Pediatric chest radiograph interpretation: how far has artificial intelligence come? A systematic literature review

Three-Stage Framework for Accurate Pediatric Chest X-ray Diagnosis Using Self-Supervision and Transfer Learning on Small Datasets

Pediatric respiratory disease diagnosis and subsequent treatment require accurate and interpretable analysis. A chest X-ray is the most cost-effective and rapid method for identifying and monitoring various thoracic diseases in children. Recent developments in self-supervised and transfer learning have shown their potential in medical imaging, including chest X-ray areas. In this article, we propose a three-stage framework with knowledge transfer from adult chest X-rays to aid the diagnosis and interpretation of pediatric thorax diseases. We conducted comprehensive experiments with different pre-training and fine-tuning strategies to develop transformer or convolutional neural network models and then evaluate them qualitatively and quantitatively. The ViT-Base/16 model, fine-tuned with the CheXpert dataset, a large chest X-ray dataset, emerged as the most effective, achieving a mean AUC of 0.761 (95% CI: 0.759-0.763) across six disease categories and demonstrating a high sensitivity (average 0.639) and specificity (average 0.683), which are indicative of its strong discriminative ability. The baseline models, ViT-Small/16 and ViT-Base/16, when directly trained on the Pediatric CXR dataset, only achieved mean AUC scores of 0.646 (95% CI: 0.641-0.651) and 0.654 (95% CI: 0.648-0.660), respectively. Qualitatively, our model excels in localizing diseased regions, outperforming models pre-trained on ImageNet and other fine-tuning approaches, thus providing superior explanations. The source code is available online and the data can be obtained from PhysioNet.

Methodological evaluation of systematic reviews based on the use of artificial intelligence systems in chest radiography

INTRODUCTION

In recent years, systems that use artificial intelligence (AI) in medical imaging have been developed, such as the interpretation of chest X-ray to rule out pathology. This has produced an increase in systematic reviews (SR) published on this topic. This article aims to evaluate the methodological quality of SRs that use AI for the diagnosis of thoracic pathology by simple chest X-ray.

MATERIAL AND METHODS

SRs evaluating the use of AI systems for the automatic reading of chest X-ray were selected. Searches were conducted (from inception to May 2022): PubMed, EMBASE, and the Cochrane Database of Systematic Reviews. Two investigators selected the reviews. From each SR, general, methodological and transparency characteristics were extracted. The PRISMA statement for diagnostic tests (PRISMA-DTA) and AMSTAR-2 were used. A narrative synthesis of the evidence was performed. Protocol registry: Open Science Framework: https://osf.io/4b6u2/.

RESULTS

After applying the inclusion and exclusion criteria, 7 SRs were selected (mean of 36 included studies per review). All the included SRs evaluated "deep learning" systems in which chest X-ray was used for the diagnosis of infectious diseases. Only 2 (29%) SRs indicated the existence of a review protocol. None of the SRs specified the design of the included studies or provided a list of excluded studies with their justification. Six (86%) SRs mentioned the use of PRISMA or one of its extensions. The risk of bias assessment was performed in 4 (57%) SRs. One (14%) SR included studies with some validation of AI techniques. Five (71%) SRs presented results in favour of the diagnostic capacity of the intervention. All SRs were rated "critically low" following AMSTAR-2 criteria.

CONCLUSIONS

The methodological quality of SRs that use AI systems in chest radiography can be improved. The lack of compliance in some items of the tools used means that the SRs published in this field must be interpreted with caution.

Evaluation of the Performance of an Artificial Intelligence (AI) Algorithm in Detecting Thoracic Pathologies on Chest Radiographs

The purpose of the study was to assess the performance of readers in diagnosing thoracic anomalies on standard chest radiographs (CXRs) with and without a deep-learning-based AI tool (Rayvolve) and to evaluate the standalone performance of Rayvolve in detecting thoracic pathologies on CXRs. This retrospective multicentric study was conducted in two phases. In phase 1, nine readers independently reviewed 900 CXRs from imaging group A and identified thoracic abnormalities with and without AI assistance. A consensus from three radiologists served as the ground truth. In phase 2, the standalone performance of Rayvolve was evaluated on 1500 CXRs from imaging group B. The average values of AUC across the readers significantly increased by 15.94%, with AI-assisted reading compared to unaided reading (0.88 ± 0.01 vs. 0.759 ± 0.07, p < 0.001). The time taken to read the CXRs decreased significantly, by 35.81% with AI assistance. The average values of sensitivity and specificity across the readers increased significantly by 11.44% and 2.95% with AI-assisted reading compared to unaided reading (0.857 ± 0.02 vs. 0.769 ± 0.02 and 0.974 ± 0.01 vs. 0.946 ± 0.01, p < 0.001). From the standalone perspective, the AI model achieved an average sensitivity, specificity, PPV, and NPV of 0.964, 0.844, 0.757, and 0.9798. The speed and performance of the readers improved significantly with AI assistance.

Artificial Intelligence in Pediatric Emergency Medicine: Applications, Challenges, and Future Perspectives

The dawn of Artificial intelligence (AI) in healthcare stands as a milestone in medical innovation. Different medical fields are heavily involved, and pediatric emergency medicine is no exception. We conducted a narrative review structured in two parts. The first part explores the theoretical principles of AI, providing all the necessary background to feel confident with these new state-of-the-art tools. The second part presents an informative analysis of AI models in pediatric emergencies. We examined PubMed and Cochrane Library from inception up to April 2024. Key applications include triage optimization, predictive models for traumatic brain injury assessment, and computerized sepsis prediction systems. In each of these domains, AI models outperformed standard methods. The main barriers to a widespread adoption include technological challenges, but also ethical issues, age-related differences in data interpretation, and the paucity of comprehensive datasets in the pediatric context. Future feasible research directions should address the validation of models through prospective datasets with more numerous sample sizes of patients. Furthermore, our analysis shows that it is essential to tailor AI algorithms to specific medical needs. This requires a close partnership between clinicians and developers. Building a shared knowledge platform is therefore a key step.

The unintended consequences of artificial intelligence in paediatric radiology

Over the past decade, there has been a dramatic rise in the interest relating to the application of artificial intelligence (AI) in radiology. Originally only 'narrow' AI tasks were possible; however, with increasing availability of data, teamed with ease of access to powerful computer processing capabilities, we are becoming more able to generate complex and nuanced prediction models and elaborate solutions for healthcare. Nevertheless, these AI models are not without their failings, and sometimes the intended use for these solutions may not lead to predictable impacts for patients, society or those working within the healthcare profession. In this article, we provide an overview of the latest opinions regarding AI ethics, bias, limitations, challenges and considerations that we should all contemplate in this exciting and expanding field, with a special attention to how this applies to the unique aspects of a paediatric population. By embracing AI technology and fostering a multidisciplinary approach, it is hoped that we can harness the power AI brings whilst minimising harm and ensuring a beneficial impact on radiology practice.

Residual networks models detection of atrial septal defect from chest radiographs

OBJECT

The purpose of this study was to explore a machine learning-based residual networks (ResNets) model to detect atrial septal defect (ASD) on chest radiographs.

METHODS

This retrospective study included chest radiographs consecutively collected at our hospital from June 2017 to May 2022. Qualified chest radiographs were obtained from patients who had finished echocardiography. These chest radiographs were labeled as positive or negative for ASD based on the echocardiographic reports and were divided into training, validation, and test dataset. Six ResNets models were employed to examine and compare by using the training dataset and was tuned using the validation dataset. The area under the curve, recall, precision and F1-score were taken as the evaluation metrics for classification result in the test dataset. Visualizing regions of interest for the ResNets models using heat maps.

RESULTS

This study included a total of 2105 chest radiographs of children with ASD (mean age 4.14 ± 2.73 years, 54% male), patients were randomly assigned to training, validation, and test dataset with an 8:1:1 ratio. Healthy children's images were supplemented to three datasets in a 1:1 ratio with ASD patients. Following the training, ResNet-10t and ResNet-18D have a better estimation performance, with precision, recall, accuracy, F1-score, and the area under the curve being (0.92, 0.93), (0.91, 0.91), (0.90, 0.90), (0.91, 0.91) and (0.97, 0.96), respectively. Compared to ResNet-18D, ResNet-10t was more focused on the distribution of the heat map of the interest region for most chest radiographs from ASD patients.

CONCLUSION

The ResNets model is feasible for identifying ASD through children's chest radiographs. ResNet-10t stands out as the preferable estimation model, providing exceptional performance and clear interpretability.

Applying Artificial Intelligence to Pediatric Chest Imaging: Reliability of Leveraging Adult-Based Artificial Intelligence Models

OBJECTIVE

The scarcity of artificial intelligence (AI) applications designed for use in pediatric patients can cause a significant health disparity in this vulnerable population. We investigated the performance of an adult-trained algorithm in detecting pneumonia in a pediatric population to explore the viability of leveraging adult-trained algorithms to accelerate pediatric AI research.

METHODS

We analyzed a publicly available pediatric chest x-ray dataset using an AI algorithm from TorchXRayVision. A 60% threshold was used to make binary predictions for pneumonia presence. Predictions were compared with dataset labels. Performance measures including true-positive rate, false-positive rate, true-negative rate, false-negative rate, sensitivity, specificity, positive predictive value (PPV), negative predictive value, accuracy, and F1-score were calculated for the complete dataset and bacterial and viral pneumonia subsets.

RESULTS

Overall (n = 5,856), the algorithm identified 3,923 cases with pneumonia (67.00%) and 1,933 (33.00%) normal cases. In comparison with the actual image labels, there were 3,411 (58.25%) true-positive cases, 512 (8.74%) false-positive cases, 1,071 (18.29%) true-negative cases, and 862 (14.72%) false-negative cases resulting in 79.83% sensitivity, 67.66% specificity, 86.95% PPV, 55.41% negative predictive value, and 76.54% accuracy and an F1-score of 0.83. Although the performance remained consistent in the bacterial pneumonia group, there was a significant decrease in PPV (69.9%) and F1-score (0.74) in the viral pneumonia group.

CONCLUSION

An adult-trained model adequately detected pneumonia in pediatric patients aged 1 to 5 years. Though models trained exclusively on pediatric images performed better, leveraging adult-based algorithms and datasets can expedite pediatric AI research.

Artificial intelligence-based clinical decision support in pediatrics

Machine learning models may be integrated into clinical decision support (CDS) systems to identify children at risk of specific diagnoses or clinical deterioration to provide evidence-based recommendations. This use of artificial intelligence models in clinical decision support (AI-CDS) may have several advantages over traditional "rule-based" CDS models in pediatric care through increased model accuracy, with fewer false alerts and missed patients. AI-CDS tools must be appropriately developed, provide insight into the rationale behind decisions, be seamlessly integrated into care pathways, be intuitive to use, answer clinically relevant questions, respect the content expertise of the healthcare provider, and be scientifically sound. While numerous machine learning models have been reported in pediatric care, their integration into AI-CDS remains incompletely realized to date. Important challenges in the application of AI models in pediatric care include the relatively lower rates of clinically significant outcomes compared to adults, and the lack of sufficiently large datasets available necessary for the development of machine learning models. In this review article, we summarize key concepts related to AI-CDS, its current application to pediatric care, and its potential benefits and risks. IMPACT: The performance of clinical decision support may be enhanced by the utilization of machine learning-based algorithms to improve the predictive performance of underlying models. Artificial intelligence-based clinical decision support (AI-CDS) uses models that are experientially improved through training and are particularly well suited toward high-dimensional data. The application of AI-CDS toward pediatric care remains limited currently but represents an important area of future research.