Guest post by Ivan Orekhov, CEO, ISS Art.
Modern Healthcare is stuck in two worlds: the digital and the paperwork one. Due to the advent of particularly strict measures, and thanks to the huge development of the last years in technology as well, OCR is being used more in Healthcare organizations world-wide. Advanced, powerful OCR solution will allow you to save time and be more efficient when creating, processing, and digitizing various documents; you can scan paper documents, especially medical records, for further editing and archiving or sharing with colleagues.
OCR: how does it work?
OCR stands for Optical Character Recognition, created in 1977 by Jonathan Bush and Todd Park. It is a useful and powerful technology that allows the recognition of texts inside images, scanned documents, and photos, and to convert them into machine-readable text. It is important to note that the recognized documents look just like the original, with the same layout and organization.
By helping healthcare organizations to achieve their paperless stage we noticed several benefits in regards to reduced time and cost, errors, easier availability and management of medical records; in addition, since document privacy and security is a huge concern, with electronic documents access becomes strictly controlled.
Let us have a brief look on how an OCR works. Typically, after a first analysis of the text, the software has a pre-processing phase in which it attempts to make the chosen document more clear and easier to be read. After this second operation, based on the use of an algorithm, the OCR solution will perform text recognition by choosing every single character and comparing it with a set of pattern images. The accuracy of the process may be increased if a lexicon – a list of words that may occur in the document – is added into the software as well.
OCR in Healthcare
Even if OCR in the medical field is not different from OCR used in other types of business, the workflow, clinical content varies widely among physicians, so the conversion process from paper to digital records remains slow. Recent advances in hardware and OCR technology, as we present in our product DocRecognizer, eliminated all other logistic problems, such as copying and editing.
It is important to note that some OCR solutions, not all of them still, are compliant with FHIR (Fast Healthcare Interoperability Resources) Standards Apps in exchanging Healthcare information digitally. As the data are collected, it must be done in a structured and standardized way; FHIR aims to simplify this using a rigorous mechanism for exchanging data.
There are very real benefits from using OCR in Healthcare organizations, as it could help digitalize clinical trial documents, patient reports, prescription slips or receipts, and lab notebook from clinical trials. A major focus on efficiency, cost and time reduction is the main benefit, as it allows physicians to reduce paperwork and increase care and productivity. By combining electronic data with the already existing patient’s information, records, lab results, health providers noticed an increased accuracy and velocity in the process, not to mention the cost savings that hospitals favour. For example, scanning and recognizing documents with the help of our tool DocRecognizer will lead to decreasing manager’s time up to 75%.
Powerful tools can process up to 250,000 documents everyday with an average length of 2 or 3 pages, thus enabling a broader range of data, ensuring that important patient’s information reaches doctors faster and efficiently. Also, the fact that they will be stored in a centralized cloud, allows the organization to have a precise digital archive storage, making it easier to find a given file; differently from scanned images, OCR documents allow a quicker keyword research, that would be otherwise not possible by having a simple, plain image.
Also, medical professionals can work remotely and send data online to a centralized OCR software, as the format allows a quick transmission of the documents from clinics to clinics. These upgrades have been shown to be essential, during our experience this year in COVID-19 Epidemia.
Usage of electronic health record and OCR software has been achieved in a certain number of countries, during recent years. For example, WHO lists Malaysia, where two hospitals have already gone paperless, Korea as 11 hospitals implemented fully Electronic Medical Records use, and China where several hospitals have successfully introduced some new technologies. Other countries are also developing electronic health records in one form or another with successful implementation. The USA is one of the most invested countries in digitalization, and for sure this puts it in an advantageous position.
In the last years, several studies on OCR systems in Healthcare organizations have been performed, to prove and demonstrate its accuracy. A study has been conducted in Wishard memorial Hospital’s Pediatric Outpatient Clinic, located in Indianapolis. One hundred and fifty typed paper forms were sent by the clinic staff, as result the OCR software was able to recognize the characters present more than 95% confidence 58% of the time.
Another interesting study has been performed on OCR and classification of pathology reports. The system was found to recognise scanned pathology reports up to 99.12%-character accuracy and 98.95% word accuracy.
Both studies show that OCR can be useful in regard to single patient’s records; but the great advantage is to create clinical registries, that can be unique and useful tools for national epidemiological and health services research. For example, the National Cancer Institute provides an excellent source of information of cancer epidemiology in the United States; genomic association projects rely on clinical registries to have detailed phenotypic information on many patients.
New technologies such as OCR solutions are currently in use by various organizations, but it’s becoming more and more relevant in the Medical field. Widespread use of electronic medical records could bring benefits to the Healthcare system in a variety of ways, since having patient’s data stored in a standardized form could help physicians in better care management and treatment of diseases. Based on all the above analysis, and our profound understanding of the medical domain, DocRecognizer is the preferred solution by delivering high quality values for healthcare organizations.
Biondich PG, Overhage JM, Dexter PR, Downs SM, Lemmon L, McDonald CJ. A modern optical character recognition system in a real world clinical setting: some accuracy and feasibility observations. Proc AMIA Symp. 2002;56-60.
Zuccon G, Nguyen AN, Bergheim A, Wickman S, Grayson N. The impact of OCR accuracy on automated cancer classification of pathology reports. Stud Health Technol Inform. 2012;178:250-6. PMID: 22797049.
Note: Please, if you decide to publish this article, let us know about it by sending an e-mail to our marketing specialist Anna (firstname.lastname@example.org) as our CEO can miss a letter from your side. Any feedback from you is also valuable for us.