Voice-Assisted Clinical Note Creation on Mobile Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Clinicians face challenges in capturing clinical information during conversations with patients due to the limitations of compact keypads on mobile devices, which distract from the conversation and require manual input of medical terminology, and existing voice dictation and automatic speech recognition solutions are not practical for real-time note-taking.
Innovation Solution
Utilizing natural language processing (NLP) on mobile devices to capture conversations and identify clinically relevant concepts in real-time, presenting them on the touch screen for selection, allowing clinicians to create shorthand clinical notes without distracting from patient interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If predictive typing tools are used to overcome compact keypad limitations, then typing speed is improved, but the suggested words are limited to common words and do not include medical terminology
Solution Approach 1:
The patent introduces a speech-to-text conversion system as an intermediary between the clinician's spoken medical terminology and the electronic medical record system. This mediator captures the full vocabulary range of medical terms naturally spoken by the clinician, converting speech to text without the limitations of predictive typing word lists, thereby resolving the contradiction between typing speed and vocabulary coverage.
2Ease of operation
If voice dictation is used to capture clinical information, then manual input burden is reduced, but patient-clinician rapport is compromised due to constant interruptions
Solution Approach 1:
The patent implements preliminary action by capturing the entire patient-clinician conversation through automatic speech recognition before any note creation occurs. This allows the clinician to maintain uninterrupted eye contact and engagement with the patient during the consultation, while the system simultaneously transcribes and structures the clinical information into the electronic medical record, eliminating the need for interruptions during or after the patient encounter.
3Loss of information
If automatic speech recognition transcribes the entire conversation, then complete conversation capture is achieved, but time is wasted verifying and manually correcting misunderstood words
Solution Approach 1:
The patent extracts only the clinically relevant information from the complete conversation transcript using natural language processing and information extraction algorithms. Instead of requiring the clinician to verify the entire conversation transcript, the system automatically identifies and extracts key clinical findings, diagnoses, and treatment plans, presenting a pre-processed summary that requires minimal verification, thereby resolving the contradiction between complete information capture and verification time.
Data Source
AI summary
Methods, systems, and computer-readable media are provided for facilitating the voice-assisted creation of a shorthand clinical note on a mobile or tablet device. A microphone on the device is used to capture a conversation between a clinician and a patient. Clinically-relevant concepts in the conversation are identified, extracted, and temporarily presented on the device's touch screen interface. The concepts are selectable, and upon selection, the selected concept is populated into a clinical note display area of the touch screen interface. The shorthand clinical note may be used as a memory prompt for the later creation of a more comprehensive clinical note.