MulBTSA risk scoring systems and outperformed CURB-65. Conclusions: A deep-learning model was able to predict severe patient outcomes (ICU admission and mortality) from COVID-19 on chest radiographs. Key Results: A deep-learning model was able to predict severe patient outcomes (ICU admission and mortality) from COVID-19 from chest radiographs with an AUC of 0.79, which is comparable to traditional risk scoring systems for pneumonia. Summary Statement: This is a chest radiography-based AI model to prognosticate the risk of severe COVID-19 pneumonia outcomes.