Chest X-ray Interpretation Research Articles

Applications and potential of machine learning augmented chest X-ray interpretation in cardiology.

The chest X-ray (CXR) has a wide range of clinical indications in the field of cardiology, from the assessment of acute pathology to disease surveillance and screening. Despite many technological advancements, CXR interpretation error rates have remained constant for decades. The application of machine learning has the potential to substantially improve clinical workflow efficiency, pathology detection accuracy, error rates and clinical decision making in cardiology. To date, machine learning has been developed to improve image processing, facilitate pathology detection, optimize the clinical workflow, and facilitate risk stratification. This review explores the current and potential future applications of machine learning for chest radiography to facilitate clinical decision making in cardiology. It maps the current state of the science and considers additional potential use cases from the perspective of clinicians and technologists actively engaged in the development and deployment of deep learning driven clinical decision support systems.

Post-deployment performance of a deep learning algorithm for normal and abnormal chest X-ray classification: A study at visa screening centers in the United Arab Emirates

BackgroundChest radiographs (CXRs) are widely used to screen for infectious diseases like tuberculosis and COVID-19 among migrants. At such high-volume settings, manual CXR reporting is challenging and integrating artificial intelligence (AI) algorithms into the workflow help to rule out normal findings in minutes, allowing radiologists to focus on abnormal cases. MethodsIn this post-deployment study, all the CXRs acquired during the visa screening process across 33 centers in United Arab Emirates from January 2021 to June 2022 (18 months) were included. The qXR v2.1 chest X-ray interpretation software was used to classify the scans into normal and abnormal, and its agreement against radiologist was evaluated. Additionally, a digital survey was conducted among 20 healthcare professionals with prior AI experience to understand real-world implementation challenges and impact. ResultsThe analysis of 1309,443 CXRs from 1309,431 patients (median age: 35 years; IQR [29–42]; 1030,071 males [78.7 %]) in this study revealed a Negative Predictive Value (NPV) of 99.92 % (95 % CI: 99.92, 99.93), Positive Predictive Value (PPV) of 5.06 % (95 % CI: 4.99, 5.13) and overall percent agreement of the AI with radiologists of 72.90 % (95 % CI: 72.82, 72.98). In the survey, majority (88.2 %) of the radiologists agreed to turnaround time reduction after AI integration, while 82 % suggested that the AI improved their diagnostic accuracy. DiscussionIn contrast with the existing studies, this research uses a substantially large data. A high NPV and satisfactory agreement with human readers indicate that AI can reliably identify normal CXRs, making it suitable for routine applications.

External quality assurance of chest X-ray interpretation to strengthen diagnosis of childhood TB.

Chest X-ray (CXR) misinterpretation negatively affects the accuracy of childhood TB diagnosis. External quality assurance (EQA) could strengthen CXR reading skills. We assessed the uptake, performance and challenges of an EQA of CXR interpretation within the childhood TB-Speed decentralisation study in six resource-limited countries. Every quarter, TB suggestive or unreadable CXRs and 10% of remaining CXRs from children with presumptive TB were selected for blind re-reading by national re-readers. The proportion of CXRs selected for EQA and re-read assessed the uptake. The performance was assessed by the proportion of discordant interpretations and the sensitivity and specificity of clinicians' vs re-readers' interpretations. Challenges were retrieved from country reports. Of 513 eligible CXRs, 309 (60.8%) were selected for EQA and 278/309 (90.0%) re-read. The proportion of discordant interpretation was between 13/48 (27%) in Sierra Leone and 7/13 (53.8%) in Cote d'Ivoire during the first EQA and decreased after the EQAs periods in 3/5 countries. Clinician sensitivity reached 100% in all countries over EQA. Specificity ranged between 13% in Sierra Leone and 65% in Cambodia (first EQA) and increased in 4/5 countries after the EQA periods. CXR transfer and re-readers' workload were the main challenges. EQA can enhance CXR interpretation for childhood TB diagnosis, provided operational challenges are overcome.

Experiment on Lung Disease Classification using YOLOv8

Lung diseases are a leading cause of illness and mortality worldwide. Accurate and timely diagnosis is crucial for improving patient outcomes, but manual interpretation of chest X-rays by radiologists may consume a significant amount of time and is susceptible to errors. This study leverages the state-of-the-art YOLOv8 deep learning model to develop an automated system for classifying lung diseases from chest X-ray images. The proposed approach utilizes a large, diverse dataset of 21,165 chest X-ray images categorized into four classes: COVID-19, viral pneumonia, lung opacity, and normal. The YOLOv8-cls model is fine-tuned using transfer learning and advanced data augmentation techniques. The system achieves a high accuracy of approximately 95% across all disease classes while maintaining real-time performance. Confusion matrix analysis demonstrates strong performance in correctly identifying each condition, with COVID-19, normal, viral pneumonia and lung opacity cases correctly identified 97%, 97%, 99% and 93% of the time respectively. The outcomes validate the adaptability and reliability of deep learning, specifically YOLOv8, for automated lung disease detection, offering the potential to improve clinical workflows and patient care by providing efficient screening tools for practitioners. The study addresses limitations of previous works by utilizing a large consolidated dataset, reporting comprehensive computational efficiency metrics, and extending classification complexity without substantially impacting accuracy or speed. Future research should focus on further optimizations and extending the curated image repositories to include under-represented patient groups to enhance the model's inclusiveness and robustness.

Radiograph accelerated detection and identification of cancer in the lung (RADICAL): a mixed methods study to assess the clinical effectiveness and acceptability of Qure.ai artificial intelligence software to prioritise chest X-ray (CXR) interpretation

IntroductionDiagnosing and treating lung cancer in early stages is essential for survival outcomes. The chest X-ray (CXR) remains the primary screening tool to identify lung cancers in the UK; however,...

Collaborative nursing education between advanced practice registered nurses

The surge in healthcare demands due to the explosion of growth in the aging adult population demands that academic institutions address enhanced education of Advanced Practice Registered Nurses (APRNs). This is particularly challenging amidst nurse faculty shortages in specialty areas, stemming from factors such as lack of pay competitiveness with clinical positions, and doctoral-prepared nurses seeking faculty positions. Despite efforts to address the shortages, recruitment challenges persist, necessitating innovative approaches. This article explores the interprofessional collaborative teaching between Nurse Anesthesia (NA) and Adult Gerontology Nurse Practitioner (NP) programs, focusing on airway management, sedation techniques, chest X-ray interpretation, and ultrasound skills. To date, this collaboration between NA and NP faculty has proven effective. Both synchronous and online asynchronous teaching have received overwhelmingly positive student feedback. Ongoing communication and collaborations between NA and NP faculty facilitate teaching and educational strategies across programs, sharing faculty expertise, and mitigating reduced faculty numbers. This innovative model benefits faculty and students and provides a platform for firsthand interprofessional collaboration, fostering mutual respect and preparing students for effective interdisciplinary healthcare teamwork.

Expanding molecular diagnostic coverage for tuberculosis by combining computer-aided chest radiography and sputum specimen pooling: a modeling study from four high-burden countries

BackgroundIn 2022, fewer than half of persons with tuberculosis (TB) had access to molecular diagnostic tests for TB due to their high costs. Studies have found that the use of artificial intelligence (AI) software for chest X-ray (CXR) interpretation and sputum specimen pooling can each reduce the cost of testing. We modeled the combination of both strategies to estimate potential savings in consumables that could be used to expand access to molecular diagnostics.MethodsWe obtained Xpert testing and positivity data segmented into deciles by AI probability scores for TB from the community- and healthcare facility-based active case finding conducted in Bangladesh, Nigeria, Viet Nam, and Zambia. AI scores in the model were based on CAD4TB version 7 (Zambia) and qXR (all other countries). We modeled four ordinal screening and testing approaches involving AI-aided CXR interpretation to indicate individual and pooled testing. Setting a false negative rate of 5%, for each approach we calculated additional and cumulative savings over the baseline of universal Xpert testing, as well as the theoretical expansion in diagnostic coverage.ResultsIn each country, the optimal screening and testing approach was to use AI to rule out testing in deciles with low AI scores and to guide pooled vs individual testing in persons with moderate and high AI scores, respectively. This approach yielded cumulative savings in Xpert tests over baseline ranging from 50.8% in Zambia to 57.5% in Nigeria and 61.5% in Bangladesh and Viet Nam. Using these savings, diagnostic coverage theoretically could be expanded by 34% to 160% across the different approaches and countries.ConclusionsUsing AI software data generated during CXR interpretation to inform a differentiated pooled testing strategy may optimize TB diagnostic test use, and could extend molecular tests to more people who need them. The optimal AI thresholds and pooled testing strategy varied across countries, which suggests that bespoke screening and testing approaches may be needed for differing populations and settings.

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

IntroductionIn 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 methodsSRs 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/. ResultsAfter 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. ConclusionsThe 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.

Revolutionizing Healthcare: Qure.AI's Innovations in Medical Diagnosis and Treatment.

Qure.AI, a leading company in artificial intelligence (AI) applied to healthcare, has developed a suite of innovative solutions to revolutionize medical diagnosis and treatment. With a plethora of FDA-approved tools for clinical use, Qure.AI continually strives for innovation in integrating AI into healthcare systems. This article delves into the efficacy of Qure.AI's chest X-ray interpretation tool, "qXR,"in medicine, drawing from a comprehensive review of clinical trials conducted by various institutions. Key applications of AI in healthcare include machine learning, deep learning, and natural language processing (NLP), all of which contribute to enhanced diagnostic accuracy, efficiency, and speed. Through the analysis of vast datasets, AI algorithms assist physicians in interpreting medical data and making informed decisions, thereby improving patient care outcomes. Illustrative examples highlight AI's impact on medical imaging, particularly in the diagnosis of conditions such as breast cancer, heart failure, and pulmonary nodules. AI can significantly reduce diagnostic errors and expedite the interpretation of medical images, leading to more timely interventions and treatments. Furthermore, AI-powered predictive analytics enable early detection of diseases and facilitate personalized treatment plans, thereby reducing healthcare costs and improving patient outcomes. The efficacy of AI in healthcare is underscored by its ability to complement traditional diagnostic methods, providing physicians with valuable insights and support in clinical decision-making. As AI continues to evolve, its role in patient care and medical research is poised to expand, promising further advancements in diagnostic accuracy and treatment efficacy.

Lung Manifestation of Dengue Fever: A Retrospective Study.

Introduction Dengue fever, caused by the dengue virus transmitted by Aedes aegypti mosquitoes, is a significant public health concern globally. Its resurgence in recent years, particularly in low- and middle-income countries, has led to increased morbidity and mortality rates. Atypical manifestations, involving the cardiac, liver, gut, renal, blood, bone, nervous, and respiratory systems, in dengue, can complicate both diagnosis and management. This study aimed to investigate the incidence of lung manifestations in dengue-infected individuals and their correlation with patient outcomes. Background The prevalence of dengue fever has risen dramatically over the past two decades, with Asia bearing the brunt of the burden, particularly India. The pathophysiology of lung complications in dengue remains unclear but is thought to be related to capillary leak syndrome and thrombocytopenia. Studies suggest that respiratory symptoms may be associated with severe cases and increased mortality rates. Despite limited research in India, understanding lung manifestations in dengue is crucial for improving diagnostic accuracy and patient care. Methods A retrospective study was conducted at K.S. Hegde Hospital, a tertiary care facility located in Mangalore, India, involving patients aged 18 years and above diagnosed with dengue fever between January and December 2019. Data gathered comprised patient demographics, clinical symptoms, laboratory findings, imaging results including radiographs, computed tomography (CT) scans of the chest (if accessible), ultrasound examinations of the chest and abdomen, and 2D echocardiograms, as well as patient outcomes. Diagnosis of lung manifestation was established through clinical assessment, chest X-ray interpretation, and ultrasound of the chest. Statistical analysis was conducted using SPSS Statistics (version 20), with a significance set at p<0.05. Results Out of 255 dengue cases, 10.19% (n=26) exhibited pulmonary manifestations, with pleural effusion being the most common. Older age (>50 years) and comorbidities were associated with a higher incidence of lung involvement. Respiratory symptoms, such as breathlessness, were more prevalent in patients with pulmonary complications. Laboratory parameters indicated distinct profiles in patients with lung manifestations, including elevated total count, urea, bilirubin, and liver enzymes, and reduced platelet counts. Mortality rates were higher in patients with lung involvement, older age, and comorbidities. Discussion The study findings highlight the importance of recognizing respiratory symptoms in dengue fever, particularly in older patients and those with underlying health conditions. The association between pulmonary involvement and adverse outcomes underscores the need for early detection and appropriate management strategies. Future research should focus on elucidating the pathophysiology of lung complications in dengue and developing targeted interventions to improve patient outcomes. Conclusion Lung manifestations in dengue fever represent a significant clinical challenge and are associated with increased morbidity and mortality. Early recognition of respiratory symptoms, along with prompt diagnostic evaluation and appropriate management, is essential for improving patient prognosis. Further studies are warranted to deepen our understanding of lung involvement in dengue and optimize therapeutic approaches to mitigate its impact on patient outcomes.

Decoding Pneumonia: Leveraging CNNS for Accurate Chest X-Ray Classification

Pneumonia is a known potentially fatal lung disease that is frequently referred to as a silent killer since it can lead to lung alveoli filling with pus or fluid, mainly from fungal, viral, or bacterial infections. Chest X-rays are the primary diagnostic tool for pneumonia; however, the diagnosis becomes more complex when other pulmonary disorders such volume loss, haemorrhage, lung cancer, fluid overload, and consequences from radiation or surgery are taken into account. As a result, the interpretation of chest X-rays becomes complex, which makes the development of computer-aided diagnosis systems necessary to help physicians make decisions that are more accurate. In order to diagnose pneumonia from chest X-ray pictures, the research reported here uses a convolutional neural network (CNN) enhanced with a self-attention mechanism. 'Normal' and 'pneumonia' classes are included in the dataset used in the study methodology, and data augmentation techniques are applied to improve the model's resilience. By means of extensive evaluation metrics and visualizations, the study highlights the potential of the suggested model as a useful instrument to aid clinicians in diagnosing pneumonia, consequently reducing the difficulties linked to the interpretation of chest X-rays in the context of various pulmonary conditions.

A Comprehensive Review of Machine Learning Techniques for Pneumonia Detection from Chest X-ray Images

Pneumonia, a significant contributor to global respiratory illnesses, demands prompt diagnosis for successful treatment. Chest X-rays, being widely accessible and cost-effective, serve as the primary imaging modality for pneumonia detection. However, manual interpretation of chest X-rays is subjective and susceptible to errors due to the often subtle and diverse visual indicators of pneumonia. This study delves into the potential of deep learning, a subfield of artificial intelligence, to automate pneumonia detection utilizing chest X-ray images. By leveraging the well-established VGG-16 model, a convolutional neural network architecture renowned for its image recognition capabilities, the study aims to develop a robust and accurate classification system. This system, if successful, has the potential to alleviate the burden on healthcare professionals by assisting in the timely and efficient detection of pneumonia, ultimately contributing to improved patient outcomes. Key Words: Deep Learning, VGG16, CNN, Pneumonia, Chest X-rays .

A deep-learning-based framework for identifying and localizing multiple abnormalities and assessing cardiomegaly in chest X-ray

Accurate identification and localization of multiple abnormalities are crucial steps in the interpretation of chest X-rays (CXRs); however, the lack of a large CXR dataset with bounding boxes severely constrains accurate localization research based on deep learning. We created a large CXR dataset named CXR-AL14, containing 165,988 CXRs and 253,844 bounding boxes. On the basis of this dataset, a deep-learning-based framework was developed to identify and localize 14 common abnormalities and calculate the cardiothoracic ratio (CTR) simultaneously. The mean average precision values obtained by the model for 14 abnormalities reached 0.572-0.631 with an intersection-over-union threshold of 0.5, and the intraclass correlation coefficient of the CTR algorithm exceeded 0.95 on the held-out, multicentre and prospective test datasets. This framework shows an excellent performance, good generalization ability and strong clinical applicability, which is superior to senior radiologists and suitable for routine clinical settings.

Evaluation of a short training course of chest X-ray interpretation for the diagnosis of paediatric TB.

Chest X-ray (CXR) interpretation is challenging for the diagnosis of paediatric TB. We assessed the performance of a three half-day CXR training module for healthcare workers (HCWs) at low healthcare levels in six high TB incidence countries. Within the TB-Speed Decentralization Study, we developed a three half-day training course to identify normal CXR, CXR of good quality and identify six TB-suggestive features. We performed a pre-post training assessment on a pre-defined set of 20 CXR readings. We compared the proportion of correctly interpreted CXRs and the median reading score before and after the training using the McNemar test and a linear mixed model. Of 191 HCWs, 43 (23%) were physicians, 103 (54%) nurses, 18 (9.4%) radiology technicians and 12 (6.3%) other professionals. Of 2,840 CXRs with both assessment, respectively 1,843 (64.9%) and 2,277 (80.2%) were correctly interpreted during pre-training and post-training (P < 0.001). The median reading score improved significantly from 13/20 to 16/20 after the training, after adjusting by country, facility and profession (adjusted β = 3.31, 95% CI 2.44-4.47). Despite some limitations of the course assessment that did not include abnormal non-TB suggestive CXR, study findings suggest that a short CXR training course could improve HCWs' interpretation skills in diagnosing paediatric TB.

What is New in the Diagnosis of Childhood Tuberculosis?

The fact that almost half of the 1 million cases of childhood tuberculosis (TB)globally remain undiagnosed jeopardizes the TB elimination goal. Fortunately, there are new advances in this fieldwhich have the potential to bridge this diagnostic gap. Advances in imaging include computer assisted interpretation of chest X-rays (CXRs), point of care ultrasound (POCUS) and faster and superior computed tomography/ magnetic resonance imaging (CT/ MRI) protocols. The urine lipoarabinomannan test has proved to be a good point of care test for diagnosing TB in Human immunodeficiency virus (HIV) infected children. Stool and nasopharyngeal aspirates are emerging as acceptable alternatives for gastric lavage and induced sputum for diagnosing intrathoracic tuberculosis. Xpert MTB/RIF Ultra has improved sensitivitycompared to Xpert MTB/RIF for diagnosing both pulmonary/ extrapulmonary TB. Xpert XDR is another commercially available accurate point of care test for detecting resistance to drugs other than rifampicin in smear positive samples. Other molecular methods including new line probe assays, pyrosequencing, whole genome sequencing, and targeted next generation sequencing are extremely promising but not available commercially at present. The C-Tb skin test isan acceptable alternative to the tuberculin skin test and interferon gamma release assays for diagnosis of latent infection. There is an urgent need to incorporate some of these advances in the existingdiagnostic algorithms of childhood TB.

Clinical Application of Artificial Intelligence-Based Detection Assistance Devices for Chest X-Ray Interpretation: Current Status and Practical Considerations

Artificial intelligence (AI) technology is actively being applied for the interpretation of medical imaging, such as chest X-rays. AI-based software medical devices, which automatically detect various types of abnormal findings in chest X-ray images to assist physicians in their interpretation, are actively being commercialized and clinically implemented in Korea. Several important issues need to be considered for AI-based detection assistant tools to be applied in clinical practice: the evaluation of performance and efficacy prior to implementation; the determination of the target application, range, and method of delivering results; and monitoring after implementation and legal liability issues. Appropriate decision making regarding these devices based on the situation in each institution is necessary. Radiologists must be engaged as medical assessment experts using the software for these devices as well as in medical image interpretation to ensure the safe and efficient implementation and operation of AI-based detection assistant tools.

Validation of a Deep Learning Chest X-ray Interpretation Model: Integrating Large-Scale AI and Large Language Models for Comparative Analysis with ChatGPT.

This study evaluates the diagnostic accuracy and clinical utility of two artificial intelligence (AI) techniques: Kakao Brain Artificial Neural Network for Chest X-ray Reading (KARA-CXR), an assistive technology developed using large-scale AI and large language models (LLMs), and ChatGPT, a well-known LLM. The study was conducted to validate the performance of the two technologies in chest X-ray reading and explore their potential applications in the medical imaging diagnosis domain. The study methodology consisted of randomly selecting 2000 chest X-ray images from a single institution's patient database, and two radiologists evaluated the readings provided by KARA-CXR and ChatGPT. The study used five qualitative factors to evaluate the readings generated by each model: accuracy, false findings, location inaccuracies, count inaccuracies, and hallucinations. Statistical analysis showed that KARA-CXR achieved significantly higher diagnostic accuracy compared to ChatGPT. In the 'Acceptable' accuracy category, KARA-CXR was rated at 70.50% and 68.00% by two observers, while ChatGPT achieved 40.50% and 47.00%. Interobserver agreement was moderate for both systems, with KARA at 0.74 and GPT4 at 0.73. For 'False Findings', KARA-CXR scored 68.00% and 68.50%, while ChatGPT scored 37.00% for both observers, with high interobserver agreements of 0.96 for KARA and 0.97 for GPT4. In 'Location Inaccuracy' and 'Hallucinations', KARA-CXR outperformed ChatGPT with significant margins. KARA-CXR demonstrated a non-hallucination rate of 75%, which is significantly higher than ChatGPT's 38%. The interobserver agreement was high for KARA (0.91) and moderate to high for GPT4 (0.85) in the hallucination category. In conclusion, this study demonstrates the potential of AI and large-scale language models in medical imaging and diagnostics. It also shows that in the chest X-ray domain, KARA-CXR has relatively higher accuracy than ChatGPT.

Medical Interns' Skill Levels in Emergency Chest X-ray Interpretation

Objectives: The research aims to evaluate the proficiency of medical interns in identifying critical findings on chest X-rays, such as pneumothorax, pleural effusion, pulmonary edema, pneumonia, and other urgent conditions. It also seeks to determine the interns' capability to differentiate between normal and abnormal findings, as well as their ability to prioritize and communicate these findings to senior medical staff promptly. Materials and Methods: A cross-sectional online questionnaire-based form was used to collect data from medical interns across different medical schools around Saudi Arabia. Pearson correlation was carried out to see the relationship between x-ray interpretation scores and degree of confidence. A Fischer exact test was conducted to find an association between the demographic characteristics and diagnostic accuracy of X-rays. Results: A total of 545 medical interns from different universities participated in this study. The lowest diagnosis accuracy (27%) was when identifying a case of left lower lobe pneumonia. On the other hand, the most identified case (60.9%) was COVID-19 pneumonia. In cases of life-threatening conditions, only 13.1% of them have high confidence in their diagnosis of pneumoperitoneum, and 23.8% have high confidence in their diagnosis of pneumothorax. Normal CXR was answered incorrectly by 47.7% of the participants. The performance among recently started interns for some conditions is better than that of recently finished interns. Moreover, there are no significant differences in performance between those who are interested in radiology and those who are not. More than half of the participants advise doing extracurricular courses, practical case-based training in college, and volunteer work in radiology to improve their CXR interpretation skills. Conclusion: The competence of medical interns in interpreting chest X-rays varied across different medical conditions. Only 27.0% of interns correctly identified left lower lobe pneumonia, while 60.9% of them were able to identify COVID-19 pneumonia. Medical interns' competencies have limitations across different medical conditions; therefore, increasing exposure to radiological cases during internships and adequate training in college are recommended.

Deploying artificial intelligence software in an NHS trust: a how-to guide for clinicians.

Over the past 10 years, artificial intelligence (AI) has become one of the fastest-growing sectors in healthcare. There are now numerous new technologies designed to cut costs and improve diagnoses and treatment pathways. However, there is significant scepticism amongst National Health Service (NHS) staff regarding the usefulness of AI and it's cost to the NHS. This has likely resulted in underuse and slow adoption of software that may revolutionize our healthcare system and ensure its continued survival and effectiveness. Several governing bodies have put forward guidance on the safe and effective adoption of AI tools, but this rarely covers the reality of selecting and deploying new software. This article set out clear guidance on the practicalities and pitfalls of deploying digital solutions in healthcare, using the example of a deep learning algorithm designed to improve the accuracy of chest X-ray (CXR) interpretation in the emergency department.

Development and validation of open-source deep neural networks for comprehensive chest x-ray reading: a retrospective, multicentre study

Artificial intelligence (AI) systems for automated chest x-ray interpretation hold promise for standardising reporting and reducing delays in health systems with shortages of trained radiologists. Yet, there are few freely accessible AI systems trained on large datasets for practitioners to use with their own data with a view to accelerating clinical deployment of AI systems in radiology. We aimed to contribute an AI system for comprehensive chest x-ray abnormality detection. In this retrospective cohort study, we developed open-source neural networks, X-Raydar and X-Raydar-NLP, for classifying common chest x-ray findings from images and their free-text reports. Our networks were developed using data from six UK hospitals from three National Health Service (NHS) Trusts (University Hospitals Coventry and Warwickshire NHS Trust, University Hospitals Birmingham NHS Foundation Trust, and University Hospitals Leicester NHS Trust) collectively contributing 2 513 546 chest x-ray studies taken from a 13-year period (2006-19), which yielded 1 940 508 usable free-text radiological reports written by the contemporary assessing radiologist (collectively referred to as the "historic reporters") and 1 896 034 frontal images. Chest x-rays were labelled using a taxonomy of 37 findings by a custom-trained natural language processing (NLP) algorithm, X-Raydar-NLP, from the original free-text reports. X-Raydar-NLP was trained on 23 230 manually annotated reports and tested on 4551 reports from all hospitals. 1 694 921 labelled images from the training set and 89 238 from the validation set were then used to train a multi-label image classifier. Our algorithms were evaluated on three retrospective datasets: a set of exams sampled randomly from the full NHS dataset reported during clinical practice and annotated using NLP (n=103 328); a consensus set sampled from all six hospitals annotated by three expert radiologists (two independent annotators for each image and a third consultant to facilitate disagreement resolution) under research conditions (n=1427); and an independent dataset, MIMIC-CXR, consisting of NLP-annotated exams (n=252 374). X-Raydar achieved a mean AUC of 0·919 (SD 0·039) on the auto-labelled set, 0·864 (0·102) on the consensus set, and 0·842 (0·074) on the MIMIC-CXR test, demonstrating similar performance to the historic clinical radiologist reporters, as assessed on the consensus set, for multiple clinically important findings, including pneumothorax, parenchymal opacification, and parenchymal mass or nodules. On the consensus set, X-Raydar outperformed historical reporter balanced accuracy with significance on 27 of 37 findings, was non-inferior on nine, and inferior on one finding, resulting in an average improvement of 13·3% (SD 13·1) to 0·763 (0·110), including a mean 5·6% (13·2) improvement in critical findings to 0·826 (0·119). Our study shows that automated classification of chest x-rays under a comprehensive taxonomy can achieve performance levels similar to those of historical reporters and exhibit robust generalisation to external data. The open-sourced neural networks can serve as foundation models for further research and are freely available to the research community. Wellcome Trust.