Designed with interoperability in mind
- HL7 CDA
- DICOM SR
- SNOMED CT
We have spent thousands of hours with clinicians to create and refine structured reporting text editor that allows for the use of standardized terminologies, e.g. SNOMED CT while maintaining the ability to customize the template when describing unusual cases. According to surveys we conduct with our users, templates created using upmedic cover 80% reporting needs, the remaining 20% can be created by on-the-spot template customization of our Report Creator. Our templates use predefined, well-known phrases. We also take care of punctuation (adjustable), so reports look clean and professional. This helps patients to receive complete documentation that they trust and are more satisfied with the quality of medical services. Template-driven reporting acts like todo lists for clinicians making sure organs are described in the same order. Why bother? We love introducing ideas of text usability to medical documentation.
- Customizable structured templates containing most popular phrases.
- Reports based on customizable templates.
- Highlight parts of the document that are of special importance to the overall diagnosis (connotations).
- Medical forms for scales, calculations performed during creating documentation.
Standard terminology – SNOMED CT
SNOMED CT (SNOMED Clinical Terms)is the collection of clinical healthcare terminology, providing also codes, terms, synonyms and definitions used in medical documentation. It is considered to be the most comprehensive one and multilingual. SNOMED CT was created by an international non-profit standards development organization - International Health Terminology Standards Development Organisation (IHTSDO). We use SNOMED CT to automatically analyze medical documentation and provide unique reporting capabilities at the level of a single observation.
Use AI only if it is worth it
Training a machine learning model is a very tedious and time as well as money consuming task. As a result we obtain an image analysis mechanism that automates recognizing very well known pathology. However, there is a huge class of pathologies that are too rare/complex to create a model to recognize them. upmedic templates are created and refined by humans. During reporting time, templates can be used by AI models to select phrases related to what model found in the imagery, leaving the freedom for the clinician to adjust any diagnosis, provide additional findings, and correct its mistakes. upmedic uses AI/ML when we clearly see an application for it: we use NLP to support radiologists with automatic reasoning in medical documentation. Our models allow for knowledge discovery in structured medical documentation, and they can also extract ICD-10 codes, SNOMED CT concepts in unstructured documentation automatically.
BYOT Bring Your Own Templates
Seamless onboarding: send us your templates and we will add them to your upmedic account, so you can benefit from productivity improvements from day 1!