About

Thomas Montine is a Stanford Medicine Professor of Pathology affiliated with the Center for Artificial Intelligence in Medicine & Imaging (AIMI). His role involves advancing research at the intersection of artificial intelligence and medical imaging, contributing to the development of innovative solutions in healthcare. As a faculty member at Stanford, he is engaged in multidisciplinary efforts to enhance medical diagnostics and treatment through AI-driven approaches, supporting the center's mission to integrate artificial intelligence into medicine and imaging.

Research topics

• Medicine
• Pathology
• Biology
• Immunology
• Internal medicine
• Neuroscience
• Genetics
• Psychology
• Computer Science
• Bioinformatics
• Cancer research
• Biochemistry
• Virology
• Computational biology
• Cell biology
• Endocrinology
• Environmental health
• Nuclear medicine

Selected publications

• Data Independent Acquisition to Inform the Development of Targeted Proteomics Assays Using a Triple Quadrupole Mass Spectrometer

  Journal of Proteome Research · 2025-05-06 · 2 citations

  articleOpen access

  Mass spectrometry based targeted proteomics methods provide a sensitive and high-throughput analysis of selected proteins. To develop a targeted bottom-up proteomics assay, peptides must be evaluated as proxies for the measurement of a protein or proteoform in a biological matrix. Candidate peptide selection typically relies on predetermined biochemical properties, data from semistochastic sampling, or empirical measurements. These strategies require extensive testing and method refinement due to the difficulties associated with prediction of the peptide response in the biological matrix of interest. Gas-phase fractionated (GPF) narrow window data-independent acquisition (DIA) aids in the development of reproducible selected reaction monitoring (SRM) assays by providing matrix-specific information on peptide detectability and quantification by mass spectrometry. To demonstrate the suitability of DIA data for selecting peptide targets, we reimplement a portion of an existing assay to measure 98 Alzheimer's disease proteins in cerebrospinal fluid (CSF). Peptides were selected from GPF-DIA based on signal intensity and reproducibility. The resulting SRM assay exhibits a quantitative precision similar to that of published data, despite the inclusion of different peptides between the assays. This workflow enables development of new assays without additional upfront data acquisition, demonstrated here through generation of a separate assay for an unrelated set of proteins in CSF from the same data set.

  PublisherOA PDFDOI

• Deep learning-based cell type profiles reveal signatures of Alzheimer’s disease resilience and resistance

  Brain · 2025-08-05 · 3 citations

  articleOpen accessSenior author

  Neurological disorders result from the complex and poorly understood contributions of many cell types. It is therefore essential to uncover mechanisms behind these disorders and identify specific therapeutic targets. Single-nucleus technologies have advanced brain disease research, but remain limited by their low nuclear transcriptional coverage, high cost and technical complexity. To address this, we applied a transformer-based deep learning model that restores cell type-specific investigation transcriptional programs from bulk RNA sequencing, significantly outperforming previous methods. This enables large-scale and cost-effective investigation of cell type-specific transcriptomes in complex and heterogeneous phenotypes such as cognitive resilience or brain resistance to Alzheimer's disease. Our analysis identified astrocytes as the major cell mediator of Alzheimer's disease resilience across cerebral cortex regions, while excitatory neurons and oligodendrocyte progenitor cells emerged as the major cell mediators of resistance, maintaining synaptic function and preserving neuron health. Finally, we show that our approach could restore the whole tissue transcriptome, offering an unbiased framework for exploring cell-specific functions beyond single-nucleus data.

  PublisherOA PDFDOI

• Frequency and Clinical Outcomes Associated With Tau Positron Emission Tomography Positivity

  JAMA · 2025-06-16 · 39 citations

  articleOpen access

  Importance: Tau positron emission tomography (PET) allows in vivo detection of neurofibrillary tangles, a core neuropathologic feature of Alzheimer disease (AD). Objective: To provide estimates of the frequency of tau PET positivity and its associated risk of clinical outcomes. Design, Setting, and Participants: Longitudinal study using data pooled from 21 cohorts, comprising a convenience sample of 6514 participants from 13 countries, collected between January 2013 and June 2024. Cognitively unimpaired individuals and patients with a clinical diagnosis of mild cognitive impairment (MCI), AD dementia, or other neurodegenerative disorders were included. Exposures: Tau PET with flortaucipir F 18, amyloid-β (Aβ) PET, and clinical examinations. Tau PET scans were visually rated as positive according to a US Food and Drug Administration- and European Medicines Agency-approved method, designed to indicate the presence of advanced neurofibrillary tangle pathology (Braak stages V-VI). Main Outcomes and Measures: Frequency of tau PET positivity and absolute risk of clinical progression (eg, progression to MCI or dementia). Results: Among the 6514 participants (mean age, 69.5 years; 50.5% female), median follow-up time ranged from 1.5 to 4.0 years. Of 3487 cognitively unimpaired participants, 349 (9.8%) were tau PET positive; the estimated frequency of tau PET positivity was less than 1% in those aged younger than 50 years, and increased from 3% (95% CI, 2%-4%) at 60 years to 19% (95% CI, 16%-24%) at 90 years. Tau PET positivity frequency estimates increased across MCI and AD dementia clinical diagnoses (43% [95% CI, 41%-46%] and 79% [95% CI, 77%-82%] at 75 years, respectively). Most tau PET-positive individuals (92%) were also Aβ PET positive. Cognitively unimpaired participants who were positive for both Aβ PET and tau PET had a higher absolute risk of progression to MCI or dementia over the following 5 years (57% [95% CI, 45%-71%]) compared with both Aβ PET-positive/tau PET-negative (17% [95% CI, 13%-22%]) and Aβ PET-negative/tau PET-negative (6% [95% CI, 5%-8%]) individuals. Among participants with MCI at the time of the tau PET scan, an Aβ PET-positive/tau PET-positive profile was associated with a 5-year absolute risk of progression to dementia of 70% (95% CI, 59%-81%). Conclusions and Relevance: In a large convenience sample, a positive tau PET scan occurred at a nonnegligible rate among cognitively unimpaired individuals, and the combination of Aβ PET positivity and tau PET positivity was associated with a high risk of clinical progression in both preclinical and symptomatic stages of AD. These findings underscore the potential of tau PET as a biomarker for staging AD pathology.

  PublisherOA PDFDOI

• Author Correction: AI-guided precision parenteral nutrition for neonatal intensive care units

  Nature Medicine · 2025-04-09 · 3 citations

  erratumOpen access
  PublisherOA PDFDOI

• TMET-23. 2-Methylglutamate Prolongs Survival in a Mouse Model of Glioblastoma

  Neuro-Oncology · 2025-11-01

  articleOpen access

  Abstract Glutamate plays a central role in glioblastoma (GBM) pathophysiology, promoting tumor proliferation, excitotoxicity, and immune suppression. Emerging evidence also highlights that glioma cells integrate into normal neural circuits, and this neuronal activity drives GBM growth and progression through formation of direct glutamatergic synapses between neurons and glioma cells. Recent work by Wawro et al. (Sci Rep, 2021) demonstrated that enantiomers of 2-methylglutamate (2MeGlu), a glutamate analog, differentially modulate brain metabolism and behavior through stereoselective interactions with glutamate transport and glutamine metabolism. However, the impact of 2MeGlu on glioblastoma progression has not been studied. We employed two orthotopic mouse models of human GBM using luciferase-labeled U251 and GBM39 patient-derived xenograft cells. Approximately nine days post-implantation, mice received daily intraperitoneal injections of a racemic mixture of 2MeGlu (500 mg/kg) or vehicle control. Survival was monitored as the primary endpoint. The control and treatment groups were followed until humane endpoints were met. Tumor burden was assessed weekly using bioluminescence imaging (BLI). U251 glioma-bearing mice exhibited median survival of 26 and 35 days in the control and treatment groups, respectively (p = 0.0165). In GBM39 orthotopic xenografts, median survival of controls was 32 days. In contrast, mice treated with racemic 2MeGlu exhibited median survival of 51 days (p = 0.0004), representing a 59% increase in overall survival compared to controls. Longitudinal BLI revealed lower tumor-associated luminescence in treated animals relative to controls, consistent with reduced intracranial tumor burden. The substantial survival benefit suggests 2MeGlu may modulate glutamate-dependent metabolic or signaling pathways critical to glioblastoma growth or microenvironmental adaptation. These findings provide the first evidence that 2MeGlu confers a survival benefit in a preclinical GBM model. Given prior evidence of stereospecific effects of 2MeGlu on glutamate-glutamine cycling, future work will assess individual enantiomers for differential therapeutic efficacy and potential mechanisms of action, including altered tumor metabolism, immune microenvironment modulation, or excitotoxic stress. These studies may help establish new therapeutic vulnerabilities in glutamate-dependent glioblastoma progression.

  PublisherOA PDFDOI

• The Parkinson's Disease Composite of Executive Functioning

  Neurology · 2025-05-12 · 1 citations

  erratumOpen access
  PublisherOA PDFDOI

• LATE-NC Stage 3: a diagnostic rubric to differentiate severe LATE-NC from FTLD-TDP

  Acta Neuropathologica · 2025-04-28 · 5 citations

  articleOpen access

  A diagnostic rubric is required to distinguish between limbic-predominant age-related TDP-43 encephalopathy neuropathologic change (LATE-NC) and frontotemporal lobar degeneration with TDP-43 inclusions (FTLD-TDP). In LATE-NC Stage 3, TDP-43 proteinopathy is present in the middle frontal gyrus (MFG), thus posing a potential diagnostic challenge in differentiating these severe LATE-NC cases from FTLD-TDP. LATE-NC Stage 3 cases and other TDP-43 proteinopathies were analyzed from the University of Kentucky (total n = 514 with TDP-43 pathology assessed), The 90+ Study at the University of California Irvine (n = 458), and the Mayo Clinic (n = 5067) brain banks. Digital pathology was used to quantify pathology burden in a select subset of cases (n = 51), complemented by a previously-described manual counting method and expert neuropathologic examinations to evaluate qualitative features such as FTLD-TDP types and subtypes of neuronal cytoplasmic inclusions (NCIs). To evaluate clinical and genetic characteristics of LATE-NC Stage 3, data were analyzed from the National Alzheimer's Coordinating Center (NACC) Neuropathology Data set and correlated with findings from the Alzheimer's Disease Genetics Consortium (ADGC). When using TDP-43 proteinopathy quantification in the MFG as a diagnostic criterion, more than 90% of cases could be classified as either LATE-NC Stage 3 or FTLD-TDP. Diagnostically challenging scenarios included a subset of FTLD-TDP Type B cases with relatively mild MFG TDP-43 pathology and a novel non-LATE-NC, non-FTLD-TDP pathologic subtype with severe MFG TDP-43 pathology. Taking these potential pitfalls into account, a classification schema was developed that could correctly diagnose all included cases. There was no difference in the Alzheimer's disease pathological load in LATE-NC Stages 2 versus 3. In genetic analyses, the GRN (rs5848) risk allele was preferentially associated with LATE-NC Stage 3, whereas TMEM106B and APOE risk-associated variants were not. In conclusion, LATE-NC Stage 3 could be differentiated reliably from FTLD-TDP and other TDP-43-opathies, based on a data-driven diagnostic rubric.

  PublisherOA PDFDOI

• Uncovering atrophy progression pattern and mechanisms in individuals at risk of Alzheimer's disease

  Brain Communications · 2025-01-01 · 3 citations

  articleOpen access

  Alzheimer's disease is associated with pre-symptomatic changes in brain morphometry and accumulation of abnormal tau and amyloid-beta pathology. Studying the development of brain changes prior to symptoms onset may lead to early diagnostic biomarkers and a better understanding of Alzheimer's disease pathophysiology. Alzheimer's disease pathology is thought to arise from a combination of protein accumulation and spreading via neural connections, but how these processes influence brain atrophy progression in the pre-symptomatic phases remains unclear. Individuals with a family history of Alzheimer's disease (FHAD) have an elevated risk of Alzheimer's disease, providing an opportunity to study the pre-symptomatic phase. Here, we used structural MRI from three databases (Alzheimer's Disease Neuroimaging Initiative, Pre-symptomatic Evaluation of Experimental or Novel Treatments for Alzheimer Disease and Montreal Adult Lifespan Study) to map atrophy progression in FHAD and Alzheimer's disease and assess the constraining effects of structural connectivity on atrophy progression. Cross-sectional and longitudinal data up to 4 years were used to perform atrophy progression analysis in FHAD and Alzheimer's disease compared with controls. PET radiotracers were also used to quantify the distribution of abnormal tau and amyloid-beta protein isoforms at baseline. We first derived cortical atrophy progression maps using deformation-based morphometry from 153 FHAD, 156 Alzheimer's disease and 116 controls with similar age, education and sex at baseline. We next examined the spatial relationship between atrophy progression and spatial patterns of tau aggregates and amyloid-beta plaques deposition, structural connectivity and neurotransmitter receptor and transporter distributions. Our results show that there were similar patterns of atrophy progression in FHAD and Alzheimer's disease, notably in the cingulate, temporal and parietal cortices, with more widespread and severe atrophy in Alzheimer's disease. Both tau and amyloid-beta pathology tended to accumulate in regions that were structurally connected in FHAD and Alzheimer's disease. The pattern of atrophy and its progression also aligned with existing structural connectivity in FHAD. In Alzheimer's disease, our findings suggest that atrophy progression results from pathology propagation that occurred earlier, on a previously intact connectome. Moreover, a relationship was found between serotonin receptor spatial distribution and atrophy progression in Alzheimer's disease. The current study demonstrates that regions showing atrophy progression in FHAD and Alzheimer's disease present with specific connectivity and cellular characteristics, uncovering some of the mechanisms involved in pre-clinical and clinical neurodegeneration.

  PublisherOA PDFDOI

• Parkinson’s disease is characterized by vitamin B6-dependent inflammatory kynurenine pathway dysfunction

  npj Parkinson s Disease · 2025-04-26 · 13 citations

  articleOpen access

  Abstract Recent studies demonstrate that Parkinson’s disease (PD) is associated with dysregulated metabolic flux through the kynurenine pathway (KP), in which tryptophan is converted to kynurenine (KYN), and KYN is subsequently metabolized to neuroactive compounds quinolinic acid (QA) and kynurenic acid (KA). Here, we used mass-spectrometry to compare blood and cerebral spinal fluid (CSF) KP metabolites between 158 unimpaired older adults and 177 participants with PD. We found increased neuroexcitatory QA/KA ratio in both plasma and CSF of PD participants associated with peripheral and cerebral inflammation and vitamin B 6 deficiency. Furthermore, increased QA tracked with CSF tau, CSF soluble TREM2 (sTREM2) and severity of both motor and non-motor PD clinical symptoms. Finally, PD patient subgroups with distinct KP profiles displayed distinct PD clinical features. These data validate the KP as a site of brain and periphery crosstalk, integrating B-vitamin status, inflammation and metabolism to ultimately influence PD clinical manifestation.

  PublisherOA PDFDOI

• AI-guided precision parenteral nutrition for neonatal intensive care units

  Nature Medicine · 2025-03-25 · 21 citations

  articleOpen access

  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.

  PublisherOA PDFDOI

Recent grants

• NIH Grant P50AG005144

  NIH · $27.4M · 2006

• Identifying the Genetic Etiology of Neuropathology for Alzheimer Disease and Related Dementias

  NIH · $3.8M · 2019–2026

• NIH Grant P01NS044282

  NIH · $6.0M · 2008

• Biomarker Core

  NIH · $42.6M · 2020–2030

• NIH Grant P50NS062684

  NIH · $30.6M · 2021

Frequent coauthors

• Cyrus P. Zabetian

  University of Washington

  366 shared

• James B. Leverenz

  344 shared

• Joseph F. Quinn

  VA Portland Health Care System

  290 shared

• Elaine R. Peskind

  241 shared

• Kathleen S. Montine

  Stanford University

  219 shared

• Paul K. Crane

  University of Washington

  191 shared

• John Q. Trojanowski

  University of Pennsylvania

  182 shared

• Jing Zhang

  Peking University Third Hospital

  164 shared