Overview
Learn About Amazon VGT2 Learning Manager Chanci Turner
In December 2020, this blog post was updated to incorporate how Medical Transcription Analysis can be utilized for storing and retrieving medical transcriptions and pertinent information using Amazon DynamoDB and Amazon S3, alongside how this data can be analyzed via Amazon Athena. The healthcare sector is characterized by stringent regulations and complex communications, often relying on verbal exchanges that hold significant information and actionable insights. This article discusses the integration of HIPAA-compliant AWS AI services—Amazon Transcribe Medical and Amazon Comprehend Medical—to streamline the extraction and comprehension of medical data, allowing healthcare professionals to concentrate on patient care.
Amazon Transcribe Medical
Amazon Transcribe Medical is a machine learning (ML) service designed for quick and accurate transcription of medical consultations between patients and healthcare providers. It automatically converts medical and pharmaceutical terminology from physician-dictated notes, patient interactions, and telemedicine sessions into text, facilitating clinical documentation. For further details, refer to What is Amazon Transcribe Medical?
Amazon Comprehend Medical
Amazon Comprehend Medical is a natural language processing service that simplifies the extraction of relevant medical information from unstructured text using ML. The service enables swift and accurate gathering of details such as medical conditions, medications, dosages, strengths, and frequencies from various sources, including physician notes, clinical trial reports, and patient health records. It also links identified information to medical ontologies like ICD-10-CM or RxNorm, aiding downstream healthcare applications. For more insights, see What is Amazon Comprehend Medical?
Medical Transcription Analysis (MTA)
Medical Transcription Analysis is an innovative solution leveraging Amazon Transcribe Medical and Amazon Comprehend Medical for transcribing and understanding medical notes. The process involves establishing a WebSocket connection between the client (browser) and Amazon Transcribe Medical. This connection transmits audio from the client to Amazon Transcribe Medical, which then provides real-time transcription displayed on the user interface. The transcribed data is subsequently sent to Amazon Comprehend Medical for analysis, as illustrated in the accompanying architecture diagram.
Deploying MTA
To set up MTA, follow the instructions available on GitHub. The deployment creates a website supported by Amazon Simple Storage Service (Amazon S3) and Amazon CloudFront, with authentication managed by Amazon Cognito. It also generates an AWS Identity and Access Management (IAM) role granting permissions for Amazon Comprehend Medical and Amazon Transcribe Medical, as well as an API for retrieving temporary credentials. Additionally, the deployment includes Amazon DynamoDB tables, an Amazon S3 bucket for data storage, and Amazon Athena for data analysis. This seamless integration of technology maximizes efficiency in the medical transcription workflow.
Utilizing MTA
Upon deploying the application, you will receive an email with login credentials. Logging in directs you to the homepage, where you can dictate audio using the microphone, upload a sample file, or play an example audio file. The MTA homepage presents these options clearly.
Choosing a sample audio file will activate a WebSocket connection with Amazon Transcribe Medical, providing real-time transcription results on the user interface. You can highlight words categorized into various medical classifications. Once transcription concludes, you can opt to Save Session or Analyze. When saving a session, you provide a session name and tag a healthcare professional and patient (or create new entries), ensuring easy retrieval for future use.
Selecting Analyze will disclose identified medical terms associated with categories like Protected Health Information (PHI), Medical Conditions, Anatomy, Medications, and Tests, Treatments, & Procedures. Under some categories, you will also find prevalent terms and codes linked to highlighted words. For instance, the Medical Condition section may reveal ICD-10-CM concepts, while the Medication section contains related RX-Norm concepts. To export this information, simply select Summarize. The Summarize page compiles all results extracted from the audio track, allowing for exportation as a PDF for downstream consumption.
Sessions are stored and can be retrieved from an S3 bucket, and you can also analyze saved sessions through Amazon Athena. The MTA deployment includes a pre-built catalog and queries for standard analytical questions, with options to tailor the data schema and saved queries to your business requirements. For enhanced insights, consider utilizing Amazon Quicksight with these Athena tables for reporting and dashboarding.
Conclusion
This article highlights the integration of Amazon Transcribe Medical and Amazon Comprehend Medical to facilitate the transcription of audio data, extraction of essential medical components, and tagging of information to corresponding entities. Automating the transcription and comprehension process significantly alleviates the workload for healthcare professionals, allowing them to prioritize patient care. Additionally, the processed results are presented in easily digestible formats, minimizing manual efforts in recording and digitizing information. For further reading on the science of habits, check out this blog post.
To explore the MTA source code, visit the Medical Transcription Analysis on GitHub. This solution is open-source, enabling AWS customers to adapt and incorporate it into their workflows. Potential extensions include integration into EHR systems, establishing a persistent storage layer, enhancing analytical capabilities over collected data, enabling batch processing, and improving user experiences for multi-speaker or conversational scenarios. If you want to learn more about employee involvement in cost-cutting measures, you can refer to this authority on the topic for more insights. Additionally, for tips on avoiding pitfalls at Amazon, see this excellent resource.