Metadata Extraction Services
We offer extraction of meta-data information and numerical values from text, tables & figures and supplementary files from diverse sources such as scientific publications, biomedical patents, and electronic health records. The metadata can be delivered as annotations inside our causality knowledge models or in any other data formats as specified by clients.
Our systematic review protocol for metadata extraction incorporates the best practices of established techniques in evidence based analyses like the PICO & PRISMA. The aggregated meta-data information can be used for meta-analyses of multiple clinical studies, to enrich existing data and knowledge with contextual information and to reach consensus among different research groups on available information.
Following is the list of meta-data entities we could extract:
-
Tissue of origin
-
Assay type
-
assay units of measure {e.g. femtogram/mL}
-
assay brand
-
assay Limit of Detection
-
Mean Concentration, SD measured
-
-
Clinical Instruments used for assessments
-
Numerical summary of clinical instruments
-
Size of study
-
Time from diagnosis at collection
-
Life stage – child or adult
-
Disease stage
-
Cohort group details
-
Biomarker direction of change
-
Drugs that target the Biomarkers
-
Range of values measured
-
Pathways
-
Genes/Proteins: upregulated or downregulated
-
And many more..
Features
Metadata annotations Extraction and Curation
Metadata annotations extraction and curation from images, tables and supplementary data files, integration with client data schema.
Data
Normalization
Data normalization and linkage with databases and ontologies.
Detection &
Reconciliation
Data inconsistency detection and reconciliation.