Automated Clinical Documentation Text Macro Rendering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated clinical documentation systems face challenges in accurately and efficiently processing conversational speech into formatted medical reports, particularly in identifying and rendering physician-specific text macros, which are customized boilerplate texts, and in facilitating the verification and correction of these macros by medical professionals.
Innovation Solution
The system encodes transcripts and text macros into vector representations, uses an attention mechanism to predict word-by-word reports, and differentiates text macros from regular content for accurate rendering and efficient verification, allowing for the decoupling of TM modeling and rendering to improve data efficiency and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If text macros are treated as regular content in automated clinical documentation, then the system can process general text, but it cannot accurately identify and render physician-specific customized boilerplate texts
Solution Approach 1:
The patent segments text macro processing from regular content processing by introducing sentinel tokens that mark the beginning and end of text macros. This segmentation allows the system to identify and handle physician-specific customized boilerplate texts separately from general clinical content, improving accuracy without requiring complete system redesign
Solution Approach 2:
The patent introduces sentinel tokens as intermediary elements that facilitate the identification and rendering of text macros. These special tokens act as markers that enable the automated clinical documentation system to distinguish text macros from regular content, solving the identification accuracy problem while maintaining system modularity
2Ease of operation
If the system processes all text uniformly without differentiating text macros, then processing is simpler, but verification and correction of text macros by medical professionals becomes inefficient
Solution Approach 1:
The patent employs visual differentiation (analogous to color changes) by rendering text macros with distinct formatting compared to regular content. This visual distinction allows medical professionals to quickly identify text macros during review, facilitating efficient verification and correction without requiring manual searching or complex interaction
3Adaptability or versatility
If text macros are integrated into the general transcription process, then the system can handle diverse content, but it loses the ability to efficiently manage physician-specific customized content
Solution Approach 1:
The patent applies preliminary action by pre-processing text to identify and mark text macros with sentinel tokens before the main transcription and rendering processes. This preliminary identification enables the system to efficiently manage physician-specific customized content throughout subsequent processing stages, maintaining both adaptability and productivity
Data Source
AI summary
A method, computer program product, and computer system for encoding, by a computing device, a transcript and text macros into vector representations. A word by word report may be predicted based upon, at least in part, the encoding. An attention mechanism may be queried based upon, at least in part, a decoder state. An attention distribution may be produced over an encoder output. An interpolation of the encoder output may be produced based upon, at least in part, the attention distribution. The interpolation of the encoder output may be input into a decoder for report modeling that includes text macro location and content.