Jonathan D. Reiss

Jonathan D. Reiss

Auto-generated profile, based on publicly available information

Jonathan D. Reiss

Department of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA

Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
16 January 2025 in Nature Machine Intelligence
Verified
Journal Article
Open access
PDF file
SMSamson J. Mataraso
Samson J. Mataraso
CECamilo A. Espinosa
Camilo A. Espinosa
DSDavid Seong
David Seong
SRS. Momsen Reincke
S. Momsen Reincke
EBEloise Berson
Eloise Berson
NANima Aghaeepour
Nima Aghaeepour
Omics studies produce a large number of measurements, enabling the development, validation and interpretation of systems-level biological models. Large cohorts are required to power these complex models; yet, the cohort size remains limited due to clinical and budgetary constraints. We introduce clinical and omics multimodal analysis enhanced with transfer learning (COMET), a machine learning framework that incorporates large, observational electronic health record databases and transfer learning to improve the analysis of small datasets from omics studies. By pretraining on electronic health record data and adaptively blending both early and late fusion strategies, COMET overcomes the limitations of existing multimodal machine learning methods. Using two independent datasets, we showed that COMET improved the predictive modelling performance and biological discovery compared with the analysis of omics data with traditional methods. By incorporating electronic health record data into omics analyses, COMET enables more precise patient classifications, beyond the simplistic binary reduction to cases and controls. This framework can be broadly applied to the analysis of multimodal omics studies and reveals more powerful biological insights from limited cohort sizes.

15 February 2023 in Science Translational Medicine
Verified
Journal Article
Closed access
DFDavide De Francesco
Davide De Francesco
JRJonathan D. Reiss
Jonathan D. Reiss
JRJacquelyn Roger
Jacquelyn Roger
ATAlice S. Tang
Alice S. Tang
ACAlan L. Chang
Alan L. Chang
NANima Aghaeepour
Nima Aghaeepour
Although prematurity is the single largest cause of death in children under 5 years of age, the current definition of prematurity, based on gestational age, lacks the precision needed for guiding care decisions. Here, we propose a longitudinal risk assessment for adverse neonatal outcomes in newborns based on a deep learning model that uses electronic health records (EHRs) to predict a wide range of outcomes over a period starting shortly before conception and ending months after birth. By linking the EHRs of the Lucile Packard Children’s Hospital and the Stanford Healthcare Adult Hospital, we developed a cohort of 22,104 mother-newborn dyads delivered between 2014 and 2018. Maternal and newborn EHRs were extracted and used to train a multi-input multitask deep learning model, featuring a long short-term memory neural network, to predict 24 different neonatal outcomes. An additional cohort of 10,250 mother-newborn dyads delivered at the same Stanford Hospitals from 2019 to September 2020 was used to validate the model. Areas under the receiver operating characteristic curve at delivery exceeded 0.9 for 10 of the 24 neonatal outcomes considered and were between 0.8 and 0.9 for 7 additional outcomes. Moreover, comprehensive association analysis identified multiple known associations between various maternal and neonatal features and specific neonatal outcomes. This study used linked EHRs from more than 30,000 mother-newborn dyads and would serve as a resource for the investigation and prediction of neonatal outcomes. An interactive website is available for independent investigators to leverage this unique dataset: https://maternal-child-health-associations.shinyapps.io/shiny_app/ .

25 March 2025 in Nature Medicine
Verified
Journal Article
Open access
PDF file
TPThanaphong Phongpreecha
Thanaphong Phongpreecha
MGMarc Ghanem
Marc Ghanem
JRJonathan D. Reiss
Jonathan D. Reiss
TOTomiko T. Oskotsky
Tomiko T. Oskotsky
SMSamson J. Mataraso
Samson J. Mataraso
NANima Aghaeepour
Nima Aghaeepour
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.

10 October 2019 in Biomedicine Hub
Verified
Journal Article
Open access
PDF file
JRJonathan Reiss
Jonathan Reiss
MSMridu Sinha
Mridu Sinha
JGJeffrey Gold
Jeffrey Gold
JBJulie Bykowski
Julie Bykowski
SLShelley M. Lawrence
Shelley M. Lawrence
Introduction: Accurately diagnosing and treating infants with mild forms of hypoxic ischemic encephalopathy (HIE) is important, as the majority of neonates with signs and symptoms of HIE after birth do not meet clinical criteria for moderate or severe disease. Emerging evidence, however, suggests that infants with mild HIE (mHIE) have an increased risk for neurodevelopmental impairment (NDI). Methods: This retrospective descriptive study examined all inborn infants ≥35 week’s gestational age at a single, level III neonatal intensive care unit (NICU) in California between January 1, 2012, and December 31, 2015. International Classification of Diseases codes were used as a proxy to identify neonates with mHIE but who did not receive therapeutic hypothermia (TH). Short- and long-term neurodevelopmental outcomes were documented, including abnormal (1) brain magnetic resonance imaging within 10 days of birth suggestive of HIE, (2) electroencephalogram with electrographic seizures, (3) neurologic discharge examination, or (4) NDI following NICU discharge. Results: Over the 4-year study period, 25 infants met inclusion criteria. Eight of 25 (32%) infants demonstrated neurologic impairment, defined by an abnormality in at least one of the four categories. The remaining 17 infants were without documented evidence for adverse outcomes. Conclusion: Our results indicate that children with mHIE are at significant risk for neurologic injury and may benefit from more aggressive interventions. Further prospective studies should be completed to determine the efficacy of TH in this specific patient population.

28 August 2022 in Nutrients
Verified
Journal Article
Open access
PDF file
Chengyin YeChengyin Ye
Chengyin Ye
JWJinghua Wu
Jinghua Wu
JRJonathan D. Reiss
Jonathan D. Reiss
TSTiffany J. Sinclair
Tiffany J. Sinclair
DSDavid K. Stevenson
David K. Stevenson
KSKarl Sylvester
Karl Sylvester
Objective: To assess the longitudinal metabolic patterns during the evolution of bronchopulmonary dysplasia (BPD) development. Methods: A case-control dataset of preterm infants (<32-week gestation) was obtained from a multicenter database, including 355 BPD cases and 395 controls. A total of 72 amino acid (AA) and acylcarnitine (AC) variables, along with infants’ calorie intake and growth outcomes, were measured on day of life 1, 7, 28, and 42. Logistic regression, clustering methods, and random forest statistical modeling were utilized to identify metabolic variables significantly associated with BPD development and to investigate their longitudinal patterns that are associated with BPD development. Results: A panel of 27 metabolic variables were observed to be longitudinally associated with BPD development. The involved metabolites increased from 1 predominant different AC by day 7 to 19 associated AA and AC compounds by day 28 and 16 metabolic features by day 42. Citrulline, alanine, glutamate, tyrosine, propionylcarnitine, free carnitine, acetylcarnitine, hydroxybutyrylcarnitine, and most median-chain ACs (C5:C10) were the most associated metabolites down-regulated in BPD babies over the early days of life, whereas phenylalanine, methionine, and hydroxypalmitoylcarnitine were observed to be up-regulated in BPD babies. Most calorie intake and growth outcomes revealed similar longitudinal patterns between BPD cases and controls over the first 6 weeks of life, after gestational adjustment. When combining with birth weight, the derived metabolic-based discriminative model observed some differences between those with and without BPD development, with c-statistics of 0.869 and 0.841 at day 7 and 28 of life on the test data. Conclusions: The metabolic panel we describe identified some metabolic differences in the blood associated with BPD pathogenesis. Further work is needed to determine whether these compounds could facilitate the monitoring and/or investigation of early-life metabolic status in the lung and other tissues for the prevention and management of BPD.

11 March 2015 in Leukemia & Lymphoma
Verified
Journal Article
Closed access
JRJonathan Reiss
Jonathan Reiss
BLBrian Link
Brian Link
JRJia Ruan
Jia Ruan
RFRichard Furman
Richard Furman
MCMorton Coleman
Morton Coleman
PMPeter Martin
Peter Martin
Existing data suggest that myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) peak in incidence 5-10 years following exposure to ionizing radiation, while most publications report less than 5 years of follow-up after radioimmunotherapy (RIT). We report the rate of secondary MDS/AML among 60 patients treated with two front-line sequential chemotherapy-RIT trials with over 11 years of follow-up. Among 35 patients evaluated after fludarabine-RIT and 25 patients evaluated after CVP (cyclophosphamide, vincristine, prednisone)-RIT treatment, the crude, cumulative and Kaplan-Meier rates of MDS/AML at 11 years of follow-up from the combined trials were 0.12/person, 0.010/person-year and 14% (95% confidence interval [CI] 5-24%), respectively. Additionally, we found that patients treated with RIT consolidation appear to have durable remissions but that relapses after 10 years do occur. Studies of efficacy and secondary MDS/AML that report fewer than 10 years of follow-up likely underestimate risk.