Clinical registers
Journals
Annotation & data extraction
02. Data & Reference
- Pre-trained NLP
- Advanced PDF and table handling
- Integrated software platform
- Highlighting functionality
01. AI Pre-Processing
- API reference check-up
- Data validation
03. Manual annotation
- Fully integrated workflow environment
- In-document variable selection feature
- Structured variable mapping
04. Review process
- Double human review
- Escalation & discussion workflow
- Decision tracing documentation feature
05. Release
- Final human control
- Automated data extraction
- Formatted data delivery
- Data visualizations and basic analytics
AI
Supervised AIs
Review & Performance
Graphic Design
- Keywords
- Document structure
Web development
- Basic NLPs (OpenNLP, SpiCy)
- Tokenization and basic recognition
Web development
- RNN, Transformers, LSTM
- Curated internal training data
Brand strategy
- Accuracy evaluation
- Continuous monitoring
- Superior results with individual training data
Heuristic Algorithms
Sci-Essentials AI
Clinical Trial
Use case 1-A Clinical Trials: Search automisation, pre-selection & classification
Research
Handle complexity and volume of evidence ased research
Improve Data
Significantly improve data quality and cost of delivery
Documentation
Insure gapless documentation and reproduceability
Customer
- Homeopathic preparation producer conducting research for RCTs in preparation for dossiers and online publications.
Pain points
- Search & selection required a large amount of manual work & time.
- Selection criteria require detailed documentation and data quality assurances.
Outcomes
- Improved search results while reducing manual effort by >40% .
- Implementation of double-blind human review process.
- Audit ready process documentation.
Automated database & register search
Semi-automated search result cleaning
AI supported & human controlled pre-classification
AI highlighting in titles & abstracts
Double reviewed RCT criteria assesment
Research
AI highlighting of all relevant data fields.
Improve Data
Semi-automated annotation of criteria.
Documentation
Fully documented process with human control & review
Data Extraction
Automated data extraction and database delivery
Documentation
Data cleaning & visualization
Clinical Trial
Use case 2-A Clinical Trials: Data annotation & variable extraction
Customer
- Clinical research agency conducting systematic search for RCTs in preparation for systematic reviews and potential meta analysis.
Pain points
- Search & selection required a large amount of time intensive manual work. Processes require detailed documentation and data quality assurances.
Outcomes
- Improved search results with >40% reduction in search time
- AI supported highlighting & annotation results in >60% efficiency
- Superior data quality with documented workflow
Automated detection of classification criteria
AI highlighting of annotation variables
Double review process for human controll
Software solution for document handling and variable definition
Fully automated data extraction and delivery process
01. AI Pre-Processing
- API reference check-up
- Data validation
02. Data & Reference
- Pre-trained NLP
- Advanced PDF and table handling
- Integrated software platform
- Highlighting functionality
03. Manual annotation
- Fully integrated workflow environment
- In-document variable selection feature
- Structured variable mapping
04. Review process
- Double human review
- Escalation & discussion workflow
- Decision tracing documentation feature
05. Release
- Final human control
- Automated data extraction
- Formatted data delivery
- Data visualizations and basic analytics
Use case 4-A: Regulatory affairs
Customer demand
Compliance
- Ensure compliance with rules & regulatory with regard to:
- Sources
- Inclusion
- Search & annotation strategy regarding future fillings
Documentation
- Web-based software platform to conduct systematic search & annotation
- Fulfill documentation requirements
Data quality
- Deliver proof of 100% data quality
- Insure completeness & correctness of data pool
- Implement double review process
- Execute systematic release process
Release
- Full documentation of processes and data according to regulatory requirements
- Prerequisites of regulatory fillings.
