← Back to all news
arXiv Machine Learning By Aravind V. Kuruvikkattil, Lalitha Pranathi Pulavarthy, Rashmita Kudamala, Saptarshi Purkayastha

Machine Learning for Pre-Culture ESBL Risk Stratification to Guide Empiric Antibiotic Selection: A 12-Hospital Study of Enterobacteriaceae Cultures

Read the original on arXiv Machine Learning →

The study presents a cost‑sensitive XGBoost model that predicts ESBL‑producing Enterobacteriaceae at the time of culture ordering using 45 pre‑culture EHR features from 132,955 cultures across 12 hospitals. At 90% sensitivity the model achieves a 95.8% negative predictive value, potentially allowing carbapenem‑sparing therapy in non‑ICU settings while avoiding unnecessary broad‑spectrum treatment for 307 out of every 1,000 cultures, at the expense of 14 missed ESBL cases per 1,000. SHAP analysis highlights prior ESBL colonization as the strongest predictor, with minimal performance loss when socioeconomic deprivation features are omitted, supporting equitable bedside deployment.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

Related stories

arXiv Machine Learning
Oct 28

ARMD-Predict: multi-site machine learning to predict antibiotic resistance before culture results

ARMD‑Predict is a multi‑site machine‑learning system trained on 7.5 million culture‑antibiotic records from four academic medical centers to forecast antibiotic resistance before culture results are available. It offers two tiers of models: one set predicts susceptibility to individual antibiotics, while a hierarchical set first identifies the infecting organism and then its resistance profile. In internal testing, the models achieved median AUROCs of 0.77 for inpatient urine and 0.70 for inpatient blood cultures, and could flag a substantial proportion of ceftriaxone‑susceptible and resistant cases for narrower or broader therapy, respectively.

By Fateme Nateghi Haredasht, Nicholas P. Marshall, Wenyuan Chen, Lauren N. Cooper, Ziming Wei, Harrison Konsker, Jacob S. Chen, Michael Klompas, Cara McKenna, Richard J. Medford, Sanjat Kanjilal, Austin J. Schoeffler, Stephen P. Ma, Niaz Banaei, Steven M. Asch, Amy Chang, Stanley C. Deresinski, Mary K. Goldstein, Robert Tibshirani, Jonathan H. Chen
More like this →
arXiv Computer Vision
Oct 28

Multimodal Knowledge Aggregation for Microbial Keratitis: Geographic Confounding and Shortcut Risk in Pooled Multicountry Data

The study presents an open‑source multimodal model that classifies bacterial versus fungal microbial keratitis using 17,158 slit‑lamp images from 1,645 patients in the US and India. The model, which integrates cross‑modality contrastive learning, modality‑specific fine‑tuning, and seven clinical variables, achieved a macro‑F1 of 83.06% and an AUROC of 89.9% overall, outperforming five comparator methods. However, performance varied markedly by country—AUROC 83.9% in India versus 67.0% in the US—highlighting geographic confounding and the need for independent geographic validation.

By Yiqing Wang, Maria A. Woodward, Ziyun Yang, N. Venkatesh Prajna, Chunming He, Leslie M. Niziol, Mercy Pawar, Ming-Chen Lu, Guillermo Amescua, Rachel A. F. Wozniak, Sejal Amin, Abinaya Krishnan, Prabhleen Kochar, Sina Farsiu
fine-tuningmultimodal
More like this →