Artificial and human intelligence for early identification of neonatal sepsis

Sullivan, Brynne A.; Kausch, Sherry L.; Fairchild, Karen D.

doi:10.1038/s41390-022-02274-7

Cited by 12 publications

(6 citation statements)

References 93 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Late recognition of sepsis, when overt clinical signs represent organ dysfunction due to systemic inflammation, can lead to fatal or long-lasting organ damage. Therefore, precision medicine tools using predictive monitoring aim to address this challenge by guiding clinical decisions to right-time antibiotics 7 .…”

Section: Introductionmentioning

confidence: 99%

Cardiorespiratory signature of neonatal sepsis: development and validation of prediction models in 3 NICUs

et al. 2023

Self Cite

View full text Add to dashboard Cite

Background: Heart rate characteristics aid early detection of late-onset sepsis (LOS), but respiratory data contain additional signatures of illness due to infection. Predictive models using cardiorespiratory data may improve early sepsis detection. We hypothesized that heart rate (HR) and oxygenation (SpO 2 ) data contain signatures that improve sepsis risk prediction over HR or demographics alone. Methods: We analyzed cardiorespiratory data from very low birth weight (VLBW, <1500g) infants admitted to three NICUs. We developed and externally validated four machine learning models to predict LOS using features calculated every 10m: mean, standard deviation, skewness, kurtosis of HR and SpO 2 , and cross-correlation. We compared feature importance, discrimination, calibration, and dynamic prediction across models and cohorts. We built models of demographics and HR or SpO 2 features alone for comparison with HR-SpO 2 models. Results: Performance, feature importance, and calibration were similar among modeling methods. All models had favorable external validation performance. The HR-SpO 2 model performed better than models using either HR or SpO 2 alone. Demographics improved the discrimination of all physiologic data models but dampened dynamic performance. Conclusions: Cardiorespiratory signatures detect LOS in VLBW infants at 3 NICUs. Demographics risk-stratify, but predictive modeling with both HR and SpO 2 features provides the best dynamic risk prediction.

show abstract

Section: Introductionmentioning

confidence: 99%

Cardiorespiratory signature of neonatal sepsis: development and validation of prediction models in 3 NICUs

et al. 2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…47 Likewise, the early identification of sepsis in neonates was successfully accomplished through the utilization of AI models, including CNN and long short-term memory networks. 108,109 Li et al employed the XGBoost algorithm to create a predictive model named CI-Lab8 for assessing the risk of cerebral infarction (CI). This model analyzed a set of factors, including fibrinogen, age, glucose levels, mean erythrocyte HGB concentration, albumin, absolute neutrophil count, activated partial thromboplastin time, and triglycerides.…”

Section: Study Of Ai Application In Diagnosis Of Other Diseasesmentioning

confidence: 99%

Integration of artificial intelligence in clinical laboratory medicine: Advancements and challenges

Xie,

Jia,

Liu

2024

Interdisciplinary Medicine

View full text Add to dashboard Cite

Artificial intelligence (AI)‐driven analysis of comprehensive clinical parameters is bringing about a significant transformation in traditional routine clinical laboratory testing. This transformation impacts the prediction, prevention, diagnosis, and prognosis of human diseases. AI possesses the capability to efficiently analyze and process vast and intricate datasets, thereby facilitating the development of diverse and efficient diagnostic or predictive models. This advancement is fueling significant improvements in laboratory quality, automation, and the accuracy of diagnoses. In this context, we conducted a thorough review and discussion on the progression of AI applications in clinical laboratory medicine, encompassing advancements, implementation, and challenges. Our conclusion underscores that integrating AI into clinical laboratory testing will notably propel personalized precision medicine forward and enhance diagnostic accuracy, especially benefiting patients for whom accurate diagnoses are elusive through traditional laboratory testing systems.

show abstract

“…Quantitative early warning algorithms should seek to make good on the promises of outcome calculators and rapid response teams by providing accurate, timely warning of an impending event where mitigation is likely to prevent damage or adverse outcomes. The challenge lies in integrating predictive analytics with clinician decisions (22). In doing so, such technology should reduce mortality and morbidity, but only if it performs as intended in the population to which it is applied.…”

Section: It Doesn't Matter What They Say In the Papers… It's Still Ro...mentioning

confidence: 99%

It doesn’t matter what they say in the papers... It’s still ROC and roll to me

King¹,

Sullivan²,

Vesoulis³

2023

Ann Transl Med

View full text Add to dashboard Cite

show abstract

Artificial and human intelligence for early identification of neonatal sepsis

Cited by 12 publications

References 93 publications

Cardiorespiratory signature of neonatal sepsis: development and validation of prediction models in 3 NICUs

Cardiorespiratory signature of neonatal sepsis: development and validation of prediction models in 3 NICUs

Integration of artificial intelligence in clinical laboratory medicine: Advancements and challenges

It doesn’t matter what they say in the papers... It’s still ROC and roll to me

Contact Info

Product

Resources

About