Data Extraction
MIDAS consistently delivers DECISION GRADE structured data from unstructured medical information at scale. The key to the quality of MIDAS data is cognitive review, a crucial step in which trained data specialists evaluate, contextualize and approve the data elements identified by the NLP automation, bringing depth and accuracy to the extracted medical data.
Image Received
The PDF of the patient’s medical history is supplied, either by the customer via API, or through the ReleasePoint Record Retrieval process. A high-quality PDF is the starting point for the MIDAS process.
Data from Image
MIDAS sends the medical record image through our multi-pass OCR process, which creates a high-quality data version of the original medical document.
Cognitive Curation
Technology alone cannot solve this problem.
Commercial NLP can only provide limited value because it cannot reliably decode the multitude of formats in world of image-based medical information.
Advanced AI
Following the curation process, MIDAS engages proprietary and commercial workflows, providing advanced text analysis and data augmentation through integration with Amazon Comprehend AI.
Delivery
Output is delivered through the MIDAS API in JSON format, ready for integration with automation tools, decision support platforms, medical data analytic systems, and other technology applications.