I. Natural Language Processing
When it comes to standardized and structured data computers are a man’s best friend. They can read and process it faster, better and more accurate than any average human being. Unstructured data, or human language, on the other hand, is something incomprehensible to computers because there are no standardized techniques to process it.
Natural Language Processing (NLP) is a sub-field of computer science, artificial intelligence, and computational linguistics focused on enabling computers to understand and process human communication and get them closer to a human-level understanding of language. And lately, the application of NLP has been increasingly large. Think of Google Assistant, Alexa, word suggestions when typing a message, machine translation, spam filtering, chat boxes - these are just a few everyday examples of NLP.
Like any other aspect of machine learning, NLP can impact one’s business in a positive way by improving the efficiency of the working process. It can be used to extract crucial patterns within different forms of text and other unstructured data that can be of significant value for one’s company. For example:
- Brand Sentiment Analysis: Understanding the emotional tone of consumer social posts in the public domain, knowing the trending opinion and having a real-time view of the customer’s pulse is a critical element of brand marketing. NLP helps to derive these insights from textual data.
- Recruitment: Semantic search of resumes to filter the best fit is far more powerful than keyword match. NLP is at the backbone of such targeted selection and recruitment methods.
- Media and Publishing: Publishing companies deliver news and content after aggregating and curating from a variety of sources. The process of aggregation is far more precise for the reader’s preferences with NLP-based selection.
- Financial Markets: With market conditions shifting daily, analysts need real-time and relevant content at their fingertips, and NLP can be an answer to provide such content more efficiently and accurately to influence timely decision making.
- Call Centers: High volumes of consumer interaction creates the need for a critical capability to prioritize which tasks to act upon first. Using voice to text, NLP and machine learning can more quickly deliver insights to the most important customer inquiries.
- E-commerce: In commerce-oriented websites and apps, NLP supports meaning-based search, allowing shoppers to search for items in their own language while still producing relevant results, even if the search terms do not directly match keywords in product records. As opposed to pure keyword-based searches, NLP strikes the critical balance between delivering a variety of highly on-target search results without inundating customers with results they’re not looking for.
II. Amazon Comprehend
So no wonder in 2017 Amazon announced its new service Amazon Comprehend - an element of the Amazon Web Services infrastructure. It uses NLP to extract insights about the content of documents in English, Spanish, French, German, Italian and Portuguese. Furthermore, it can identify natural language terms and classify text which is specialized by a certain team, business or industry. It does this by:
- returning a list of entities, such as people, places, and locations, identified in a document;
- extracting key phrases that appear in a document;
- identifying the dominant language in a document. (Amazon Comprehend can identify 100 languages);
- determining the emotional sentiment of a document may it be positive, neutral, negative, or mixed;
- parsing each word in your document and thus determining the part of speech for the word.
You can use Amazon Comprehend to examine the content of a collection of documents to determine common themes. For example, you can give Amazon Comprehend a collection of news articles, and it will determine the subjects, such as sports, politics, or entertainment. Moreover, it can tell you what customers think of your products, whether they feel positive, negative, neutral, or mixed about it. Last but not least it can help you discover the topics that your customers are talking about on your forums and message boards and let you know how they feel about it.
III. Amazon Comprehend Medical
In 2018 Amazon made some bold steps into health care one of which was Amazon Medical Comprehend - a logical extension of its NLP platform. Most of the historical patient data is still stored as unstructured text. This includes all kinds of medical notes such as prescriptions, observation, and administration of drugs and therapies, test results, x-rays, CT-scans, surgical history, immunization history, family history, audio interview transcripts, printed pathology and radiology reports. It may take hours or even days for doctors and other professionals to sort through such an enormous quantity of data when trying to extract meaningful information. Amazon Comprehend Medical combines text analysis and machine learning to read patient records. Once those records are digitized and uploaded to Comprehend Medical, it picks out and organizes information about diagnoses, treatments, medication dosage and symptoms. Extracting and correlating patterns from patient records can help healthcare providers and researchers save money, make treatment decisions faster and manage clinical trials better. For example:
- Amazon Comprehend Medical is expected to give health care providers and health researchers the ability to share and analyze data sets that were previously not practical to do so due to the disparate nature of healthcare data and unstructured patient-specific notes. Simplified, streamlined access to healthcare data is critical to identifying predictors of disease, assessing new therapies, identifying best practices, fine-tuning treatment guidelines and evaluating current therapies and procedures.
- Being able to quickly identify selection criteria to recruit patients for trials is crucial. By using Amazon Comprehend Medical’s ability to recognize complex medical information in unstructured text, identification of appropriate patients for trial recruitment becomes much more efficient from a time and cost perspective.
- Comprehend Medical improves pharmacovigilance and post-market surveillance to monitor adverse drug events by detecting relevant information in clinical text.
- Patients going through a certain kind of therapeutic treatment are to benefit as well. Amazon Comprehend Medical allows a more effective approach to monitoring how patients respond to certain therapies by analyzing their narratives or follow-up notes by health care providers.
IV. Privacy
Of course, the uploading of medical records to the cloud for machine-learning analysis might provoke questions by patients about how Comprehend Medical will ensure their privacy. Amazon assures that patient data is encrypted and can only be unlocked by customers who have a key, and that no data processed will be stored or used for training its algorithms. Furthermore, it has built-in HIPAA compliance so that health care providers will be able to better address data privacy and protected health information requirements.
V. Partners and opinions
Amazon is already working with companies like Deloitte and PricewaterhouseCoopers, as well Seattle’s Fred Hutchinson Cancer Research Center and Roche Diagnostics to process unstructured data in order to find lifesaving insights, personalize healthcare, enable decision support, and population analytics. Fred Hutch was able to evaluate millions of clinical notes to extract and index medical conditions, medications, and choice of cancer therapeutic options, reducing the time to process each document from hours to seconds.
“Curing cancer is, inherently, an issue of time,” said Matthew Trunnell, Chief Information Officer, Fred Hutchinson Cancer Research Center. “For cancer patients and the researchers dedicated to curing them, time is the limiting resource. The process of developing clinical trials and connecting them with the right patients requires research teams to sift through and label mountains of unstructured medical record data. Amazon Comprehend Medical will reduce this time burden from hours per record to seconds. This is a vital step toward getting researchers rapid access to the information they need when they need it so they can find actionable insights to advance lifesaving therapies for patients.”
“Roche’s NAVIFY decision support portfolio provides solutions that accelerate research and enable personalized healthcare. With petabytes of unstructured data being generated in hospital systems every day, our goal is to take this information and convert it into useful insights that can be efficiently accessed and understood,” said Anish Kejariwal, Director of Software Engineering for Roche Diagnostics Information Solutions. “Amazon Comprehend Medical provides the functionality to help us with quickly extracting and structuring information from medical documents so that we can build a comprehensive, longitudinal view of patients, and enable both decision support and population analytics.”
“Amazon Comprehend Medical provides us the ability to realize better results, quicker and with less overhead. By using Amazon Comprehend Medical, our customers are able to focus more on building smarter applications and extracting critical insights and less on annotating, training and re-training models. The ability to perform a very manual task accurately at scale, and securely, allows us to create more impactful solutions and better patient and clinical outcomes. For example, one of our pharmaceutical clients is using Amazon Comprehend Medical on a limited sample size to help extract information that allows them to identify medically relevant events. In preliminary findings, we are seeing a significantly faster throughput than in the past,” said Matt Rich, Healthcare AI Lead at PricewaterhouseCoopers.
“It (Amazon Comprehend Medical) allows us to adopt a scalable, cost-effective, and secure model that had previously been a challenge with prior medical Natural Language Processing tools. We are working to apply the information extraction and classification services towards applications in real-world evidence, pharmacovigilance, competitive intelligence, and provider efficiency, which will help us to mine the information we need to extract meaningful insights and continue to drive transformation in the industry,” shared Dan Housman, Chief Technology Officer of ConvergeHEALTH by Deloitte.
VI. Conclusion
Based on partners’ opinions and recent reviews one could say that Amazon is on the right way to make a significant impact on the medical records environment. Unlike other companies whose focus is on building their own closed personal health record ecosystem, Amazon is striving to deliver a more open and interoperable healthcare ecosystem. Ultimately, this richness of information may be able to one day help consumers with managing their own health, including medication management, proactively scheduling care visits, or empowering them to make informed decisions about their health and eligibility.