Artificial intelligence in diagnostic and interventional radiology: Where are we now?

Artificial intelligence in interventional radiology: Current concepts and future trends

While artificial intelligence (AI) is already well established in diagnostic radiology, it is beginning to make its mark in interventional radiology. AI has the potential to dramatically change the daily practice of interventional radiology at several levels. In the preoperative setting, recent advances in deep learning models, particularly foundation models, enable effective management of multimodality and increased autonomy through their ability to function minimally without supervision. Multimodality is at the heart of patient-tailored management and in interventional radiology, this translates into the development of innovative models for patient selection and outcome prediction. In the perioperative setting, AI is manifesting itself in applications that assist radiologists in image analysis and real-time decision making, thereby improving the efficiency, accuracy, and safety of interventions. In synergy with advances in robotic technologies, AI is laying the groundwork for an increased autonomy. From a research perspective, the development of artificial health data, such as AI-based data augmentation, offers an innovative solution to this central issue and promises to stimulate research in this area. This review aims to provide the medical community with the most important current and future applications of AI in interventional radiology.

Artificial intelligence in emergency neuroradiology: Current applications and perspectives

Emergency neuroradiology provides rapid diagnostic decision-making and guidance for management for a wide range of acute conditions involving the brain, head and neck, and spine. This narrative review aims at providing an up-to-date discussion about the state of the art of applications of artificial intelligence in emergency neuroradiology, which have substantially expanded in depth and scope in the past few years. A detailed analysis of machine learning and deep learning algorithms in several tasks related to acute ischemic stroke involving various imaging modalities, including a description of existing commercial products, is provided. The applications of artificial intelligence in acute intracranial hemorrhage and other vascular pathologies such as intracranial aneurysm and arteriovenous malformation are discussed. Other areas of emergency neuroradiology including infection, fracture, cord compression, and pediatric imaging are further discussed in turn. Based on these discussions, this article offers insight into practical considerations regarding the applications of artificial intelligence in emergency neuroradiology, calling for more development driven by clinical needs, attention to pediatric neuroimaging, and analysis of real-world performance.

Prediction of Local Tumor Progression After Thermal Ablation of Colorectal Cancer Liver Metastases Based on Magnetic Resonance Imaging Δ-Radiomics

PURPOSE

This study aimed to enhance the predictability of local tumor progression (LTP) postthermal ablation in patients with colorectal cancer liver metastases (CRLMs). A sophisticated approach integrating magnetic resonance imaging (MRI) Δ-radiomics and clinical feature-based modeling was employed.

MATERIALS AND METHODS

In this retrospective study, 37 patients with CRLM were included, encompassing a total of 57 tumors. Radiomics features were derived by delineating the images of lesions pretreatment and images of the ablation zones posttreatment. The change in these features, termed Δ-radiomics, was calculated by subtracting preprocedure values from postprocedure values. Three models were developed using the least absolute shrinkage and selection operators (LASSO) and logistic regression: the preoperative lesion model, the postoperative ablation area model, and the Δ model. Additionally, a composite model incorporating identified clinical features predictive of early treatment success was created to assess its prognostic utility for LTP.

RESULTS

LTP was observed in 20 out of the 57 lesions (35%). The clinical model identified, tumor size (P = 0.010), and ΔCEA (P = 0.044) as factors significantly associated with increased LTP risk postsurgery. Among the three models, the Δ model demonstrated the highest AUC value (T2WI AUC in training, 0.856; Delay AUC, 0.909; T2WI AUC in testing, 0.812; Delay AUC, 0.875), whereas the combined model yielded optimal performance (T2WI AUC in training, 0.911; Delay AUC, 0.954; T2WI AUC in testing, 0.847; Delay AUC, 0.917). Despite its superior AUC values, no significant difference was noted when comparing the performance of the combined model across the two sequences (P = 0.6087).

CONCLUSIONS

Combined models incorporating clinical data and Δ-radiomics features serve as valuable indicators for predicting LTP following thermal ablation in patients with CRLM.

The Role and Future of Artificial Intelligence in Robotic Image-Guided Interventions

Artificial intelligence and robotics are transforming interventional radiology, driven by advancements in computer vision, robotics and procedural automation. Historically focused on diagnostics, AI now also enhances procedural capabilities in IR, enabling future robotic systems to handle complex tasks such as catheter manipulation or needle placement with increasing precision and reliability. Early robotic systems in IR demonstrated improved accuracy in both vascular and percutaneous interventions, though none were equipped with automatic decision-making. This review tends to show the potential in improving procedural outcomes with AI for robotics, though challenges remain. Techniques like reinforcement learning and haptic vision are under investigation to address several issues, training robots to adapt based on real-time feedback from the environment. As AI-driven robotics evolve, IR could shift towards a model where human expertise oversees the technology rather than performs the intervention itself.

Potential strength and weakness of artificial intelligence integration in emergency radiology: a review of diagnostic utilizations and applications in patient care optimization

Artificial intelligence (AI) and its recent increasing healthcare integration has created both new opportunities and challenges in the practice of radiology and medical imaging. Recent advancements in AI technology have allowed for more workplace efficiency, higher diagnostic accuracy, and overall improvements in patient care. Limitations of AI such as data imbalances, the unclear nature of AI algorithms, and the challenges in detecting certain diseases make it difficult for its widespread adoption. This review article presents cases involving the use of AI models to diagnose intracranial hemorrhage, spinal fractures, and rib fractures, while discussing how certain factors like, type, location, size, presence of artifacts, calcification, and post-surgical changes, affect AI model performance and accuracy. While the use of artificial intelligence has the potential to improve the practice of emergency radiology, it is important to address its limitations to maximize its advantages while ensuring the safety of patients overall.

Edge Computing for AI-Based Brain MRI Applications: A Critical Evaluation of Real-Time Classification and Segmentation

Medical imaging plays a pivotal role in diagnostic medicine with technologies like Magnetic Resonance Imagining (MRI), Computed Tomography (CT), Positron Emission Tomography (PET), and ultrasound scans being widely used to assist radiologists and medical experts in reaching concrete diagnosis. Given the recent massive uplift in the storage and processing capabilities of computers, and the publicly available big data, Artificial Intelligence (AI) has also started contributing to improving diagnostic radiology. Edge computing devices and handheld gadgets can serve as useful tools to process medical data in remote areas with limited network and computational resources. In this research, the capabilities of multiple platforms are evaluated for the real-time deployment of diagnostic tools. MRI classification and segmentation applications developed in previous studies are used for testing the performance using different hardware and software configurations. Cost-benefit analysis is carried out using a workstation with a NVIDIA Graphics Processing Unit (GPU), Jetson Xavier NX, Raspberry Pi 4B, and Android phone, using MATLAB, Python, and Android Studio. The mean computational times for the classification app on the PC, Jetson Xavier NX, and Raspberry Pi are 1.2074, 3.7627, and 3.4747 s, respectively. On the low-cost Android phone, this time is observed to be 0.1068 s using the Dynamic Range Quantized TFLite version of the baseline model, with slight degradation in accuracy. For the segmentation app, the times are 1.8241, 5.2641, 6.2162, and 3.2023 s, respectively, when using JPEG inputs. The Jetson Xavier NX and Android phone stand out as the best platforms due to their compact size, fast inference times, and affordability.

Climate change and artificial intelligence in healthcare: Review and recommendations towards a sustainable future

The rapid advancement of artificial intelligence (AI) in healthcare has revolutionized the industry, offering significant improvements in diagnostic accuracy, efficiency, and patient outcomes. However, the increasing adoption of AI systems also raises concerns about their environmental impact, particularly in the context of climate change. This review explores the intersection of climate change and AI in healthcare, examining the challenges posed by the energy consumption and carbon footprint of AI systems, as well as the potential solutions to mitigate their environmental impact. The review highlights the energy-intensive nature of AI model training and deployment, the contribution of data centers to greenhouse gas emissions, and the generation of electronic waste. To address these challenges, the development of energy-efficient AI models, the adoption of green computing practices, and the integration of renewable energy sources are discussed as potential solutions. The review also emphasizes the role of AI in optimizing healthcare workflows, reducing resource waste, and facilitating sustainable practices such as telemedicine. Furthermore, the importance of policy and governance frameworks, global initiatives, and collaborative efforts in promoting sustainable AI practices in healthcare is explored. The review concludes by outlining best practices for sustainable AI deployment, including eco-design, lifecycle assessment, responsible data management, and continuous monitoring and improvement. As the healthcare industry continues to embrace AI technologies, prioritizing sustainability and environmental responsibility is crucial to ensure that the benefits of AI are realized while actively contributing to the preservation of our planet.

Telehealth development in the WHO European region: Results from a quantitative survey and insights from Norway

BACKGROUND

The COVID-19 pandemic sent shock waves through societies, economies, and health systems of Member States in the WHO European Region and beyond. During the pandemic, most countries transitioned from a slow to a rapid adoption of telehealth solutions, to accommodate the public health and social measures introduced to mitigate the spread of the disease. As countries shift to a post-pandemic world, the question remains whether telehealth's importance as a mode of care provision in Europe continues to be significant.

OBJECTIVE

This paper aims to present, synthesize, and interpret results from the Telehealth Programmes section of the 2022 WHO Survey on Digital Health (2022 WHO/Europe DH Survey). We specifically analyze the implementation and use of teleradiology, telemedicine, and telepsychiatry. Norwegian telehealth experiences will be used to illustrate survey findings, and we discuss some of the relevant barriers and facilitators that impact the use of telehealth services.

METHODS

The survey tool was revised from the 2015 WHO Global Survey on eHealth, updated to reflect recent progress and policy priorities.The 2022 WHO/Europe DH Survey was conducted by WHO and circulated to Member States in its European Region from April to October 2022.

RESULTS

The data analysis revealed that teleradiology, telemedicine, and telepsychiatry are the telehealth services most commonly used in the WHO European Region in 2022. Funding remains the most significant barrier to the implementation of telehealth in the Region, followed by infrastructure and capacity/human resources. The survey results highlighted in this study are presented in the following sections: (1) telehealth strategies and financing, (2) telehealth programmes and services offered by Member States of the WHO European Region, (3) barriers to implementing telehealth services, and (4) monitoring and evaluation of telehealth.

CONCLUSION

Based on WHO's 2022 survey, the use of telehealth in the WHO European Region is on the rise. However, merely having telehealth in place is not sufficient for its successful and sustained use for care provision. Responses also uncovered regional differences and barriers that need to be overcome. Successful implementation and scaling of telehealth requires rethinking the design of health and social care systems to create robust, trustworthy, and person-centred digital health and care services.

Evaluation of a deep learning-based software to automatically detect and quantify breast arterial calcifications on digital mammogram

PURPOSE

The purpose of this study was to evaluate an artificial intelligence (AI) software that automatically detects and quantifies breast arterial calcifications (BAC).

MATERIALS AND METHODS

Women who underwent both mammography and thoracic computed tomography (CT) from 2009 to 2018 were retrospectively included in this single-center study. Deep learning-based software was used to automatically detect and quantify BAC with a BAC AI score ranging from 0 to 10-points. Results were compared using Spearman correlation test with a previously described BAC manual score based on radiologists' visual quantification of BAC on the mammogram. Coronary artery calcification (CAC) score was manually scored using a 12-point scale on CT. The diagnostic performance of the marked BAC AI score (defined as BAC AI score ≥ 5) for the detection of marked CAC (CAC score ≥ 4) was analyzed in terms of sensitivity, specificity, accuracy and area under the receiver operating characteristic curve (AUC).

RESULTS

A total of 502 women with a median age of 62 years (age range: 42-96 years) were included. The BAC AI score showed a very strong correlation with the BAC manual score (r = 0.83). Marked BAC AI score had 32.7 % sensitivity (37/113; 95 % confidence interval [CI]: 24.2-42.2), 96.1 % specificity (374/389; 95 % CI: 93.7-97.8), 71.2 % positive predictive value (37/52; 95 % CI: 56.9-82.9), 83.1 % negative predictive value (374/450; 95 % CI: 79.3-86.5), and 81.9 % accuracy (411/502; 95 % CI: 78.2-85.1) for the diagnosis of marked CAC. The AUC of the marked BAC AI score for the diagnosis of marked CAC was 0.64 (95 % CI: 0.60-0.69).

CONCLUSION

The automated BAC AI score shows a very strong correlation with manual BAC scoring in this external validation cohort. The automated BAC AI score may be a useful tool to promote the integration of BAC into mammography reports and to improve awareness of a woman's cardiovascular risk status.

Paranoid About Androids: A Review of Robotics in Radiology

In tandem with the ever-increasing global population, the demand for diagnostic radiology service provision is on the rise and at a disproportionate rate compared to the number of radiologists available to practice. The current "revolution in robotics" promises to alleviate personnel shortages in many sectors of industry, including medicine. Despite negative depictions of robots in popular culture, their multiple potential benefits cannot be overlooked, in particular when it comes to health service provision. The type of robots used for interventional procedures are largely robotic-assistance devices, such as the Da Vinci surgical robot. Advances have also been made with regards to robots for image-guided percutaneous needle placement, which have demonstrated superior accuracy compared to manual methods. It is likely that artificial intelligence will come to play a key role in the field of robotics and will result in an increase in the levels of robotic autonomy attainable. However, this concept is not without ethical and legal considerations, most notably who is responsible should an error occur; the physician, the robot manufacturer, software engineers, or the robot itself? Efforts have been made to legislate in order to protect against the potentially harmful effects of unexplainable "black-box" decision outputs of artificial intelligence systems. In order to be accepted by patients, studies have shown that the perceived level of trustworthiness and predictability of robots is crucial. Ultimately, effective, widespread implementation of medical robotic systems will be contingent on developers remaining cognizant of factors that increase human acceptance, as well as ensuring compliance with regulations.

Coronary artery disease detection using deep learning and ultrahigh-resolution photon-counting coronary CT angiography

PURPOSE

The purpose of this study was to evaluate the diagnostic performance of automated deep learning in the detection of coronary artery disease (CAD) on photon-counting coronary CT angiography (PC-CCTA).

MATERIALS AND METHODS

Consecutive patients with suspected CAD who underwent PC-CCTA between January 2022 and December 2023 were included in this retrospective, single-center study. Non-ultra-high resolution (UHR) PC-CCTA images were analyzed by artificial intelligence using two deep learning models (CorEx, Spimed-AI), and compared to human expert reader assessment using UHR PC-CCTA images. Diagnostic performance for global CAD assessment (at least one significant stenosis ≥ 50 %) was estimated at patient and vessel levels.

RESULTS

A total of 140 patients (96 men, 44 women) with a median age of 60 years (first quartile, 51; third quartile, 68) were evaluated. Significant CAD on UHR PC-CCTA was present in 36/140 patients (25.7 %). The sensitivity, specificity, accuracy, positive predictive value), and negative predictive value of deep learning-based CAD were 97.2 %, 81.7 %, 85.7 %, 64.8 %, and 98.9 %, respectively, at the patient level and 96.6 %, 86.7 %, 88.1 %, 53.8 %, and 99.4 %, respectively, at the vessel level. The area under the receiver operating characteristic curve was 0.90 (95 % CI: 0.83-0.94) at the patient level and 0.92 (95 % CI: 0.89-0.94) at the vessel level.

CONCLUSION

Automated deep learning shows remarkable performance for the diagnosis of significant CAD on non-UHR PC-CCTA images. AI pre-reading may be of supportive value to the human reader in daily clinical practice to target and validate coronary artery stenosis using UHR PC-CCTA.

Intra-arterial administration of PSMA-targeted radiopharmaceuticals for brain tumors: is the era of interventional theranostics next?

In recent years, prostate-specific membrane antigen (PSMA), a transmembrane glycoprotein, has emerged as a promising biomarker for theranostics, integrating diagnosis and therapy. PSMA's overexpression in various tumors, including brain metastases and high-grade gliomas, suggests its potential in neuro-oncology. Pruis et al. conducted a proof-of-concept study comparing intra-arterial (IA) and intravenous (IV) administration of 68Ga-PSMA-11 in brain tumor patients, aiming to enhance radioligand therapy (RLT) outcomes. Ten patients underwent IV and super-selective IA (ssIA) tracer administration, showing higher tumor uptake and more favorable biodistribution after ssIA administration on positron emission tomography (PET). Dosimetry modeling on the basis of PET data resulted in median absorbed radiation doses per tumor per cycle notably higher with ssIA with respect to IV administration, indicating its potential for RLT optimization. Challenges persist, notably in penetrating intact blood-brain barriers and targeting tumor cells effectively. To overcome these limitations, novel approaches like convection-enhanced delivery and focused ultrasound warrant exploration. Safety concerns, though minimal in this study, underscore the need for larger trials and AI-assisted procedures. PSMA's role in neuro-oncological theranostics is promising, but future research must address specificity and compare it with emerging targets.

Clinical value of sequential circulating tumor DNA analysis using next-generation sequencing and epigenetic modifications for guiding thermal ablation for colorectal cancer metastases: a prospective study

INTRODUCTION

While thermal ablation is now a standard treatment option for oligometastatic colorectal cancer patients, selecting those who will benefit most from locoregional therapies remains challenging. This proof-of-concept study is the first to assess the feasibility of routine testing of ctDNA before and after thermal ablation with curative intent, analyzed by next-generation sequencing (NGS) and methylation specific digital droplet PCR (ddPCR). Our prospective study primary objective was to assess the prognostic value of ctDNA before thermal ablation.

METHODS

This single-center prospective study from November 2021 to June 2022 included colorectal cancer patients referred for curative-intent thermal ablation. Cell-free DNA was tested at different time points by next-generation sequencing and detection of WIF1 and NPY genes hypermethylation using ddPCR. The ctDNA was considered positive if either a tumor mutation or hypermethylation was detected; recurrence-free survival was used as the primary endpoint.

RESULTS

The study enrolled 15 patients, and a total of 60 samples were analyzed. The median follow-up after ablation was 316 days, and median recurrence-free survival was 250 days. CtDNA was positive for 33% of the samples collected during the first 24 h. The hazard ratio for progression according to the presence of baseline circulating tumor DNA was estimated at 0.14 (CI 95%: 0.03-0.65, p = 0.019). The dynamics are provided, and patients with no recurrence were all negative at H24 for ctDNA.

DISCUSSION

This study shows the feasibility of routine testing of ctDNA before and after thermal ablation with curative intent. We report that circulating tumor DNA is detectable in patients with low tumor burden using 2 techniques. This study emphasizes the potential of ctDNA for discerning patients who are likely to benefit from thermal ablation from those who may not, which could shape future referrals. The dynamics of ctDNA before and after ablation shed light on the need for further research and larger studies.

Toward Precision Diagnosis: Machine Learning in Identifying Malignant Orbital Tumors With Multiparametric 3 T MRI

BACKGROUND

Orbital tumors present a diagnostic challenge due to their varied locations and histopathological differences. Although recent advancements in imaging have improved diagnosis, classification remains a challenge. The integration of artificial intelligence in radiology and ophthalmology has demonstrated promising outcomes.

PURPOSE

This study aimed to evaluate the performance of machine learning models in accurately distinguishing malignant orbital tumors from benign ones using multiparametric 3 T magnetic resonance imaging (MRI) data.

MATERIALS AND METHODS

In this single-center prospective study, patients with orbital masses underwent presurgery 3 T MRI scans between December 2015 and May 2021. The MRI protocol comprised multiparametric imaging including dynamic contrast-enhanced (DCE), diffusion-weighted imaging (DWI), intravoxel incoherent motion (IVIM), as well as morphological imaging acquisitions. A repeated nested cross-validation strategy using random forest classifiers was used for model training and evaluation, considering 8 combinations of explanatory features. Shapley additive explanations (SHAP) values were used to assess feature contributions, and the model performance was evaluated using multiple metrics.

RESULTS

One hundred thirteen patients were analyzed (57/113 [50.4%] were women; average age was 51.5 ± 17.5 years, range: 19-88 years). Among the 8 combinations of explanatory features assessed, the performance on predicting malignancy when using the most comprehensive model, which is the most exhaustive one incorporating all 46 explanatory features-including morphology, DWI, DCE, and IVIM, achieved an area under the curve of 0.9 [0.73-0.99]. When using the streamlined "10-feature signature" model, performance reached an area under the curve of 0.88 [0.71-0.99]. Random forest feature importance graphs measured by the mean of SHAP values pinpointed the 10 most impactful features, which comprised 3 quantitative IVIM features, 4 quantitative DCE features, 1 quantitative DWI feature, 1 qualitative DWI feature, and age.

CONCLUSIONS

Our findings demonstrate that a machine learning approach, integrating multiparametric MRI data such as DCE, DWI, IVIM, and morphological imaging, offers high-performing models for differentiating malignant from benign orbital tumors. The streamlined 10-feature signature, with a performance close to the comprehensive model, may be more suitable for clinical application.

Computer-assisted urine cytology: Faster, cheaper, better?

Recent advancements in computer-assisted diagnosis (CAD) have catalysed significant progress in pathology, particularly in the realm of urine cytopathology. This review synthesizes the latest developments and challenges in CAD for diagnosing urothelial carcinomas, addressing the limitations of traditional urinary cytology. Through a literature review, we identify and analyse CAD models and algorithms developed for urine cytopathology, highlighting their methodologies and performance metrics. We discuss the potential of CAD to improve diagnostic accuracy, efficiency and patient outcomes, emphasizing its role in streamlining workflow and reducing errors. Furthermore, CAD tools have shown potential in exploring pathological conditions, uncovering novel biomarkers and prognostic/predictive features previously unknown or unseen. Finally, we examine the practical issues surrounding the integration of CAD into clinical practice, including regulatory approval, validation and training for pathologists. Despite the promising results, challenges remain, necessitating further research and validation efforts. Overall, CAD presents a transformative opportunity to revolutionize diagnostic practices in urine cytopathology, paving the way for enhanced patient care and outcomes.

Molecular imaging for non-invasive risk stratification of renal masses

Anatomic imaging with contrast-enhanced computed tomography (CT) and magnetic resonance imaging (MRI) has long been the mainstay of renal mass characterization. However, those modalities are often unable to adequately characterize indeterminate, solid, enhancing renal masses - with some exceptions, such as the development of the clear-cell likelihood score on multi-parametric MRI. As such, molecular imaging approaches have gained traction as an alternative to anatomic imaging. Mitochondrial imaging with 99mTc-sestamibi single-photon emission computed tomography/CT is a cost-effective means of non-invasively identifying oncocytomas and other indolent renal masses. On the other end of the spectrum, carbonic anhydrase IX agents, most notably the monoclonal antibody girentuximab - which can be labeled with positron emission tomography radionuclides such as zirconium-89 - are effective at identifying renal masses that are likely to be aggressive clear cell renal cell carcinomas. Renal mass biopsy, which has a relatively high non-diagnostic rate and does not definitively characterize many oncocytic neoplasms, nonetheless may play an important role in any algorithm targeted to renal mass risk stratification. The combination of molecular imaging and biopsy in selected patients with other advanced imaging methods, such as artificial intelligence/machine learning and the abstraction of radiomics features, offers the optimal way forward for maximization of the information to be gained from risk stratification of indeterminate renal masses. With the proper application of those methods, inappropriately aggressive therapy for benign and indolent renal masses may be curtailed.

Emerging technologies in pediatric radiology: current developments and future prospects

Radiological imaging is a crucial diagnostic tool for the pediatric population. However, it is associated with several unique challenges in this age group compared to adults. These challenges mainly come from the fact that children are not small-sized adults and differ in development, anatomy, physiology, and pathology compared to adults. This paper reviews relevant articles published between January 2015 and October 2023 to analyze challenges associated with imaging technologies currently used in pediatric radiology, emerging technologies, and their role in resolving the challenges and future prospects of pediatric radiology. In recent decades, imaging technologies have advanced rapidly, developing advanced ultrasound, computed tomography, magnetic resonance, nuclear imaging, teleradiology, artificial intelligence, machine learning, three-dimensional printing, radiomics, and radiogenomics, among many others. By prioritizing the unique needs of pediatric patients while developing such technologies, we can significantly alleviate the challenges faced in pediatric radiology.

An Introductory Guide to Artificial Intelligence in Interventional Radiology: Part 1 Foundational Knowledge

Artificial intelligence (AI) is rapidly evolving and has transformative potential for interventional radiology (IR) clinical practice. However, formal training in AI may be limited for many clinicians and therefore presents a challenge for initial implementation and trust in AI. An understanding of the foundational concepts in AI may help familiarize the interventional radiologist with the field of AI, thus facilitating understanding and participation in the development and deployment of AI. A pragmatic classification system of AI based on the complexity of the model may guide clinicians in the assessment of AI. Finally, the current state of AI in IR and the patterns of implementation are explored (pre-procedural, intra-procedural, and post-procedural).

Deep Learning for Pneumothorax Detection on Chest Radiograph: A Diagnostic Test Accuracy Systematic Review and Meta Analysis

BACKGROUND

Pneumothorax is a common acute presentation in healthcare settings. A chest radiograph (CXR) is often necessary to make the diagnosis, and minimizing the time between presentation and diagnosis is critical to deliver optimal treatment. Deep learning (DL) algorithms have been developed to rapidly identify pathologic findings on various imaging modalities.

PURPOSE

The purpose of this systematic review and meta-analysis was to evaluate the overall performance of studies utilizing DL algorithms to detect pneumothorax on CXR.

METHODS

A study protocol was created and registered a priori (PROSPERO CRD42023391375). The search strategy included studies published up until January 10, 2023. Inclusion criteria were studies that used adult patients, utilized computer-aided detection of pneumothorax on CXR, dataset was evaluated by a qualified physician, and sufficient data was present to create a 2 × 2 contingency table. Risk of bias was assessed using the QUADAS-2 tool. Bivariate random effects meta-analyses and meta-regression modeling were performed.

RESULTS

Twenty-three studies were selected, including 34 011 patients and 34 075 CXRs. The pooled sensitivity and specificity were 87% (95% confidence interval, 81%, 92%) and 95% (95% confidence interval, 92%, 97%), respectively. The study design, use of an institutional/public data set and risk of bias had no significant effect on the sensitivity and specificity of pneumothorax detection.

CONCLUSIONS

The relatively high sensitivity and specificity of pneumothorax detection by deep-learning showcases the vast potential for implementation in clinical settings to both augment the workflow of radiologists and assist in more rapid diagnoses and subsequent patient treatment.

Radiologic Technology Students' Perceptions on Adoption of Artificial Intelligence Technology in Radiology

Study Purpose

This study aims to analyze radiologic technology student's perceptions of artificial intelligence (AI) and its applications in radiology.

Methods

A quantitative cross-sectional survey was conducted. A pre-validated survey questionnaire with 17 items related to students perceptions of AI and its applications was used. The sample included radiologic technology students from three universities in Saudi Arabia. The survey was conducted online for several weeks, resulting in a sample of 280 radiologic technology students.

Results

Of the participants, 63.9% were aware of AI and its applications. T-tests revealed a statistically significant difference (p = 0.0471) between genders with male participants reflecting slightly higher AI awareness than female participants. Regarding the choice of radiology as specialization, 35% of the participants stated that they would not choose radiology, whereas 65% preferred it. Approximately 56% of the participants expressed concerns about the potential replacement of radiology technologists with AI, and 62.1% strongly agreed on the necessity of incorporating known ethical principles into AI.

Conclusion

The findings reflect a positive evaluation of the applications of this technology, which is attributed to its essential support role. However, tailored education and training programs are necessary to prepare future healthcare professionals for the increasing role of AI in medical sciences.

Cinematic Rendering of Gastrointestinal Stromal Tumours: A Review of Current Possibilities and Future Developments

Gastrointestinal stromal tumours (GISTs) are defined as CD117-positive primary, spindled or epithelioid, mesenchymal tumours of the gastrointestinal tract, omentum, or mesentery. While computed tomography (CT) is the recommended imaging modality for GISTs, overlap in imaging features between GISTs and other gastrointestinal tumours often make radiological diagnosis and subsequent selection of the optimal therapeutic approach challenging. Cinematic rendering is a novel CT post-processing technique that generates highly photorealistic anatomic images based on a unique lighting model. The global lighting model produces high degrees of surface detail and shadowing effects that generate depth in the final three-dimensional display. Early studies have shown that cinematic rendering produces high-quality images with enhanced detail by comparison with other three-dimensional visualization techniques. Cinematic rendering shows promise in improving the visualization of enhancement patterns and internal architecture of abdominal lesions, local tumour extension, and global disease burden, which may be helpful for lesion characterization and pretreatment planning. This article discusses and illustrates the application of cinematic rendering in the evaluation of GISTs and the unique benefit of using cinematic rendering in the workup of GIST with a specific emphasis on tumour characterization and preoperative planning.

Assessing the Impact of AI Education on Hispanic Healthcare Professionals' Perceptions and Knowledge

This study investigates the awareness and perceptions of artificial intelligence (AI) among Hispanic healthcare-related professionals, focusing on integrating AI in healthcare. The study participants were recruited from an asynchronous course offered twice within a year at the University of Puerto Rico Medical Science Campus, titled "Artificial Intelligence and Machine Learning Applied to Health Disparities Research", which aimed to bridge the gaps in AI knowledge among participants. The participants were divided into Experimental (n = 32; data-illiterate) and Control (n = 18; data-literate) groups, and pre-test and post-test surveys were administered to assess knowledge and attitudes toward AI. Descriptive statistics, power analysis, and the Mann-Whitney U test were employed to determine the influence of the course on participants' comprehension and perspectives regarding AI. Results indicate significant improvements in knowledge and attitudes among participants, emphasizing the effectiveness of the course in enhancing understanding and fostering positive attitudes toward AI. Findings also reveal limited practical exposure to AI applications, highlighting the need for improved integration into education. This research highlights the significance of educating healthcare professionals about AI to enable its advantageous incorporation into healthcare procedures. The study provides valuable perspectives from a broad spectrum of healthcare workers, serving as a basis for future investigations and educational endeavors aimed at AI implementation in healthcare.

Assessing radiologists' and radiographers' perceptions on artificial intelligence integration: opportunities and challenges

OBJECTIVES

The objective of this study was to evaluate radiologists' and radiographers' opinions and perspectives on artificial intelligence (AI) and its integration into the radiology department. Additionally, we investigated the most common challenges and barriers that radiologists and radiographers face when learning about AI.

METHODS

A nationwide, online descriptive cross-sectional survey was distributed to radiologists and radiographers working in hospitals and medical centres from May 29, 2023 to July 30, 2023. The questionnaire examined the participants' opinions, feelings, and predictions regarding AI and its applications in the radiology department. Descriptive statistics were used to report the participants' demographics and responses. Five-points Likert-scale data were reported using divergent stacked bar graphs to highlight any central tendencies.

RESULTS

Responses were collected from 258 participants, revealing a positive attitude towards implementing AI. Both radiologists and radiographers predicted breast imaging would be the subspecialty most impacted by the AI revolution. MRI, mammography, and CT were identified as the primary modalities with significant importance in the field of AI application. The major barrier encountered by radiologists and radiographers when learning about AI was the lack of mentorship, guidance, and support from experts.

CONCLUSION

Participants demonstrated a positive attitude towards learning about AI and implementing it in the radiology practice. However, radiologists and radiographers encounter several barriers when learning about AI, such as the absence of experienced professionals support and direction.

ADVANCES IN KNOWLEDGE

Radiologists and radiographers reported several barriers to AI learning, with the most significant being the lack of mentorship and guidance from experts, followed by the lack of funding and investment in new technologies.

CT and MRI of abdominal cancers: current trends and perspectives in the era of radiomics and artificial intelligence

Abdominal cancers continue to pose daily challenges to clinicians, radiologists and researchers. These challenges are faced at each stage of abdominal cancer management, including early detection, accurate characterization, precise assessment of tumor spread, preoperative planning when surgery is anticipated, prediction of tumor aggressiveness, response to therapy, and detection of recurrence. Technical advances in medical imaging, often in combination with imaging biomarkers, show great promise in addressing such challenges. Information extracted from imaging datasets owing to the application of radiomics can be used to further improve the diagnostic capabilities of imaging. However, the analysis of the huge amount of data provided by these advances is a difficult task in daily practice. Artificial intelligence has the potential to help radiologists in all these challenges. Notably, the applications of AI in the field of abdominal cancers are expanding and now include diverse approaches for cancer detection, diagnosis and classification, genomics and detection of genetic alterations, analysis of tumor microenvironment, identification of predictive biomarkers and follow-up. However, AI currently has some limitations that need further refinement for implementation in the clinical setting. This review article sums up recent advances in imaging of abdominal cancers in the field of image/data acquisition, tumor detection, tumor characterization, prognosis, and treatment response evaluation.

Detection and severity quantification of pulmonary embolism with 3D CT data using an automated deep learning-based artificial solution

PURPOSE

The purpose of this study was to propose a deep learning-based approach to detect pulmonary embolism and quantify its severity using the Qanadli score and the right-to-left ventricle diameter (RV/LV) ratio on three-dimensional (3D) computed tomography pulmonary angiography (CTPA) examinations with limited annotations.

MATERIALS AND METHODS

Using a database of 3D CTPA examinations of 1268 patients with image-level annotations, and two other public datasets of CTPA examinations from 91 (CAD-PE) and 35 (FUME-PE) patients with pixel-level annotations, a pipeline consisting of: (i), detecting blood clots; (ii), performing PE-positive versus negative classification; (iii), estimating the Qanadli score; and (iv), predicting RV/LV diameter ratio was followed. The method was evaluated on a test set including 378 patients. The performance of PE classification and severity quantification was quantitatively assessed using an area under the curve (AUC) analysis for PE classification and a coefficient of determination (R²) for the Qanadli score and the RV/LV diameter ratio.

RESULTS

Quantitative evaluation led to an overall AUC of 0.870 (95% confidence interval [CI]: 0.850-0.900) for PE classification task on the training set and an AUC of 0.852 (95% CI: 0.810-0.890) on the test set. Regression analysis yielded R² value of 0.717 (95% CI: 0.668-0.760) and of 0.723 (95% CI: 0.668-0.766) for the Qanadli score and the RV/LV diameter ratio estimation, respectively on the test set.

CONCLUSION

This study shows the feasibility of utilizing AI-based assistance tools in detecting blood clots and estimating PE severity scores with 3D CTPA examinations. This is achieved by leveraging blood clots and cardiac segmentations. Further studies are needed to assess the effectiveness of these tools in clinical practice.

External validation of an artificial intelligence solution for the detection of elbow fractures and joint effusions in children

PURPOSE

The purpose of this study was to conduct an external validation of an artificial intelligence (AI) solution for the detection of elbow fractures and joint effusions using radiographs from a real-life cohort of children.

MATERIALS AND METHODS

This single-center retrospective study was conducted on 758 radiographic sets (1637 images) obtained from consecutive emergency room visits of 712 children (mean age, 7.27 ± 3.97 [standard deviation] years; age range, 7 months and 10 days to 15 years and 10 months), referred for a trauma of the elbow. For each set, fracture and/or effusion detection by eleven senior radiologists (reference standard) and AI solution was recorded. Diagnostic performance of the AI solution was measured via four different approaches: fracture detection (presence/absence of fracture as binary variable), fracture enumeration, fracture localization and lesion detection (fracture and/or a joint effusion used as constructed binary variable).

RESULTS

The sensitivity of the AI solution for each of the four approaches was >89%. Greatest sensitivity of the AI solution was obtained for lesion detection (95.0%; 95% confidence interval: 92.1-96.9). The specificity of the AI solution ranged between 63% (for lesion detection) and 77% (for fracture detection). For all four approaches, the negative predictive values were >92% and the positive predictive values ranged between 54% (for fracture enumeration and localization) and 73% (for lesion detection). Specificity was lower for plastered children for all approaches (P < 0.001).

CONCLUSION

The AI solution demonstrates high performances for detecting elbow's fracture and/or joint effusion in children. However, in our context of use, 8% of the radiographic sets ruled-out by the algorithm concerned children with a genuine traumatic elbow lesion.

Exploring the potential of ChatGPT as an adjunct for generating diagnosis based on chief complaint and cone beam CT radiologic findings

AIM

This study aimed to assess the performance of OpenAI's ChatGPT in generating diagnosis based on chief complaint and cone beam computed tomography (CBCT) radiologic findings.

MATERIALS AND METHODS

102 CBCT reports (48 with dental diseases (DD) and 54 with neoplastic/cystic diseases (N/CD)) were collected. ChatGPT was provided with chief complaint and CBCT radiologic findings. Diagnostic outputs from ChatGPT were scored based on five-point Likert scale. For diagnosis accuracy, the scoring was based on the accuracy of chief complaint related diagnosis and chief complaint unrelated diagnoses (1-5 points); for diagnosis completeness, the scoring was based on how many accurate diagnoses included in ChatGPT's output for one case (1-5 points); for text quality, the scoring was based on how many text errors included in ChatGPT's output for one case (1-5 points). For 54 N/CD cases, the consistence of the diagnosis generated by ChatGPT with pathological diagnosis was also calculated. The constitution of text errors in ChatGPT's outputs was evaluated.

RESULTS

After subjective ratings by expert reviewers on a five-point Likert scale, the final score of diagnosis accuracy, diagnosis completeness and text quality of ChatGPT was 3.7, 4.5 and 4.6 for the 102 cases. For diagnostic accuracy, it performed significantly better on N/CD (3.8/5) compared to DD (3.6/5). For 54 N/CD cases, 21(38.9%) cases have first diagnosis completely consistent with pathological diagnosis. No text errors were observed in 88.7% of all the 390 text items.

CONCLUSION

ChatGPT showed potential in generating radiographic diagnosis based on chief complaint and radiologic findings. However, the performance of ChatGPT varied with task complexity, necessitating professional oversight due to a certain error rate.

Radiation reduction for interventional radiology imaging: a video frame interpolation solution

PURPOSE

The aim of this study was to diminish radiation exposure in interventional radiology (IR) imaging while maintaining image quality. This was achieved by decreasing the acquisition frame rate and employing a deep neural network to interpolate the reduced frames.

METHODS

This retrospective study involved the analysis of 1634 IR sequences from 167 pediatric patients (March 2014 to January 2022). The dataset underwent a random split into training and validation subsets (at a 9:1 ratio) for model training and evaluation. Our approach proficiently synthesized absent frames in simulated low-frame-rate sequences by excluding intermediate frames from the validation subset. Accuracy assessments encompassed both objective experiments and subjective evaluations conducted by nine radiologists.

RESULTS

The deep learning model adeptly interpolated the eliminated frames within IR sequences, demonstrating encouraging peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM) results. The average PSNR values for angiographic, subtraction, and fluoroscopic modes were 44.94 dB, 34.84 dB, and 33.82 dB, respectively, while the corresponding SSIM values were 0.9840, 0.9194, and 0.7752. Subjective experiments conducted with experienced interventional radiologists revealed minimal discernible differences between interpolated and authentic sequences.

CONCLUSION

Our method, which interpolates low-frame-rate IR sequences, has shown the capability to produce high-quality IR images. Additionally, the model exhibits potential for reducing the frame rate during IR image acquisition, consequently mitigating radiation exposure.

CRITICAL RELEVANCE STATEMENT

This study presents a critical advancement in clinical radiology by demonstrating the effectiveness of a deep neural network in reducing radiation exposure during pediatric interventional radiology while maintaining image quality, offering a potential solution to enhance patient safety.

KEY POINTS

• Reducing radiation: cutting IR image to reduce radiation. • Accurate frame interpolation: our model effectively interpolates missing frames. • High visual quality in terms of PSNR and SSIM, making IR procedures safer without sacrificing quality.

Reference standard for the evaluation of automatic segmentation algorithms: Quantification of inter observer variability of manual delineation of prostate contour on MRI

PURPOSE

The purpose of this study was to investigate the relationship between inter-reader variability in manual prostate contour segmentation on magnetic resonance imaging (MRI) examinations and determine the optimal number of readers required to establish a reliable reference standard.

MATERIALS AND METHODS

Seven radiologists with various experiences independently performed manual segmentation of the prostate contour (whole-gland [WG] and transition zone [TZ]) on 40 prostate MRI examinations obtained in 40 patients. Inter-reader variability in prostate contour delineations was estimated using standard metrics (Dice similarity coefficient [DSC], Hausdorff distance and volume-based metrics). The impact of the number of readers (from two to seven) on segmentation variability was assessed using pairwise metrics (consistency) and metrics with respect to a reference segmentation (conformity), obtained either with majority voting or simultaneous truth and performance level estimation (STAPLE) algorithm.

RESULTS

The average segmentation DSC for two readers in pairwise comparison was 0.919 for WG and 0.876 for TZ. Variability decreased with the number of readers: the interquartile ranges of the DSC were 0.076 (WG) / 0.021 (TZ) for configurations with two readers, 0.005 (WG) / 0.012 (TZ) for configurations with three readers, and 0.002 (WG) / 0.0037 (TZ) for configurations with six readers. The interquartile range decreased slightly faster between two and three readers than between three and six readers. When using consensus methods, variability often reached its minimum with three readers (with STAPLE, DSC = 0.96 [range: 0.945-0.971] for WG and DSC = 0.94 [range: 0.912-0.957] for TZ, and interquartile range was minimal for configurations with three readers.

CONCLUSION

The number of readers affects the inter-reader variability, in terms of inter-reader consistency and conformity to a reference. Variability is minimal for three readers, or three readers represent a tipping point in the variability evolution, with both pairwise-based metrics or metrics with respect to a reference. Accordingly, three readers may represent an optimal number to determine references for artificial intelligence applications.

French community grid for the evaluation of radiological artificial intelligence solutions (DRIM France Artificial Intelligence Initiative)

PURPOSE

The purpose of this study was to validate a national descriptive and analytical grid for artificial intelligence (AI) solutions in radiology.

MATERIALS AND METHODS

The RAND-UCLA Appropriateness Method was chosen by expert radiologists from the DRIM France IA group for this statement paper. The study, initiated by the radiology community, involved seven steps including literature review, template development, panel selection, pre-panel meeting survey, data extraction and analysis, second and final panel meeting, and data reporting.

RESULTS

The panel consisted of seven software vendors, three for bone fracture detection using conventional radiology and four for breast cancer detection using mammography. A consensus was reached on various aspects, including general target, main objective, certification marking, integration, expression of results, forensic aspects and cybersecurity, performance and scientific validation, description of the company and economic details, possible usage scenarios in the clinical workflow, database, specific objectives and targets of the AI tool.

CONCLUSION

The study validates a descriptive and analytical grid for radiological AI solutions consisting of ten items, using breast cancer and bone fracture as an experimental guide. This grid would assist radiologists in selecting relevant and validated AI solutions. Further developments of the grid are needed to include other organs and tasks.

CT and MRI of Gastrointestinal Stromal Tumors: New Trends and Perspectives

Gastrointestinal stromal tumors (GISTs) are defined as mesenchymal tumors of the gastrointestinal tract that express positivity for CD117, which is a c-KIT proto-oncogene antigen. Expression of the c-KIT protein, a tyrosine kinase growth factor receptor, allows the distinction between GISTs and other mesenchymal tumors such as leiomyoma, leiomyosarcoma, schwannoma and neurofibroma. GISTs can develop anywhere in the gastrointestinal tract, as well as in the mesentery and omentum. Over the years, the management of GISTs has improved due to a better knowledge of their behaviors and risk or recurrence, the identification of specific mutations and the use of targeted therapies. This has resulted in a better prognosis for patients with GISTs. In parallel, imaging of GISTs has been revolutionized by tremendous progress in the field of detection, characterization, survival prediction and monitoring during therapy. Recently, a particular attention has been given to radiomics for the characterization of GISTs using analysis of quantitative imaging features. In addition, radiomics has currently many applications that are developed in conjunction with artificial intelligence with the aim of better characterizing GISTs and providing a more precise assessment of tumor burden. This article sums up recent advances in computed tomography and magnetic resonance imaging of GISTs in the field of image/data acquisition, tumor detection, tumor characterization, treatment response evaluation, and preoperative planning.

Opportunities and challenges of traditional Chinese medicine doctors in the era of artificial intelligence

With the exponential advancement of artificial intelligence (AI) technology, the realm of medicine is experiencing a paradigm shift, engendering a multitude of prospects and trials for healthcare practitioners, encompassing those devoted to the practice of traditional Chinese medicine (TCM). This study explores the evolving landscape for TCM practitioners in the AI era, emphasizing that while AI can be helpful, it cannot replace the role of TCM practitioners. It is paramount to underscore the intrinsic worth of human expertise, accentuating that artificial intelligence (AI) is merely an instrument. On the one hand, AI-enabled tools like intelligent symptom checkers, diagnostic assistance systems, and personalized treatment plans can augment TCM practitioners' expertise and capacity, improving diagnosis accuracy and treatment efficacy. AI-empowered collaborations between Western medicine and TCM can strengthen holistic care. On the other hand, AI may disrupt conventional TCM workflow and doctor-patient relationships. Maintaining the humanistic spirit of TCM while embracing AI requires upholding professional ethics and establishing appropriate regulations. To leverage AI while retaining the essence of TCM, practitioners need to hone holistic analytical skills and see AI as complementary. By highlighting promising applications and potential risks of AI in TCM, this study provides strategic insights for stakeholders to promote the integrated development of AI and TCM for better patient outcomes. With proper implementation, AI can become a valuable assistant for TCM practitioners to elevate healthcare quality.

Accurate non-invasive grading of nonfunctional pancreatic neuroendocrine tumors with a CT derived radiomics signature

PURPOSE

The purpose of this study was to develop a radiomics-signature using computed tomography (CT) data for the preoperative prediction of grade of nonfunctional pancreatic neuroendocrine tumors (NF-PNETs).

MATERIALS AND METHODS

A retrospective study was performed on patients undergoing resection for NF-PNETs between 2010 and 2019. A total of 2436 radiomic features were extracted from arterial and venous phases of pancreas-protocol CT examinations. Radiomic features that were associated with final pathologic grade observed in the surgical specimens were subjected to joint mutual information maximization for hierarchical feature selection and the development of the radiomic-signature. Youden-index was used to identify optimal cutoff for determining tumor grade. A random forest prediction model was trained and validated internally. The performance of this tool in predicting tumor grade was compared to that of EUS-FNA sampling that was used as the standard of reference.

RESULTS

A total of 270 patients were included and a fusion radiomic-signature based on 10 selected features was developed using the development cohort (n = 201). There were 149 men and 121 women with a mean age of 59.4 ± 12.3 (standard deviation) years (range: 23.3-85.0 years). Upon internal validation in a new set of 69 patients, a strong discrimination was observed with an area under the curve (AUC) of 0.80 (95% confidence interval [CI]: 0.71-0.90) with corresponding sensitivity and specificity of 87.5% (95% CI: 79.7-95.3) and 73.3% (95% CI: 62.9-83.8) respectively. Of the study population, 143 patients (52.9%) underwent EUS-FNA. Biopsies were non-diagnostic in 26 patients (18.2%) and could not be graded due to insufficient sample in 42 patients (29.4%). In the cohort of 75 patients (52.4%) in whom biopsies were graded the radiomic-signature demonstrated not different AUC as compared to EUS-FNA (AUC: 0.69 vs. 0.67; P = 0.723), however greater sensitivity (i.e., ability to accurately identify G2/3 lesion was observed (80.8% vs. 42.3%; P < 0.001).

CONCLUSION

Non-invasive assessment of tumor grade in patients with PNETs using the proposed radiomic-signature demonstrated high accuracy. Prospective validation and optimization could overcome the commonly experienced diagnostic uncertainty in the assessment of tumor grade in patients with PNETs and could facilitate clinical decision-making.

Radiomics and artificial intelligence for soft-tissue sarcomas: Current status and perspectives

This article proposes a summary of the current status of the research regarding the use of radiomics and artificial intelligence to improve the radiological assessment of patients with soft tissue sarcomas (STS), a heterogeneous group of rare and ubiquitous mesenchymal malignancies. After a first part explaining the principle of radiomics approaches, from raw image post-processing to extraction of radiomics features mined with unsupervised and supervised machine-learning algorithms, and the current research involving deep learning algorithms in STS, especially convolutional neural networks, this review details their main research developments since the formalisation of 'radiomics' in oncologic imaging in 2010. This review focuses on CT and MRI and does not involve ultrasonography. Radiomics and deep radiomics have been successfully applied to develop predictive models to discriminate between benign soft-tissue tumors and STS, to predict the histologic grade (i.e., the most important prognostic marker of STS), the response to neoadjuvant chemotherapy and/or radiotherapy, and the patients' survivals and probability for presenting distant metastases. The main findings, limitations and expectations are discussed for each of these outcomes. Overall, after a first decade of publications emphasizing the potential of radiomics through retrospective proof-of-concept studies, almost all positive but with heterogeneous and often non-replicable methods, radiomics is now at a turning point in order to provide robust demonstrations of its clinical impact through open-science, independent databases, and application of good and standardized practices in radiomics such as those provided by the Image Biomarker Standardization Initiative, without forgetting innovative research paths involving other '-omics' data to better understand the relationships between imaging of STS, gene-expression profiles and tumor microenvironment.