
pogonici - Fotolia
Artificial intelligence makes scanning medical records easier
Medical-records scanning can be costly and time-consuming for healthcare organizations that want to go paperless. Artificial intelligence can ease the process.
HIPAA requirements and clinical needs have driven the digitization of paper-based records, many of which could have ended up in a secure warehouse or shredded. However, the costs and hassle associated with scanning medical records are often overwhelming for organizations that are going paperless.
Even after paper charts are scanned and loaded into the EHR, physicians still find it difficult to interact with the documents due to the lack of adequate image classification. Fortunately, modern scanning services have introduced smarter classification capabilities to appropriately categorize documents and make it easier for physicians to find what they are looking for.
Despite the improvements in scanning hardware speeds and lower storage prices, there has not been a significant reduction in the cost of medical-records scanning. A common complaint from physicians and nurses is they still have to flip through several pages when they need something specific once the charts are scanned. This creates inefficiencies and causes physicians to question the value of scanning medical records.
The services commonly used to digitize medical records require the use of scanning applications that can detect different formats and templates using optical character recognition to classify the documents based on what it reads. However, the process is not always accurate. And, in many cases, not all paper documents adhere to the same layout, since they can originate from different sources. This can lead to charts with multiple pages being categorized into multiple groups. Hospitals and medical practices then pay higher fees to ensure documents are properly classified through manual intervention by the scanning vendor. The practice may also take on that burden and have their staff perform the extraction of the needed documents.
One approach to reducing the cost of classification some scanning vendors are adopting is to outsource classification of the scanned medical records. This helps to a certain extent, but still requires the team helping with the classification to be knowledgeable in medical terminologies to classify the content into the right groups.
In recent months, a newer method of classification has been gaining traction as an improved and cost-effective method of content classification. Scanning vendors are now advertising the use of artificial intelligence (AI) to classify healthcare documents more accurately, while also keeping costs down.
Where to store legacy medical records
Organizations that move to electronic charts typically maintain their scanned paper charts in a stand-alone document management system, or import those documents into their EHR. The decision to maintain the records in either system depends on the practice type and specialty. Specialists who may only see patients once or twice are likely to keep legacy charts outside of the EHR. Hospitals and primary care physicians or specialists who treat patients for longer periods tend to pull the scanned images into their patient health record system.
Optical character recognition which has been used for years in the scanning field, is a subset of machine learning. However, the new AI capabilities being introduced offer smart and fast classification of images based on the content of the scanned documents. Smart document scanning or capture services take advantage of optical character recognition capabilities to recognize page content.
Using AI, the system can automatically tag and classify the document accurately. AI can process the medical terminology and detect based on what is written or typed in the document to decide whether the scanned image is a referral letter, MRI report or progress note from another physician. These new automation capabilities can classify captured documents more accurately, while also minimizing costly manual intervention.
Advancements in AI now give hospitals and medical organizations that may have held back on scanning medical records access to products that provide their end users with a better experience, while also keeping costs down.