Sorting out AI, ML, DL, and NLP

TOPICS:

Posted By: Roberta Mullin February 5, 2020

The alphabet soup of acronyms in the world Artificial Intelligence. What are they? What do they mean? What is the difference between them? This is our ongoing reporting on AI and how it is being integrated into healthcare technology. We are seeking out the thought leaders and innovations that are moving the needle forward using artificial intelligence. Read more posts on Artificial Intelligence in Healthcare.

Follow the hashtag #AIinHealthcare.

Listen

From The Incrementalist, host Dr. Nick van Terheyden aka Dr. Nick talks to Dr. Brian Powers, Internal Medicine Resident at Brigham and Women’s discussing the use of AI and the risk of bias that arises from the data used to create and program these algorithms. And how we combat this bias with a team based approach and vigilance.

In the News

Current Health Collaborates with Mass General Hospital to Reduce Maternal Mortality Rates in Sub-Saharan Africa
Current Health (@HeyCurrent), which offers the leading FDA-cleared, artificial intelligence (AI)-powered patient management platform, announced that it will collaborate with Massachusetts General Hospital (MGH) and the Mbarara University of Science and Technology (MUST) in a study aimed at reducing maternal deaths and severe complications following cesarean delivery in Sub-Saharan Africa.

FDA Clears Eko’s AFib and Heart Murmur Detection Algorithms, Making It the First AI-Powered Stethoscope to Screen for Serious Heart Conditions
Eko (@Eko_Health), a digital health company applying artificial intelligence (AI) in the fight against heart disease, announced that the U.S. Food and Drug Administration has cleared a suite of algorithms that, when combined with Eko’s digital stethoscopes, will enable healthcare providers in the U.S. to more accurately screen for heart conditions during routine physical exams. If left undiagnosed, these heart conditions can lead to stroke and heart failure.

NantHealth and NantOmics Reveal a Novel AI Based Machine-Learning Digital Pathology Software for Lung Cancer by Identifying Tumor Infiltrating Killer Cells From Whole Slide Images
NantHealth, Inc. (@NantHealth), a next-generation, evidence-based, personalized healthcare company and NantOmics, LLC, the leader in molecular analysis, presented a novel artificial intelligence platform for aiding pathologists in image-based lung cancer subtyping at the Society for Imaging Science and Technology’s International Symposium on Electronic Imaging 2020. This novel machine vision software platform accurately subtypes lung cancer pathology and achieves high concordance with analysis performed by trained medical pathologists.

Deep Learning Artificial Intelligence Leader Vyasa Analytics Selected for 2020 MassChallenge HealthTech Accelerator
Vyasa Analytics (@vyasaanalytics), an innovative provider of highly scalable deep-learning AI analytics software, has been selected to join the 2020 MassChallenge HealthTech (@MassChallengeHT) program, a startup accelerator that matches later-stage digital health startups with industry experts to accelerate innovation and transform healthcare. Vyasa was selected to join this cohort, alongside 26 other digital health startups, because of its innovation in the field that can help tackle some of healthcare’s biggest challenges and improve patient experience.

OptraHEALTH Collaborates with Microsoft Azure to Serve Genomic Health Industry with the Launch of “GeneFAX™ Pro”
OptraHEALTH® (@Optra_Health) announced that GeneFAX™, a conversational Genetic ChatBOT platform powered by Microsoft Azure AI (@AzureMktPlace), is now available as GeneFAX™ Pro on Microsoft Azure Marketplace, an online store providing applications and services for use on Azure. The availability of GeneFAX™ Pro on Azure Marketplace will benefit genetic testing providers for their DTC as well as Diagnostics tests.

AI Therapeutics Announces that a Common LAM-002 Mechanism in Cancer and Neurodegenerative Diseases Shows Antitumor Activity in the Clinic and Hope for ALS
AI Therapeutics (@aitherapeutics) is a clinical-stage biopharmaceutical company that has created an artificial intelligence-driven drug development platform for matching drugs to new indications. The company has made significant recent progress with its four clinical assets and proprietary Guardian Angel™ algorithm.

CPSI Partners With WorldQuant Predictive to Improve Healthcare Outcomes
CPSI (@cpsiehr), a community healthcare solutions company, and WorldQuant Predictive (@WQPredictive), a predictive analytics firm, announced an innovative partnership that will use advanced technology, including machine learning and artificial intelligence, to derive meaningful insights from real-world clinical evidence to improve healthcare outcomes for patients.

To Read

Black Book
Providers’ Underuse of Analytics Could Be Costing Healthcare Consumers Billions of Dollars, Black Book Survey
The healthcare analytics market is soaring, exceeding $14 billion spent in 2019 and anticipated to grow at a CAGR of 23.5% to $40 billion by 2024. With the recent explosion of various analytics applications and models implemented, 95% of hospitals and physician group leaders have access to supportive decision-making tools, sharply up from 63% in 2016. Yet according to a recent industry survey, utilization of advanced analytics for key decision-making and strategic planning is self-reported as negligible by eight in ten surveyed healthcare managers.

To Follow

The Basics and Resources

From the leading text book around the world, Artificial Intelligence: A Modern Approach.

Artificial Intelligence is composed of six different disciplines:

  1. Natural Language Processing to enable it to communicate successfully in English
  2. Knowledge Representation to store what it knows or hears
  3. Automated Reasoning to use the stored information to answer questions and to draw new conclusions
  4. Machine Learning to adapt to new circumstances and to detect and extrapolate patterns
  5. Computer Vision to perceive objects
  6. Robotics to manipulate objects and move about

To build a generally intelligent agent, you need machine learning in addition to the other aspects mentioned above.

Machine Learning is roughly the science of prediction. Given certain knowns (features), you wish to predict some unknowns (targets). The unknown could be structured (e.g. numeric) or unstructured (e.g. a string response).

Deep Learning is a sub field of machine learning where concepts are learned hierarchically. The simplest concepts emerge first, followed by more complicated concepts that build on the simpler ones. Usually, this leads to a simple layered hierarchy of concepts.

Optum Resource Library

Natural Language Processing: AI with an ROI
Health care providers need to see a return on any analytic investment they make. Natural language processing (NLP) is one way AI can help providers convert the potential within their health data into quality improvement and cost savings. Natural language processing is an AI technology that actually makes sense for health care.

IBM Analytics

Navigating the A.I. and Cognitive Maze – If you work in the area of Artificial Intelligence (AI) and Cognitive Computing, you might use buzz words and phrases which to others might be perceived as confusing jargon. This article attempts to explain what these terms mean, how they relate to one other and where they all fit along the AI and cognitive time continuum. I include a glossary of my top 20 useful AI/cognitive terms — and advice on getting started on your AI/cognitive journey.