Timothy Bergquist

Timothy Bergquist

Auto-generated profile, based on publicly available information

Timothy Bergquist

Institute for Genomic Health, Icahn School of Medicine at Mount Sinai, New York, NY, USA

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
17 July 2023 in Nature Machine Intelligence
Verified
Journal Article
Open access
PDF file
AKAlexandros Karargyris
Alexandros Karargyris
RURenato Umeton
Renato Umeton
MSMicah J. Sheller
Micah J. Sheller
AAAlejandro Aristizabal
Alejandro Aristizabal
JGJohnu George
Johnu George
PMPeter Mattson
Peter Mattson
Medical artificial intelligence (AI) has tremendous potential to advance healthcare by supporting and contributing to the evidence-based practice of medicine, personalizing patient treatment, reducing costs, and improving both healthcare provider and patient experience. Unlocking this potential requires systematic, quantitative evaluation of the performance of medical AI models on large-scale, heterogeneous data capturing diverse patient populations. Here, to meet this need, we introduce MedPerf, an open platform for benchmarking AI models in the medical domain. MedPerf focuses on enabling federated evaluation of AI models, by securely distributing them to different facilities, such as healthcare organizations. This process of bringing the model to the data empowers each facility to assess and verify the performance of AI models in an efficient and human-supervised process, while prioritizing privacy. We describe the current challenges healthcare and AI communities face, the need for an open platform, the design philosophy of MedPerf, its current implementation status and real-world deployment, our roadmap and, importantly, the use of MedPerf with multiple international institutions within cloud-based technology and on-premises scenarios. Finally, we welcome new contributions by researchers and organizations to further strengthen MedPerf as an open benchmarking platform.

10 July 2021 in Journal of Clinical Oncology
Verified
Journal Article
Closed access
NSNoha Sharafeldin
Noha Sharafeldin
BBBenjamin Bates
Benjamin Bates
QSQianqian Song
Qianqian Song
VMVithal Madhira
Vithal Madhira
YYYao Yan
Yao Yan
UTUmit Topaloglu
Umit Topaloglu
PURPOSE: Variation in risk of adverse clinical outcomes in patients with cancer and COVID-19 has been reported from relatively small cohorts. The NCATS’ National COVID Cohort Collaborative (N3C) is a centralized data resource representing the largest multicenter cohort of COVID-19 cases and controls nationwide. We aimed to construct and characterize the cancer cohort within N3C and identify risk factors for all-cause mortality from COVID-19.METHODS: We used 4,382,085 patients from 50 US medical centers to construct a cohort of patients with cancer. We restricted analyses to adults ≥ 18 years old with a COVID-19–positive or COVID-19–negative diagnosis between January 1, 2020, and March 25, 2021. We followed N3C selection of an index encounter per patient for analyses. All analyses were performed in the N3C Data Enclave Palantir platform.RESULTS: A total of 398,579 adult patients with cancer were identified from the N3C cohort; 63,413 (15.9%) were COVID-19–positive. Most common represented cancers were skin (13.8%), breast (13.7%), prostate (10.6%), hematologic (10.5%), and GI cancers (10%). COVID-19 positivity was significantly associated with increased risk of all-cause mortality (hazard ratio, 1.20; 95% CI, 1.15 to 1.24). Among COVID-19–positive patients, age ≥ 65 years, male gender, Southern or Western US residence, an adjusted Charlson Comorbidity Index score ≥ 4, hematologic malignancy, multitumor sites, and recent cytotoxic therapy were associated with increased risk of all-cause mortality. Patients who received recent immunotherapies or targeted therapies did not have higher risk of overall mortality.CONCLUSION: Using N3C, we assembled the largest nationally representative cohort of patients with cancer and COVID-19 to date. We identified demographic and clinical factors associated with increased all-cause mortality in patients with cancer. Full characterization of the cohort will provide further insights into the effects of COVID-19 on cancer outcomes and the ability to continue specific cancer treatments.

10 March 2025 in Genetics in Medicine
Verified
Journal Article
Open access
TBTimothy Bergquist
Timothy Bergquist
SSSarah L. Stenton
Sarah L. Stenton
ENEmily A.W. Nadeau
Emily A.W. Nadeau
ABAlicia B. Byrne
Alicia B. Byrne
MGMarc S. Greenblatt
Marc S. Greenblatt
TPTina Pesaran
Tina Pesaran
We previously developed an approach to calibrate computational tools for clinical variant classification, updating recommendations for the reliable use of variant impact predictors to provide evidence strength up to Strong. A new generation of tools using distinctive approaches has since been released, and these methods must be independently calibrated for clinical application. Using our local posterior probability-based calibration and our established data set of ClinVar pathogenic and benign variants, we determined the strength of evidence provided by three new tools (AlphaMissense, ESM1b, VARITY) and calibrated scores meeting each evidence strength. All three tools reached the Strong level of evidence for variant pathogenicity and Moderate for benignity, though sometimes for few variants. Compared to previously recommended tools, these yielded at best only modest improvements in the tradeoffs of evidence strength and false positive predictions. At calibrated thresholds, three new computational predictors provided evidence for variant pathogenicity at similar strength to the four previously recommended predictors (and comparable with functional assays for some variants). This calibration broadens the scope of computational tools for application in clinical variant classification. Their new approaches offer promise for future advancement of the field.

14 March 2022 in Journal of Clinical Oncology
Verified
Journal Article
Open access
QSQianqian Song
Qianqian Song
BBBenjamin Bates
Benjamin Bates
YSYu Raymond Shao
Yu Raymond Shao
FHFang-Chi Hsu
Fang-Chi Hsu
FLFeifan Liu
Feifan Liu
UTUmit Topaloglu
Umit Topaloglu
PURPOSE: To provide real-world evidence on risks and outcomes of breakthrough COVID-19 infections in vaccinated patients with cancer using the largest national cohort of COVID-19 cases and controls.METHODS: We used the National COVID Cohort Collaborative (N3C) to identify breakthrough infections between December 1, 2020, and May 31, 2021. We included patients partially or fully vaccinated with mRNA COVID-19 vaccines with no prior SARS-CoV-2 infection record. Risks for breakthrough infection and severe outcomes were analyzed using logistic regression.RESULTS: A total of 6,860 breakthrough cases were identified within the N3C-vaccinated population, among whom 1,460 (21.3%) were patients with cancer. Solid tumors and hematologic malignancies had significantly higher risks for breakthrough infection (odds ratios [ORs] = 1.12, 95% CI, 1.01 to 1.23 and 4.64, 95% CI, 3.98 to 5.38) and severe outcomes (ORs = 1.33, 95% CI, 1.09 to 1.62 and 1.45, 95% CI, 1.08 to 1.95) compared with noncancer patients, adjusting for age, sex, race/ethnicity, smoking status, vaccine type, and vaccination date. Compared with solid tumors, hematologic malignancies were at increased risk for breakthrough infections (adjusted OR ranged from 2.07 for lymphoma to 7.25 for lymphoid leukemia). Breakthrough risk was reduced after the second vaccine dose for all cancers (OR = 0.04; 95% CI, 0.04 to 0.05), and for Moderna's mRNA-1273 compared with Pfizer's BNT162b2 vaccine (OR = 0.66; 95% CI, 0.62 to 0.70), particularly in patients with multiple myeloma (OR = 0.35; 95% CI, 0.15 to 0.72). Medications with major immunosuppressive effects and bone marrow transplantation were strongly associated with breakthrough risk among the vaccinated population.CONCLUSION: Real-world evidence shows that patients with cancer, especially hematologic malignancies, are at higher risk for developing breakthrough infections and severe outcomes. Patients with vaccination were at markedly decreased risk for breakthrough infections. Further work is needed to assess boosters and new SARS-CoV-2 variants.

1 June 2021 in Diagnostic Microbiology and Infectious Disease
Verified
Journal Article
Closed access
SGSaurabh Gombar
Saurabh Gombar
TBTimothy Bergquist
Timothy Bergquist
VPVikas Pejaver
Vikas Pejaver
NHNoah E. Hammarlund
Noah E. Hammarlund
KMKanagavel Murugesan
Kanagavel Murugesan
NBNiaz Banaei
Niaz Banaei

8 July 2020 in Journal of the American Medical Informatics Association
Verified
Journal Article
Open access
PDF file
TBTimothy Bergquist
Timothy Bergquist
YYYao Yan
Yao Yan
TSThomas Schaffter
Thomas Schaffter
TYThomas Yu
Thomas Yu
VPVikas Pejaver
Vikas Pejaver
SMSean Mooney
Sean Mooney
Objective: The development of predictive models for clinical application requires the availability of electronic health record (EHR) data, which is complicated by patient privacy concerns. We showcase the “Model to Data” (MTD) approach as a new mechanism to make private clinical data available for the development of predictive models. Under this framework, we eliminate researchers’ direct interaction with patient data by delivering containerized models to the EHR data. Materials and Methods: We operationalize the MTD framework using the Synapse collaboration platform and an on-premises secure computing environment at the University of Washington hosting EHR data. Containerized mortality prediction models developed by a model developer, were delivered to the University of Washington via Synapse, where the models were trained and evaluated. Model performance metrics were returned to the model developer. Results: The model developer was able to develop 3 mortality prediction models under the MTD framework using simple demographic features (area under the receiver-operating characteristic curve [AUROC], 0.693), demographics and 5 common chronic diseases (AUROC, 0.861), and the 1000 most common features from the EHR’s condition/procedure/drug domains (AUROC, 0.921). Discussion: We demonstrate the feasibility of the MTD framework to facilitate the development of predictive models on private EHR data, enabled by common data models and containerization software. We identify challenges that both the model developer and the health system information technology group encountered and propose future efforts to improve implementation. Conclusions: The MTD framework lowers the barrier of access to EHR data and can accelerate the development and evaluation of clinical prediction models.