|
1
|
Phongpreecha T, Ghanem M, Reiss JD, Oskotsky TT, Mataraso SJ, De Francesco D, Reincke SM, Espinosa C, Chung P, Ng T, Costello JM, Sequoia JA, Razdan S, Xie F, Berson E, Kim Y, Seong D, Szeto MY, Myers F, Gu H, Feister J, Verscaj CP, Rose LA, Sin LWY, Oskotsky B, Roger J, Shu CH, Shome S, Yang LK, Tan Y, Levitte S, Wong RJ, Gaudillière B, Angst MS, Montine TJ, Kerner JA, Keller RL, Shaw GM, Sylvester KG, Fuerch J, Chock V, Gaskari S, Stevenson DK, Sirota M, Prince LS, Aghaeepour N. AI-guided precision parenteral nutrition for neonatal intensive care units. Nat Med 2025:10.1038/s41591-025-03601-1. [PMID: 40133525 DOI: 10.1038/s41591-025-03601-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/23/2024] [Accepted: 02/17/2025] [Indexed: 03/27/2025]
Abstract
One in ten neonates are admitted to neonatal intensive care units, highlighting the need for precise interventions. However, the application of artificial intelligence (AI) in guiding neonatal care remains underexplored. Total parenteral nutrition (TPN) is a life-saving treatment for preterm neonates; however, implementation of the therapy in its current form is subjective, error-prone and resource-consuming. Here, we developed TPN2.0-a data-driven approach that optimizes and standardizes TPN using information collected routinely in electronic health records. We assembled a decade of TPN compositions (79,790 orders; 5,913 patients) at Stanford to train TPN2.0. In addition to internal validation, we also validated our model in an external cohort (63,273 orders; 3,417 patients) from a second hospital. Our algorithm identified 15 TPN formulas that can enable a precision-medicine approach (Pearson's R = 0.94 compared to experts), increasing safety and potentially reducing cost. A blinded study (n = 192) revealed that physicians rated TPN2.0 higher than current best practice. In patients with high disagreement between the actual prescriptions and TPN2.0, standard prescriptions were associated with increased morbidities (for example, odds ratio = 3.33; P value = 0.0007 for necrotizing enterocolitis), while TPN2.0 recommendations were linked to reduced risk. Finally, we demonstrated that TPN2.0 employing a transformer architecture enabled guideline-adhering, physician-in-the-loop recommendations that allow collaboration between the care team and AI.
Collapse
Affiliation(s)
- Thanaphong Phongpreecha
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
- Department of Pathology, Stanford University, Stanford, CA, USA
| | - Marc Ghanem
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - Jonathan D Reiss
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Tomiko T Oskotsky
- Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, USA
- Department of Pediatrics, University of California, San Francisco, CA, USA
| | - Samson J Mataraso
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - Davide De Francesco
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - S Momsen Reincke
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - Camilo Espinosa
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - Philip Chung
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
| | - Taryn Ng
- Department of Pharmacy, Lucile Packard Children's Hospital, Palo Alto, CA, USA
| | - Jean M Costello
- Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, USA
| | | | - Sheila Razdan
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Keck School of Medicine of USC, Division of Neonatal and Infant Critical Care Unit, Children's Hospital Los Angeles, Los Angeles, CA, USA
| | - Feng Xie
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - Eloise Berson
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
- Department of Pathology, Stanford University, Stanford, CA, USA
| | - Yeasul Kim
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - David Seong
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - May Y Szeto
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Faith Myers
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Hannah Gu
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - John Feister
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | | | - Laura A Rose
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Lucas W Y Sin
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
| | - Boris Oskotsky
- Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, USA
| | - Jacquelyn Roger
- Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, USA
| | - Chi-Hung Shu
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - Sayane Shome
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - Liu K Yang
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
- Department of Pediatrics, Stanford University, Stanford, CA, USA
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
| | - Yuqi Tan
- Department of Pathology, Stanford University, Stanford, CA, USA
- Department of Microbiology and Immunology, Stanford University, Stanford, CA, USA
| | - Steven Levitte
- Division of Pediatric Gastroenterology, Hepatology and Nutrition, Stanford University, Stanford, CA, USA
| | - Ronald J Wong
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Brice Gaudillière
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
| | - Martin S Angst
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA
| | | | - John A Kerner
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Roberta L Keller
- Department of Pediatrics, University of California, San Francisco, CA, USA
| | - Gary M Shaw
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Karl G Sylvester
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Janene Fuerch
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Valerie Chock
- Department of Pediatrics, Stanford University, Stanford, CA, USA
| | - Shabnam Gaskari
- Department of Pharmacy, Lucile Packard Children's Hospital, Palo Alto, CA, USA
| | | | - Marina Sirota
- Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, USA
- Department of Pediatrics, University of California, San Francisco, CA, USA
| | | | - Nima Aghaeepour
- Department of Anesthesiology, Pain and Perioperative Medicine, Stanford University, Stanford, CA, USA.
- Department of Pediatrics, Stanford University, Stanford, CA, USA.
- Department of Biomedical Data Science, Stanford University, Stanford, CA, USA.
| |
Collapse
|
|
4
|
Butler JJ, Puleo J, Harrington MC, Dahmen J, Rosenbaum AJ, Kerkhoffs GMMJ, Kennedy JG. From technical to understandable: Artificial Intelligence Large Language Models improve the readability of knee radiology reports. Knee Surg Sports Traumatol Arthrosc 2024; 32:1077-1086. [PMID: 38488217 DOI: 10.1002/ksa.12133] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/01/2024] [Revised: 02/19/2024] [Accepted: 02/22/2024] [Indexed: 04/23/2024]
Abstract
PURPOSE The purpose of this study was to evaluate the effectiveness of an Artificial Intelligence-Large Language Model (AI-LLM) at improving the readability of knee radiology reports. METHODS Reports of 100 knee X-rays, 100 knee computed tomography (CT) scans and 100 knee magnetic resonance imaging (MRI) scans were retrieved. The following prompt command was inserted into the AI-LLM: 'Explain this radiology report to a patient in layman's terms in the second person:[Report Text]'. The Flesch-Kincaid reading level (FKRL) score, Flesch reading ease (FRE) score and report length were calculated for the original radiology report and the AI-LLM generated report. Any 'hallucination' or inaccurate text produced by the AI-LLM-generated report was documented. RESULTS Statistically significant improvements in mean FKRL scores in the AI-LLM generated X-ray report (12.7 ± 1.0-7.2 ± 0.6), CT report (13.4 ± 1.0-7.5 ± 0.5) and MRI report (13.5 ± 0.9-7.5 ± 0.6) were observed. Statistically significant improvements in mean FRE scores in the AI-LLM generated X-ray report (39.5 ± 7.5-76.8 ± 5.1), CT report (27.3 ± 5.9-73.1 ± 5.6) and MRI report (26.8 ± 6.4-73.4 ± 5.0) were observed. Superior FKRL scores and FRE scores were observed in the AI-LLM-generated X-ray report compared to the AI-LLM-generated CT report and MRI report, p < 0.001. The hallucination rates in the AI-LLM generated X-ray report, CT report and MRI report were 2%, 5% and 5%, respectively. CONCLUSIONS This study highlights the promising use of AI-LLMs as an innovative, patient-centred strategy to improve the readability of knee radiology reports. The clinical relevance of this study is that an AI-LLM-generated knee radiology report may enhance patients' understanding of their imaging reports, potentially reducing the responder burden placed on the ordering physicians. However, due to the 'hallucinations' produced by the AI-LLM-generated report, the ordering physician must always engage in a collaborative discussion with the patient regarding both reports and the corresponding images. LEVEL OF EVIDENCE Level IV.
Collapse
Affiliation(s)
- James J Butler
- Department of Orthopaedic Surgery, Foot and Ankle Division, NYU Langone Health, New York City, New York, USA
| | - James Puleo
- Albany Medical Center, Albany, New York, USA
| | | | - Jari Dahmen
- Department of Orthopaedic Surgery and Sports Medicine, Amsterdam Movement Sciences, Amsterdam UMC, University of Amsterdam, Location AMC, Amsterdam, The Netherlands
- Academic Center for Evidence-Based Sports Medicine, Amsterdam UMC, Amsterdam, The Netherlands
- Amsterdam Collaboration for Health and Safety in Sports, International Olympic Committee Research Center, Amsterdam UMC, Amsterdam, The Netherlands
| | | | - Gino M M J Kerkhoffs
- Department of Orthopaedic Surgery and Sports Medicine, Amsterdam Movement Sciences, Amsterdam UMC, University of Amsterdam, Location AMC, Amsterdam, The Netherlands
- Academic Center for Evidence-Based Sports Medicine, Amsterdam UMC, Amsterdam, The Netherlands
- Amsterdam Collaboration for Health and Safety in Sports, International Olympic Committee Research Center, Amsterdam UMC, Amsterdam, The Netherlands
| | - John G Kennedy
- Department of Orthopaedic Surgery, Foot and Ankle Division, NYU Langone Health, New York City, New York, USA
| |
Collapse
|