Gabriel Rosenfeld
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Gabriel Rosenfeld has a total of 15 co-authors. Here are the top co-authors by number of shared publications:
Andrei Gabrielian
Office of Cyber Infrastructure and Computational Biology, National Institutes of Allergy and Infectious Diseases, 5601 Fishers Lane, Rockville, MD 20852, USA
4 shared publicationsAlex Rosenthal
Office of Cyber Infrastructure and Computational Biology, National Institutes of Allergy and Infectious Diseases, 5601 Fishers Lane, Rockville, MD 20852, USA
3 shared publicationsDarrell E. Hurt
2 shared publicationsDarrell Hurt
Office of Cyber Infrastructure and Computational Biology, National Institutes of Allergy and Infectious Diseases, 5601 Fishers Lane, Rockville, MD 20852, USA
2 shared publicationsQinlu Wang
2 shared publications- Alex Rosenthal
Office of Cyber Infrastructure and Computational Biology, National Institute of Allergy and Infectious Diseases, National Institutes of Health
1 shared publication
Latest Publications
- Open Access
- 27 September 2023
- Published by Elsevier in European Journal of Radiology Open
- Vol. 11, 100518
Predictive capabilities of baseline radiological findings for early and late disease outcomes within sensitive and multi-drug resistant tuberculosis cases
- 27 September 2023
- Published by Elsevier in European Journal of Radiology Open
- Vol. 11, 100518
Open AccessPredictive capabilities of baseline radiological findings for early and late disease outcomes within sensitive and multi-drug resistant tuberculosis cases
- Open Access
Analysis of a Large Patient-Level Dataset to Predict Outcome of Treatment for Drug-Resistant Tuberculosis
BACKGROUND: Drug-resistant (DR) tuberculosis treatment is challenging and frequently leads to poor outcomes. An international collaboration, the National Institute of Allergy and Infectious Diseases (NIAID) TB Portals develops, maintains, and supports a multi-national database of tuberculosis cases, with an emphasis on drug-resistant tuberculosis. Patient records include clinical, radiological, genomic, and socioeconomic features. Establishing factors associated with unsuccessful treatment may help optimize treatment for the most challenging infections. METHODS: Association analysis and machine learning algorithms were applied to identify important factors associated with treatment outcome and predict the outcome for three patient cohorts, selected by drug resistance level representing 1575 patients in total. The predicted probabilities of poor treatment outcome from models were calibrated as a risk score ranging from 0 to 100 corresponding to confidence level of the model for treatment outcome. RESULTS: The features most associated with treatment success in all cohorts were body mass index (BMI), onset age, employment, education, smear-negative microscopy, and percent of abnormal volume in X-ray images, confirming previously reported findings, and identifying novel factors such as pathogen genomic markers. CONCLUSIONS: The identified features might help in establishing high-risk patients at the time of admission for tuberculosis treatment. This study integrates clinical, radiological, and pathogen genomics into a patient risk model, a way of determining risk through the application of machine learning on real-world data.
Open AccessAnalysis of a Large Patient-Level Dataset to Predict Outcome of Treatment for Drug-Resistant Tuberculosis
BACKGROUND: Drug-resistant (DR) tuberculosis treatment is challenging and frequently leads to poor outcomes. An international collaboration, the National Institute of Allergy and Infectious Diseases (NIAID) TB Portals develops, maintains, and supports a multi-national database of tuberculosis cases, with an emphasis on drug-resistant tuberculosis. Patient records include clinical, radiological, genomic, and socioeconomic features. Establishing factors associated with unsuccessful treatment may help optimize treatment for the most challenging infections. METHODS: Association analysis and machine learning algorithms were applied to identify important factors associated with treatment outcome and predict the outcome for three patient cohorts, selected by drug resistance level representing 1575 patients in total. The predicted probabilities of poor treatment outcome from models were calibrated as a risk score ranging from 0 to 100 corresponding to confidence level of the model for treatment outcome. RESULTS: The features most associated with treatment success in all cohorts were body mass index (BMI), onset age, employment, education, smear-negative microscopy, and percent of abnormal volume in X-ray images, confirming previously reported findings, and identifying novel factors such as pathogen genomic markers. CONCLUSIONS: The identified features might help in establishing high-risk patients at the time of admission for tuberculosis treatment. This study integrates clinical, radiological, and pathogen genomics into a patient risk model, a way of determining risk through the application of machine learning on real-world data.
- Open Access
Radiologist observations of chest X-rays (CXR) predict sputum smear microscopy status in TB Portals, a real-world database of tuberculosis (TB) cases
The Tuberculosis (TB) Portals is an international program of 14 countries connecting clinical, genomic, and radiologist specialists to develop an openly available repository of deidentified TB cases with multi-modal data such as case clinical characteristics, pathogen genomics, and radiomics. This real-world data resource contains over 4000 TB cases, principally drug resistant cases, with over 4000 chest X-rays (CXR) images. The scope of curated data offers a case-focused perspective into the drivers of disease incorporating the chronological context of the presented CXR data. Here, we analyze a cohort consisting of new TB cases to understand the relationship between baseline sputum microscopy status and nearby Chest X rays images. The Timika score, a lung biomarker of disease severity, was derived for each CXR using available radiologist observations. The Timika score along with the radiologist observations were compared for predictive performance of baseline sputum microscopy status. Baseline sputum microscopy status is a useful marker of pre-treatment disease severity and infectiousness. The modeling results support that both the radiologist observations as well as Timika score are predictive of smear status and that Timika score performs similarly to the top 5 radiologist features by feature selection. Moreover, inferential statistical analysis identifies the factors having the greatest association with sputum smear positivity such as presence of radiologist observations in both lungs, presence of cavity, presence of nodule, and Timika score itself. The results are consistent with prior reports showing Timika Score utility for predicting baseline sputum smear and disease status. We report testing of Timika Score on the largest, openly available real-world dataset of TB cases that can serve as a reference to explore extant and new TB disease severity scores bridging radiological, microbiological, and clinical data. To illustrate, we visualize Timika score from images in our database with other cases characteristics demonstrating that this score captures lung biomarker status consistent with known clinical risk factors.
Open AccessRadiologist observations of chest X-rays (CXR) predict sputum smear microscopy status in TB Portals, a real-world database of tuberculosis (TB) cases
The Tuberculosis (TB) Portals is an international program of 14 countries connecting clinical, genomic, and radiologist specialists to develop an openly available repository of deidentified TB cases with multi-modal data such as case clinical characteristics, pathogen genomics, and radiomics. This real-world data resource contains over 4000 TB cases, principally drug resistant cases, with over 4000 chest X-rays (CXR) images. The scope of curated data offers a case-focused perspective into the drivers of disease incorporating the chronological context of the presented CXR data. Here, we analyze a cohort consisting of new TB cases to understand the relationship between baseline sputum microscopy status and nearby Chest X rays images. The Timika score, a lung biomarker of disease severity, was derived for each CXR using available radiologist observations. The Timika score along with the radiologist observations were compared for predictive performance of baseline sputum microscopy status. Baseline sputum microscopy status is a useful marker of pre-treatment disease severity and infectiousness. The modeling results support that both the radiologist observations as well as Timika score are predictive of smear status and that Timika score performs similarly to the top 5 radiologist features by feature selection. Moreover, inferential statistical analysis identifies the factors having the greatest association with sputum smear positivity such as presence of radiologist observations in both lungs, presence of cavity, presence of nodule, and Timika score itself. The results are consistent with prior reports showing Timika Score utility for predicting baseline sputum smear and disease status. We report testing of Timika Score on the largest, openly available real-world dataset of TB cases that can serve as a reference to explore extant and new TB disease severity scores bridging radiological, microbiological, and clinical data. To illustrate, we visualize Timika score from images in our database with other cases characteristics demonstrating that this score captures lung biomarker status consistent with known clinical risk factors.
- Open Access
Current challenges in microbiome metadata collection
While the biomedical community has embraced data sharing (e.g. results, raw data) and supported establishment of large research consortia (e.g. the Human Microbiome Project) aimed to standardize the quality of important sets of microbiome sequencing data, the reusability of most microbiome data is still limited by the quality of its associated metadata. To ensure that microbiome data is indeed FAIR (Findable, Accessible, Interoperable, and Reusable), it is necessary to consider tools and approaches that make it easier to provide high-quality metadata that is fit for purpose moving forward. Such tools and approaches could be informed by current efforts to harmonize and improve the quality of extant microbiome metadata.
Open AccessCurrent challenges in microbiome metadata collection
While the biomedical community has embraced data sharing (e.g. results, raw data) and supported establishment of large research consortia (e.g. the Human Microbiome Project) aimed to standardize the quality of important sets of microbiome sequencing data, the reusability of most microbiome data is still limited by the quality of its associated metadata. To ensure that microbiome data is indeed FAIR (Findable, Accessible, Interoperable, and Reusable), it is necessary to consider tools and approaches that make it easier to provide high-quality metadata that is fit for purpose moving forward. Such tools and approaches could be informed by current efforts to harmonize and improve the quality of extant microbiome metadata.
- Open Access
- 17 March 2021
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 16 (3) , e0247906
Radiologist observations of computed tomography (CT) images predict treatment outcome in TB Portals, a real-world database of tuberculosis (TB) cases
The TB Portals program provides a publicly accessible repository of TB case data containing multi-modal information such as case clinical characteristics, pathogen genomics, and radiomics. The real-world resource contains over 3400 TB cases, primarily drug resistant cases, and CT images with radiologist annotations are available for many of these cases. The breadth of data collected offers a patient-centric view into the etiology of the disease including the temporal context of the available imaging information. Here, we analyze a cohort of new TB cases with available radiologist observations of CTs taken around the time of initial registration of the case into the database and with available follow up to treatment outcome of cured or died. Follow up ranged from 5 weeks to a little over 2 years consistent with the longest treatment regimens for drug resistant TB and cases were registered within the years 2008 to 2019. The radiologist observations were incorporated into machine learning pipelines to test various class balancing strategies on the performance of predictive models. The modeling results support that the radiologist observations are predictive of treatment outcome. Moreover, inferential statistical analysis identifies markers of TB disease spread as having an association with poor treatment outcome including presence of radiologist observations in both lungs, swollen lymph nodes, multiple cavities, and large cavities. While the initial results are promising, further data collection is needed to incorporate methods to mitigate potential confounding such as including additional model covariates or matching cohorts on covariates of interest (e.g. demographics, BMI, comorbidity, TB subtype, etc.). Nonetheless, the preliminary results highlight the utility of the resource for hypothesis generation and exploration of potential biomarkers of TB disease severity and support these additional data collection efforts.
- 17 March 2021
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 16 (3) , e0247906
Open AccessRadiologist observations of computed tomography (CT) images predict treatment outcome in TB Portals, a real-world database of tuberculosis (TB) cases
The TB Portals program provides a publicly accessible repository of TB case data containing multi-modal information such as case clinical characteristics, pathogen genomics, and radiomics. The real-world resource contains over 3400 TB cases, primarily drug resistant cases, and CT images with radiologist annotations are available for many of these cases. The breadth of data collected offers a patient-centric view into the etiology of the disease including the temporal context of the available imaging information. Here, we analyze a cohort of new TB cases with available radiologist observations of CTs taken around the time of initial registration of the case into the database and with available follow up to treatment outcome of cured or died. Follow up ranged from 5 weeks to a little over 2 years consistent with the longest treatment regimens for drug resistant TB and cases were registered within the years 2008 to 2019. The radiologist observations were incorporated into machine learning pipelines to test various class balancing strategies on the performance of predictive models. The modeling results support that the radiologist observations are predictive of treatment outcome. Moreover, inferential statistical analysis identifies markers of TB disease spread as having an association with poor treatment outcome including presence of radiologist observations in both lungs, swollen lymph nodes, multiple cavities, and large cavities. While the initial results are promising, further data collection is needed to incorporate methods to mitigate potential confounding such as including additional model covariates or matching cohorts on covariates of interest (e.g. demographics, BMI, comorbidity, TB subtype, etc.). Nonetheless, the preliminary results highlight the utility of the resource for hypothesis generation and exploration of potential biomarkers of TB disease severity and support these additional data collection efforts.
- Open Access
Estimating the scale of biomedical data generation using text mining
While the impact of biomedical research has traditionally been measured using bibliographic metrics such as citation or journal impact factor, the data itself is an output which can be directly measured to provide additional context about a publication’s impact. Data are a resource that can be repurposed and reused providing dividends on the original investment used to support the primary work. Moreover, it is the cornerstone upon which a tested hypothesis is rejected or accepted and specific scientific conclusions are reached. Understanding how and where it is being produced enhances the transparency and reproducibility of the biomedical research enterprise. Most biomedical data are not directly deposited in data repositories and are instead found in the publication within figures or attachments making it hard to measure. We attempted to address this challenge by using recent advances in word embedding to identify the technical and methodological features of terms used in the free text of articles’ methods sections. We created term usage signatures for five types of biomedical research data, which were used in univariate clustering to correctly identify a large fraction of positive control articles and a set of manually annotated articles where generation of data types could be validated. The approach was then used to estimate the fraction of PLOS articles generating each biomedical data type over time. Out of all PLOS articles analyzed (n = 129,918), ~7%, 19%, 12%, 18%, and 6% generated flow cytometry, immunoassay, genomic microarray, microscopy, and high-throughput sequencing data. The estimate portends a vast amount of biomedical data being produced: in 2016, if other publishers generated a similar amount of data then roughly 40,000 NIH-funded research articles would produce ~56,000 datasets consisting of the five data types we analyzed. One Sentence Summary: Application of a word-embedding model trained on the methods sections of research articles allows for estimation of the production of diverse biomedical data types using text mining.
Open AccessEstimating the scale of biomedical data generation using text mining
While the impact of biomedical research has traditionally been measured using bibliographic metrics such as citation or journal impact factor, the data itself is an output which can be directly measured to provide additional context about a publication’s impact. Data are a resource that can be repurposed and reused providing dividends on the original investment used to support the primary work. Moreover, it is the cornerstone upon which a tested hypothesis is rejected or accepted and specific scientific conclusions are reached. Understanding how and where it is being produced enhances the transparency and reproducibility of the biomedical research enterprise. Most biomedical data are not directly deposited in data repositories and are instead found in the publication within figures or attachments making it hard to measure. We attempted to address this challenge by using recent advances in word embedding to identify the technical and methodological features of terms used in the free text of articles’ methods sections. We created term usage signatures for five types of biomedical research data, which were used in univariate clustering to correctly identify a large fraction of positive control articles and a set of manually annotated articles where generation of data types could be validated. The approach was then used to estimate the fraction of PLOS articles generating each biomedical data type over time. Out of all PLOS articles analyzed (n = 129,918), ~7%, 19%, 12%, 18%, and 6% generated flow cytometry, immunoassay, genomic microarray, microscopy, and high-throughput sequencing data. The estimate portends a vast amount of biomedical data being produced: in 2016, if other publishers generated a similar amount of data then roughly 40,000 NIH-funded research articles would produce ~56,000 datasets consisting of the five data types we analyzed. One Sentence Summary: Application of a word-embedding model trained on the methods sections of research articles allows for estimation of the production of diverse biomedical data types using text mining.