Artificial intelligence-based clinical decision support in pediatrics

Ramgopal, Sriram; Sanchez-Pinto, L. Nelson; Horvat, Christopher M.; Carroll, Michael S.; Luo, Yuan; Florin, Todd A.

doi:10.1038/s41390-022-02226-1

Cited by 67 publications

(45 citation statements)

References 88 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The widespread interest in telehealth during the pandemic has also highlighted the importance of healthcare professionals' competency in its implementation [29,30]. Genomics and arti cial intelligence are expected to play a crucial role in predicting the prognosis of severe diseases, particularly cancer, and supporting clinical decision-making [31,32]. Mental health in children has gained increased attention during the pandemic, leading to proposed solutions utilizing various technologies [33,34].…”

Section: Discussionmentioning

confidence: 99%

Prioritizing Educational Initiatives on Emerging Technologies for Italian Pediatricians: bibliometric review and a survey

Tozzi

Gesualdo

Pandolfi

et al. 2023

Preprint

View full text Add to dashboard Cite

Background Emerging technologies have demonstrated outstanding potential in improving healthcare, yet their full integration remains a challenge for all medical specialties, including pediatrics. To support the swift implementation of technologies, we identified the current trends through a bibliometric review, and we conducted a survey on Italian pediatricians to gauge educational needs and willingness to integrate technologies into clinical practice. Methods A working group of pediatricians representing various backgrounds designed and coordinated the study. To identify relevant topics for educational strategy development, we focused on virtual reality, telehealth, natural language processing, smartphone applications, robotics, genomics, and artificial intelligence. A bibliometric analysis limited to 2018-2023 was performed to identify trends and emerging applications within each topic. Based on the results, a questionnaire was developed and made available online to all Italian pediatricians. The results were analyzed through descriptive analysis and a multivariable logistic regression to explore associations between technology adoption and sociodemographic characteristics. Results A total of 3,253 publications were found, with Telehealth and Telemedicine having the highest number of publications and Natural Language Processing the lowest. The number of respondents to the online questionnaire was 1,540, predominantly medical doctors with over 20 years of experience working as family pediatricians. Telehealth had the highest level of knowledge (95.2%), followed by smartphone applications (89.1%) and genomics (63.2%). The greatest potential for increased use through education programs was projected for natural language processing (+43.1%), artificial intelligence (+39.6%), and virtual and mixed reality (+38.1%). Female respondents and older individuals were less likely to use emerging technologies. Hospital pediatricians and residents were more likely to use AI. Conclusions We developed a replicable strategy to identify emerging themes in medical technologies relevant to pediatrics and assess the educational needs of pediatricians. A significant gap still exists between current and potential usage of emerging technologies among Italian pediatricians although they showed a positive attitude towards implementing these technologies following specific education programs. The study highlights the need for comprehensive education programs on emerging technologies in pediatrics and recommends addressing gender and age disparities in technology adoption.

show abstract

Section: Discussionmentioning

confidence: 99%

Prioritizing Educational Initiatives on Emerging Technologies for Italian Pediatricians: bibliometric review and a survey

Tozzi

Gesualdo

Pandolfi

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…A CDS tool for pneumonia, for example, would take clinical data in combination with radiology data (such as CXRs and their interpretation) to calculate a predicted probability for this outcome. When this probability occurs within certain stakeholder-defined risk parameters, CDS tools may inform the clinician with respect to the best course of action (Supplementary Figure) (15). NLP offers a mechanism by which to rapidly interpret CXR reports and incorporate the encoded result into a CDS tool.…”

Section: Introductionmentioning

confidence: 99%

The development of a novel natural language processing tool to identify pediatric chest radiograph reports with pneumonia

Rixe

Frisch

Wang

et al. 2023

Front. Digit. Health

View full text Add to dashboard Cite

ObjectiveChest radiographs are frequently used to diagnose community-acquired pneumonia (CAP) for children in the acute care setting. Natural language processing (NLP)-based tools may be incorporated into the electronic health record and combined with other clinical data to develop meaningful clinical decision support tools for this common pediatric infection. We sought to develop and internally validate NLP algorithms to identify pediatric chest radiograph (CXR) reports with pneumonia.Materials and methodsWe performed a retrospective study of encounters for patients from six pediatric hospitals over a 3-year period. We utilized six NLP techniques: word embedding, support vector machines, extreme gradient boosting (XGBoost), light gradient boosting machines Naïve Bayes and logistic regression. We evaluated their performance of each model from a validation sample of 1,350 chest radiographs developed as a stratified random sample of 35% admitted and 65% discharged patients when both using expert consensus and diagnosis codes.ResultsOf 172,662 encounters in the derivation sample, 15.6% had a discharge diagnosis of pneumonia in a primary or secondary position. The median patient age in the derivation sample was 3.7 years (interquartile range, 1.4–9.5 years). In the validation sample, 185/1350 (13.8%) and 205/1350 (15.3%) were classified as pneumonia by content experts and by diagnosis codes, respectively. Compared to content experts, Naïve Bayes had the highest sensitivity (93.5%) and XGBoost had the highest F1 score (72.4). Compared to a diagnosis code of pneumonia, the highest sensitivity was again with the Naïve Bayes (80.1%), and the highest F1 score was with the support vector machine (53.0%).ConclusionNLP algorithms can accurately identify pediatric pneumonia from radiography reports. Following external validation and implementation into the electronic health record, these algorithms can facilitate clinical decision support and inform large database research.

show abstract

“…

Recent advances of artificial intelligence (AI) in pediatrics hold great potential and primarily involve clinically oriented solutions, such as early-warning and clinical decision support systems. 1,2 Structured development and careful implementation of these solutions into the clinical workflow are challenging tasks currently withholding their actual use despite showing promising results. 3 Presently, clinical AI models are almost exclusively developed on readily available data acquired for different purposes, introducing bias and impairing model quality.

…”

mentioning

confidence: 99%