MediViewer Modules: IMMJ OCR & Document Classification
At IMMJ we understand the need to search for key information contained within a vast amount of scanned imagery.
In general the average medical record contains 200 individual sheets of paper which when scanned in duplex presents a user with 400 individual images to work through to find the information you want.
- Although the use of bar codes and folder TAB’s are commonly applied to split large batches of images into sub sections, more often than not from our experience medical records are rarely afforded much more than six dividers meaning that batches are still difficult manage.
- Optical Character Recognition helps but does not specifically address the issues users face as the records then begin to grow when the addition of day forward scanning and information harvested from other clinical applications is also added to the electronic patient record.
- Most current systems rely heavily on complex indexing at the point of scanning, but this is not always accurate due to human error and is not cost effective. MediViewer’s document classification module removes the need to consider large scale indexing costs at the point of scanning but rather uses built in heuristic algorithms to understand the various document types.
- IMMJ’s Document Classification module was designed and developed from the ground up by our team of ex-NHS staff who have all deep experience in implementing EDRM technology and understand the challenges of having to reclassify document types post the initial scanning process.
- Simply put, Document Classification enables vast amounts of clinical documentation when scanned to be searched upon in context of the actual types of documents they are. When used in a live environment it means a user can for example retrieve selected documents from a scanned multiple volume of notes in a matter of seconds. These searches and filters can be easily saved for future use at an outpatient appointment or inpatient ward round for example.
- What we term as the “Taxonomy” that is applied in the configuration, can be easily adjusted, added to by a trained user on an ongoing basis and not set in stone. This means that the degree of classification can be a living breathing entity, and when adjusted or added to is applied upon each and every search the user makes within the system.
- Classification is not only applied to scanned imagery but in addition to all day forward document capture and digital documentation harvested into the MediViewer document repository.
- MediViewer’s Classification Module can be used to reprocess previously scanned patient records providing the path for a solution and data migration to MediViewer long after the physical patient records were scanned.