Category Archives: Systems

Complex adaptive systems, forecasting systems, systems design, systems architecture, modules, modular systems, modular design, system design patterns, forecasting systems that mine big data

Deep Learning

Singularity, Deep Learning, and AI

DRAFT: January, 2019

CREDIT: The Deep Learning Revolution, by Terrence J. Sejnowski

CREDIT: https://en.wikipedia.org/wiki/Deep_learning

====================

Sadly, I was giving up on speech recognition just as it was emerging. I gave up, after 30 years of waiting, around 1995. Bad idea.

Speech recognition stopped being just cute, and exploded onto the world scene during the late 1990’s. It has taken almost two decades to commercialize, but the technologies birthed in the late 1990’s have now yielded commercial grade results. 

Why then? Why the late 1990’s?  

In reading “The Deep Learning Revolution”, by Terrence J. Senjnowski, I learned why: an underlying technology called “deep learning” had come of age. 

“Deep Learning” was birthed in the late 1990’s, but the research leading up to the term goes back to the 1980”s (and the foundations of all this goes back to 1965).

Many trace the current revolution in deep learning to October 2012. Researchers proved successful in a large-scale ImageNet competition. Their approach won the ICPR contest on analysis of large medical images for cancer detection.

In 2013 and 2014, the error rate on the ImageNet task using deep learning was further reduced, following a similar trend in large-scale speech recognition. The Wolfram Image Identification project publicized these improvements.

Image classification was then extended to the more challenging task of generating descriptions (captions) for images, often as a combination of CNNs and LSTMs.

IN 2015, spectacular practical application began to burst on the scene in 2015. Speech recognition was one. Facial recognition was a second. Pattern recognition can identify cats, dogs and dog breeds, and applications that allow medical diagnosticians to improve their diagnoses. 

Today, applications address computer vision, speech recognition, natural language processing, audio recognition, social network filtering, machine translation, bioinformatics, drug design, medical image analysis, material inspection and board game programs, where they have produced results comparable to and in some cases superior to human experts.

It turns out that new approaches to Deep Learning have broad applicability. But one of those applications that has broken into mass commercialization is …. speech recognition. These breakthroughs trace back to breakthroughs in “speaker recognition” – results that were achieved at SRI.

To understand the massive improvements, consider this: In 2015, Google Voice Search experienced a dramatic performance jump of 49%.

Or consider this: All major commercial speech recognition systems (e.g., Microsoft Cortana, Xbox, Skype Translator, Amazon Alexa, Google Now, Apple Siri, Baidu and iFlyTek voice search, and a range of Nuance speech products, etc.) are based on deep learning.

More on speaker recognition: The recent history traces back to breakthroughs at SRI in the late 1990’s. The research arms of NSA and DARPA needed answers. To get the answers, they turned to SRI international. SRI made the biggest breakthroughs. They cracked “speaker recognition” at that time. They failed, however, to crack “speech recognition”. That came later, around 2003.

Specifically, important papers were published in the late 1990’s describing how deep learning could solve the nagging issues of speaker and speech recognition. 

The deep learning method used was called long short-term memory (LSTM). (Hochreiter and Schmidhuber, 1997.)

Deep learning for speech recognition came later, in the early 21st century. In 2003, LSTM started to become competitive with traditional speech recognizers on certain tasks.. Later it was combined with connectionist temporal classification (CTC) in stacks of LSTM RNNs.

Google Voice Search drew upon “CTC-trained LSTM” – in other words, the LSTM technologies birthed in the late 1990’s had by 2015 yielded commercial-grade results.

Today, lay people understand the power of speech recognition by using “Siri” – or by using the voice transcription technologies on their iPhones. Everyone has noted the vast improvements in the last several years. All of these improvements are due to Deep Learning. 

Let me step back at this point and trace the breakthroughs by researchers. I begin with a glossary:

AI – Artificial Intelligence

ANN – Artificial Neural Networks

DNN – Deep Learning Networks – a variant of artificial intelligence in which software “learns to recognize patterns in distinct layers

RNN – Recurrent Neural Networks

“Deep” – The “deep” in “deep learning” refers to the number of layers through which the data is transformed.

“Layers” – each layer represents a level of abstraction that allows the machine to group like data from unlike data (the machine classifies). Each successive layer uses the output from the previous layer as input. The “deep” in “deep learning” refers to the number of layers through which the data is transformed. Deep learning helps to disentangle these abstractions and pick out which features improve performance.[1]

Pattern Recognition

Image Recognition (In 2011, deep learning-based image recognition has become “superhuman”, producing more accurate results than human contestants.)

Speech Recognition (and ASR – Automatic Speech recognition)

Speaker Recognition (In 1998, deep learning-based speaker recognition was proven to be effective)

Visual Recognition – recognizing object, faces handwritten zip codes etc

Facial Recognition – for example, Facebook’s AI lab performs tasks such as automatically tagging uploaded pictures with the names of the people in them.

Object recognition – (in 1992, a method of extracting 3D objects from a cluttered scene

Medical Imaging – where each neural-network layer operates both independently and in concert, separating aspects such as color, size and shape before integrating the outcomes” of medical imaging

Deep Learning Techniques

Supervised – uses classifications

Unsupervised.  – uses pattern recognition (without human assistance)

Backpropogation (Backprop) – passing information in the reverse direction and adjusting the network to reflect that information.

LSTM – long short-term memory 

CTC – connectionist temporal classification

CAP – the chain of transformations from input to output. CAPs describe potentially causal connections between input and output.

More precisely, deep learning systems have a substantial credit assignment path (CAP) depth. The CAP is the chain of transformations from input to output. CAPs describe potentially causal connections between input and output.

Applications 

TAMER – in 2008, proposed new methods for robots or computer programs to learn how to perform tasks by interacting with a human instructor.[

TAMER (Deep TAMER) – in 2018, is a new algorithm using deep learning to provide a robot the ability to learn new tasks through observation. (robots learn a task with a human trainer, watching video streams or observing a human perform a task in-person). The robot practices the task with the help of some coaching from the trainer, who provided feedback such as “good job” and “bad job.”

CRESCEPTRON, in 1991, a method for performing 3-D object recognition in cluttered scenes. 

Hardware

GPU – in 2009, Nvidia graphics processing units (GPUs) were used by Google Brain to create capable DNNs. This increased the speed of deep-learning systems by about 100 times.

Training Sets

TIMIT (Automatic speech recognition trainer)

MNIST (image classification trainer)

The MNIST database is composed of handwritten digits. it includes 60,000 training examples and 10,000 test examples. As with TIMIT, its small size lets users test multiple configurations.

With this glossary, a few simple statements can pinpoint why the current revolution is exploding:

Hardware has advanced, thanks to GPU commercialized in 2009.

Software has advanced, thanks to GPU-based successes in cancer image identification in 2012. 

Pattern recognition has advanced, with speech recognition leading the way. The TIMIT training set has allowed exponential progress, especially in 2015, leading the way. 

Robotics have advanced, thanks to deep TAMER breakthroughs in 2018. 

Voice Recognition Explodes

CREDIT: The Deep Learning Revolution, by Terrence J. Sejnowski

CREDIT: https://en.wikipedia.org/wiki/Deep_learning

====================

Sadly, I was giving up on voice recognition just as it was emerging. I gave up, after 30 years of waiting, around 1995. Bad idea.

Voice recognition stopped being just cute, and exploded onto the world scene during the late 1990’s. It has taken almost two decades to commercialize, but the technologies birthed in the late 1990’s have now yielded commercial grade results. 

Why then? Why the late 1990’s?  

In reading “The Deep Learning Revolution”, by Terrence J. Senjnowski, I learned why: an underlying technology called “deep learning” had come of age. 

“Deep Learning” was birthed in the late 1990’s, but the research leading up to the term goes back to the 1980”s.

It turns out that new approaches to Deep Learning have broad applicability. But one of those applications that has broken into mass commercialization is …. voice recognition. 

To understand the massive improvements, consider this: In 2015, Google Voice Search experienced a dramatic performance jump of 49%.

Or consider this: All major commercial speech recognition systems (e.g., Microsoft Cortana, Xbox, Skype Translator, Amazon Alexa, Google Now, Apple Siri, Baidu and iFlyTek voice search, and a range of Nuance speech products, etc.) are based on deep learning.

The recent history traces back to breakthroughs at SRI in the late 1990’s. The research arms of NSA and DARPA needed answers. To get the answers, they turned to SRI international. SRI made the biggest breakthroughs. They cracked “speaker recognition” at that time. They failed, however, to crack “speech recognition”. That came later.

Specifically, important papers were published in the late 1990’s describing how deep learning could solve the nagging issues of speaker and voice recognition. The deep learning method used was called long short-term memory (LSTM). (Hochreiter and Schmidhuber, 1997.)

Deep learning for speech recognition came later, in the early 21st century. In 2003, LSTM started to become competitive with traditional speech recognizers on certain tasks.. Later it was combined with connectionist temporal classification (CTC) in stacks of LSTM RNNs.

Google Voice Search drew upon “CTC-trained LSTM” – in other words, the LSTM technologies birthed in the late 1990’s had by 2015 yielded commercial-grade results.

Today, lay people understand the power of speech recognition by using “Siri” – or by using the voice transcription technologies on their iPhones. Everyone has noted the vast improvements in the last several years. All of these improvements are due to Deep Learning. 

Microbiome Science Advances

On January 28, the New York Times published a major article on recent advances in microbiome research.

The article says that breakthroughs began in 2014, when scientists began finding evidence that the micro biome is linked to Alzheimer’s, Parkinson’s, depression, schizophrenia, autism, and other conditions.

The article also describes the early 2000’s, when major advances came from figuring out how to sequenced DNA from microbes in the micro biome. Apparently, a gene called SHANK3 Is particularly central to autism research.

Also apparently, Researchers have isolated one particular bacteria, lactobacillus reuteri. They seem to have identified compounds that are released. These compounds send a signal to nerve endings in the intestines. The Vegas nerve send these signals from we got to the brain, where they alter production of a hormone called Oxsee Tosun. This hormone apparently promote social bonds.

The article is below:

Digital Immortality

In this week’s Sunday NYT Magazine, a discussion was recorded about the future of technology. One of my favorite writers, Sid Mukerjee, discussed chronic disease. In that discussion, he touched on a notion of immortality that I have been pondering for some time.

Here is what he said, and after is what I say in response.

MUKHERJEE: “In terms of longevity, the diseases that are most likely to kill us are neurological diseases and heart disease and cancer. In some other countries, there is tuberculosis and malaria and other infectious diseases, but here it’s the chronic diseases that dominate. There are three ways to think about these chronic diseases. One is the disease-specific way. So, you attack Alzheimer’s as Alzheimer’s; you attack cancer as cancer. The second one is that you forget about the disease-specific manners of attacking diseases and you attack longevity or aging reversal in general. You change diet, change genes, change whatever else — we might call them “trans factors,” which would simply override the “cis factors” that existed for individual diseases. And the third option is some combination of that and some digital form of immortality, which is that you record yourself forever, that you clone yourself and somehow pass along that recording. Which is to say that the body is just a repository of memories, images, times. And as a repository, there’s nothing special about it. The body per se, the mortal coil, is just a coil.

This is the first time I have heard a major thinker put immortality into this context. And yet – its so obvious to do so!

For example:

– wouldn’t it be fair to say that every autobiography ever written would be a sincere attempt by the writer to achieve some form of immortality?

– in like manner, isn’t the task of the biographer, in part, to immortalize their subject?

– more broadly, how do societies around the world remember their ancestors? Their memories are their attempts to allow ancestors to live forever!

This point is nicely illustrated by the Irish culture. In my work on the History of Ireland, the centrality of “oral tradition” was crystal clear. I came continually across how the Irish told stories to revere their ancestors. The Irish would distill their ancestors into a wide variety of stories that helped the present generation understand the past.

So, by extrapolation from this point (which is obvious), can this be asked: “Can I be immortalized digitally?

Digital storage costs have plummeted. Methods of organizing and tagging video and audio recordings are now commonplace. Search engines are commonplace. Pattern recognition combined with search is exploding.

So what will prevent me in the future from immortalizing myself digitally? What prevents me from storing who I am, what I did, what I learned, where I have been, what I have experienced, who I knew, who my ancestors were, who my children and grandchildren were, etc etc?

Perhaps the answer is: nothing. Nothing prevents me from being digitally immortal.

Climate Change Language

We Need A Better Language for Climate Change – that Acts as a Call to Action

============================

Below is as essay that makes the case for a new six-box classification system for global climate change – two columns and three rows. The core idea here is to move climate change out of a subject for the editorial page and into a subject for daily new – much like how storms, earthquakes and epidemics are covered. We want a language that serves as a “call-to-action”.

The news would inform the world about climate-change related occurrences that have impacts that are “major”, “disaster”, or “global disaster”, and that are either “incidents” (one-time) or “recurring”.

I worked this out with Karen . I am the scribe. Obviously, this is DRAFT 1.

=============================
Climate Change Language

CREDIT: Karen Flanders-Reid
CREDIT: https://www.nytimes.com/2018/08/08/opinion/environment/california-wildfires-trump-zinke-climate-change.html

Karen and I read today’s NYT article about California wildfires, and found ourselves musing – is the language of climate change right? Why is a “wildfire” just an isolated incident? Why isn’t it part of a larger wildfire classification system (“BREAKING NEWS: THE CALIFORNIA WILDFIRE HAS JUST BEEN RECLASSIFIED AS CATEGORY V.”?

We went on to ask: if climate change is the critical issue of our day, why Why isn’t the wildfire in California an climate change incident – part of a larger climate change classification system?

Why do the NYT editorial writers have to scream – everything is related to climate change!!!! After all, news breaks when a Hurricane is re-classified: “BREAKING NEWS: THE TROPICAL STORM OVER CUBA HAS JUST BEEN RE-CLASSIFIED BY THE WEATHER SERVICE AS A HURRICANE.”

Why doesn’t climate change have its own global classification system? How do we move from the editorial opinion desk to the news desk? How do we move from “The science is being ignored.” To “BREAKING NEWS: THE WILDFIRES IN CALIFORNIA HAVE JUST BEEN RECLASSIFIED BY THE WEATHER SERVICE FROM A CLIMATE-RELATED INCIDENT (CRI) TO A CLIMATE-RELATED DISASTER (CRD).”

EXAMPLES OF POWERFUL GLOBAL CLASSIFICATION SYSTEMS

To identify a powerful classification system, and the new language it implies, it first would be useful to identify the other global classification systems that exist – especially those with imply a call to action.

There are at least four:

Storms; Classified by the World Meteorological Organization (WMO), using the Saffir–Simpson scale:

Tropical Depression
Tropical Storm
Hurricane/Cyclone Categories 1-5

Source: https://en.wikipedia.org/wiki/Maximum_sustained_wind

Earthquakes: Classified by the US Geological Service, using the Richter Scale:
Moderate (above 8)
Strong (7-7.9)
Major (6-6.9)
Great (5-5.9)

Infectious Disease; Classified by the global centers for disease control, the classes are:

Outbreak (more incident than expected)
Epidemic (spreads rapidly to many people)
Pandemic (spreads rapidly to many people globally)

Source: https://www.webmd.com/cold-and-flu/what-are-epidemics-pandemics-outbreaks#1

A NEW GLOBAL CLASSIFICATION SYSTEM FOR CLIMATE CHANGE

To Begin

We recommend s simple structure, with easily understood terms, that evolves over time:

Starts with a few terms, and adds terms over time.
Begins classifying major occurrences only, and evolves to classify most occurrences.
Begins classifying evidence-based occurrences only (where science is conclusive that the occurrence is climate-change-related) and evolves as science becomes increasingly conclusive.

Initial Terms

“Occurrence” – a natural phenomena that occurs somewhere

“Climate-Change-Related” (CR) – a shorthand for saying that the preponderance of science indicates that a given occurrence is a contributor to or the result of climate change.

“Incident” (I) – an episodic occurrence (with a beginning, middle, and end)
“Recurring” (R) – an on-going occurrence (no end in sight)

“Major” (M) – an occurrence with sufficient size to merit being classified.
“Disaster” (D) – an occurrence, with major impacts
“Global Disaster” (G) – an occurrence with major global impacts

Initial Classification System:

Climate-related Occurrences shall be identified.

Once identified, they shall be classified in one of six classes:

Either “incidents” or “recurring”.
Either “major”, “disaster”, or “global disaster”

“Climate-Change-Related Event” (CRE) – any occurrence that is deemed to be a contributor to climate-change.

“Climate-Change-Related Outcome” (CRO) – any occurrence that is deemed to be the result of to climate-change.

All major climate-change-related occurrences would be classified as follows:

CR Incident (CRE-I): An episodic event, with a beginning, a middle, and an end.
CR Disaster (CRE-D): An episodic event, with global impacts

The Weather Service would be tasked with implementation, and aligning with the World Meteorological Organization (WMO) and other agencies around the world.

History of US Immigration

Borders
A History of Border Security, Illegal and legal immigration

Overview

Regulating the flow of immigrants into the United States has a long, and often tawdry past.

Once regulated, entry then becomes “legal” or “illegal”. And “legal” entry is now generally highly restricted, on a temporary or permanent basis to three different routes: employment, family reunification, or humanitarian protection. All other entry: “illegal”.

Once regulated, borders then become “secure” or “insecure”. Because of trade, borders needed to be highly efficient for goods, and highly “secure” for people. This distinction, between the flow of goods and the flow of people, was an almost unenforceable dilemma, where billions have been expended to do …. the best we can.

Who should regulate? The Supreme Court settled that issue in 1875, opining that this was the role of the Federal Government. Up until then, it was a state responsibility.

How should it regulate? Congress decided that racial quotas were the answer in 1917. Before that time, they actually banned Asian immigration in 1875. The essential idea was to restrict immigration by race to a % of the race’s population in the US (2% of that population was frequently used, noting that 2% of nothing is nothing). The notion of racial quotas was maintained until 1965!

Would there be any exceptions to racial quotas?

Yes, for refugees and asylum-seekers. Congress responded to American sympathies for those fleeing communism and those feeing persecution. Recognizing “refugees” added significant new complexity.

Yes, for spouses and children of American citizens.

Yes, for those born in the Western Hemisphere.

Once regulated, politicians could rail against immigrants, but they rarely provided the funds to enforce the border laws. We severely curtailed legal immigration, and illegal immigration was the easily anticipated result. In 1952, Congress specified that legal immigration be limited to 175,455 per year!

Also easily anticipated, “illegals” brought massive issues for schools, health care, housing, etc. As the number of “illegals” grew, so grew the pressure to do something, anything, to reduce the pressure. Congress has been forced to act, as they did in 1986 when they granted amnesty to approximately 3 million illegals!

So the history of immigration in the United States includes major shifts in policy in 1875 (Supreme Court rules), 1891 (Federal bureaucracy formed), 1924 (racial quotas put in place), 1986 (racial quotas replaced and amnesty granted).

“Illegals” are out of control. Estimates of illegals are 3 million illegals in 1986, 7 million in 2001, and 12 million in 2017. As a % of U.S. population, “foreign-born” dropped from 14.7% in 1910 to 4.7% in 1970, and has been rising ever since. In 2013, there were 13.1% of the population who were foreign born (CREDIT:PEW).

Discussion
Immigration became a full-fledged subject for the nation in 1875, when the Supreme Court ruled that it was a Federal responsibility. Shortly thereafter, Congress stepped up and began excluding people – literally making it “illegal” for them to enter the United States. They banned Asians in 1875 and Chinese in 1882 (the “Asian Exclusion Act” and the “Chinese Exclusion Act” set the stage for all restrictions on immigration that would follow.

In 1891, the Federal Government took a big step: they created a bureaucracy to execute the laws. The Immigration Act of 1891 established a Commissioner of Immigration in the Treasury Department. With the two exceptions noted above, states regulated immigration before 1890.

Before then, this “nation of immigrants” actually had an immigration hiatus from 1790 to 1815, when “foreign-born” reached a low. Immigration as we now know it began with some force in 1830, when “foreign-born reached 9.7% of the population. By 1850, census estimates place immigrants at 1.7 million people, and “foreign-born” at 2.2 million. Between 1870 and 1910, foreign born hovered between 13% and 15% of population. It then started to dip, moving to 4.7% in 1970. It has been climbing since, reaching 13.1% in 2013.

Since then, waves of immigration brought the country waves of immigrants:

Between 1850 and 1930, 25 million Europeans immigrated. Italians, Greeks, Hungarians, Poles, and others speaking Slavic languages made up the bulk of this migration. But among them were 5 million Germans, 3.5 million British, and 4.5 million Irish. 2.5 to 4 million Jews were among them.

The twentieth century began with debates about immigration, and we have been debating the subject ever since.

In 1907, Congress created The Dillingham Commission to investigate the effects of immigration on the country. They wrote forty volumes on the subject.

In 1917, Congress changed the nation’s basic policy about immigration. We began setting “quotas” and limiting access based on literacy. The first such law was a literacy requirement in 1917.

In 1921, Congress adopted the Emergency Quota Act, set quotas. The National Origins Formula assigned quotas based on national origins. This complex legislation gave preference to immigrants from Central, Northern and Western Europe, severely limiting the numbers from Russia and Southern Europe, and declared all potential immigrants from Asia unworthy of entry into the United States (to our shame, this law made it virtually impossible for Jews fleeing Germany after 1934 to immigrate to the United States).

In 1924 , Congress adopted The Immigration Act of 1924. It set quotas for European immigrants so that no more than 2% of the 1890 immigrant stocks were allowed into America.

Interestingly, no quotas were set for people born in the Western Hemisphere.

This era, and its legislative framework, lasted until 1965. During this period, Congress recognized the notion of a “refugee” seeking “amnesty”. Jewish Holocaust survivors after the war, those fleeing Communist rule in Central Europe and Russia, Hungarians seeking refuge after their failed uprising in 1956, and Cubans after the 1960 revolution, and others moved the conscience of the nation.

In 1965, Congress adopted the Hart-Celler Act. It was a by-product of the civil rights revolution and a jewel in the crown of President Lyndon Johnson’s Great Society programs. It abolished the racially based quota system.The law replaced these quotas with new preferential categories. It gave particular preference to immigrants with U.S. relatives and job skills deemed critical.

In 1986, the Immigration Reform and Control Act (IRCA) was adopted. It created, for the first time, penalties for employers who hired illegal immigrants. IRCA, also granted amnesty to workers in the country illegally. In practice, amnesty was granted for about 3,000,000 illegal immigrants. Most were from Mexico. Legal Mexican immigrant family numbers were 2,198,000 in 1980, 4,289,000 in 1990 (includes IRCA), and 7,841,000 in 2000.

References

https://en.wikipedia.org/wiki/History_of_immigration_to_the_United_States

https://www.politico.com/magazine/story/2017/08/06/trump-history-of-american-immigration-215464

https://americanimmigrationcouncil.org/research/why-don’t-they-just-get-line

How U.S. immigration laws and rules have changed through history

http://assets.pewresearch.org/wp-content/uploads/sites/7/reports/39.pdf

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3407978/

SmartWatch Technology Reliably Detects Afib

The quantified self movement strikes again!

CREDIT: Cleveland Clinic Article on Detection of Afib via SmartWatch

Smartwatch Technology Reliably Detects Afib Prior to Cardioversion
Study suggests a role for KardiaBand when paired with physician review

A newly FDA-approved smartwatch accessory can record heart rhythm and successfully differentiate atrial fibrillation (AF) from normal sinus rhythm (SR) through an automated algorithm, according to a Cleveland Clinic investigation. The study, which will be presented March 11 at the American College of Cardiology’s 67th Scientific Session, also showed that the accuracy of interpretation gets even better when the accessory is supported by physician review.
The findings suggest that the wearable technology, known as KardiaBand™, can help screen patients before presentation for elective cardioversion to avoid unnecessary procedures, among other potential uses.
KardiaBand, which consists of a software app for an Apple Watch® and a sensor band that replaces one of the watch’s straps, provides a 30-second recording of an ECG rhythm strip when the wearer places a thumb on the sensor band. The app contains an algorithm for automated detection of AF.
“Our objective was to determine how accurately KardiaBand and its algorithm can differentiate AF from sinus rhythm compared with physician-interpreted 12-lead ECGs,” says senior author Khaldoun Tarakji, MD, MPH, a Cleveland Clinic electrophysiologist. In November 2017, the device became the first smartwatch healthcare accessory to be approved by the FDA, “but we wanted to test it ourselves to determine how well it would perform in clinical practice,” Dr. Tarakji explains.
Study essentials
To that end, he and Cleveland Clinic colleagues prospectively enrolled 100 consecutive patients (mean age, 68 ± 11 years) with chronic AF who were scheduled to undergo cardioversion. Upon presenting for the cardioversion procedure, all patients were given a KardiaBand-equipped smartwatch and trained in its use, after which they underwent traditional ECG assessment and a 30-second KardiaBand recording. If cardioversion was still indicated, they underwent ECG and KardiaBand testing after the procedure. KardiaBand recordings were then compared with the physician-reviewed ECGs and also reviewed by two blinded electrophysiologists, with these readings compared to ECG interpretations.
Eight patients did not undergo cardioversion because they presented in SR; these patients were excluded. Among the remaining patients, a total of 169 pairs of ECG and KardiaBand recordings were available for comparison (each patient had two before and two after cardioversion).
Key findings
• Of the 169 pre-cardioversion KardiaBand recordings, 57 fell out as “unclassified,” meaning that the KardiaBand algorithm did not draw a conclusion of either AF or SR.
• Among the remaining 112 pairs of recordings, the reviewing electrophysiologists determined that the KardiaBand algorithm correctly detected AF with 93 percent sensitivity and 84 percent specificity compared with ECG.
• When the blinded reviewers bypassed the automated algorithm and interpreted each patient’s KardiaBand strips against his or her ECG, sensitivity rose to 99 percent and specificity was 83 percent. Further, in the 57 unclassified cases, the reviewers were able to use the strips to correctly diagnose AF versus SR with 100 percent sensitivity and 80 percent specificity.
“This study shows that KardiaBand provides excellent sensitivity and good specificity in identifying AF,” says Dr. Tarakji. “The numbers improve further with physician overview of these recordings, indicating that even unclassified KardiaBand strip recordings could be of value to reading physicians.”
Smart devices demand smart use
KardiaBand carries the benefit of enabling patients to record their rhythm at any time, as opposed to only when they are wearing a Holter monitor or at a physician’s office. “We can catch intermittent episodes when they happen, and we’re not limited to a specific duration of monitoring time,” Dr. Tarakji says. He adds that wearable devices like this can also reduce time spent responding to false alarms if a recording taken at the same time shows normal rhythm.
Yet many questions remain about how KardiaBand and similar products may ultimately be used in practice. Dr. Tarakji cites a few examples:
• Which patients are best suited to this technology? For many patients dealing with AF, KardiaBand can provide reassurance when they need it. But for others, having constant access to their ECG data may lead them to check their rhythm obsessively, raising anxiety. “In general, however, patients value the instant feedback they get,” Dr. Tarakji observes.
• Do physicians have the IT infrastructure in place to make these devices part of their practice? Wearable devices can mean a flood of event reports to clinicians’ email boxes. At Cleveland Clinic, information from patients’ KardiaBands bypasses the email system and feeds into a cloud-computing platform that physicians can access anytime.
• How should clinicians respond to short episodes, particularly in asymptomatic patients? “We currently have a gap in our clinical knowledge about whether brief, random episodes that are asymptomatic warrant anticoagulation or not,” Dr. Tarakji explains, adding that ongoing studies are trying to address this important question.
“Future studies will focus on how we can use these smart devices intelligently to make sure they’re improving quality of care rather than just producing noise for physicians,” he observes.
A parallel goal, he says, is to ensure that the devices provide value by making care delivery more efficient. Noting that patients currently need to pay for KardiaBand out of pocket, Dr. Tarakji says that “developing a richer body of research evidence is the best way we can demonstrate cost-effectiveness to healthcare payers.”
Tech like this can’t be ignored
Indeed, KardiaBand could prove cost-effective by allowing patients who are in SR to avoid needless trips for elective procedures, such as in the case of the eight patients in the study who were found to be in SR when they presented for cardioversion and did not require the procedure. Other potential uses of KardiaBand for the longitudinal management of AF patients could well prove cost-effective too.
Regardless of how quickly such cost-effectiveness evidence may come, Dr. Tarakji says clinicians cannot be passive in the face of technologies like KardiaBand. “Patients will come to us with new products, and we can’t turn away,” he observes. “We need to test these products and find ways of responding to the information they deliver in a way that improves patient outcomes, all while remaining mindful of both patient and physician satisfaction.”
The researchers report that KardiaBand’s manufacturer, AliveCor, provided smartwatches for the study but was not involved in the study’s design, implementation, data analysis or interpretation.

Well-Being – Real Time Revisited

NOTE: This post revisits a post titled “Well-Being Real Time”. The original post was May, 2014, and can be found at: http://johncreid.com/2014/05/well-being-real-time/.

Well-Being – Real Time Revisited

Well-being is arguably the central mega-trend of the 21st century. As we look to the future, we have an obligation to “unpack” this dense concept, and find its essential component parts.

We describe these components here as “ACE” – ACT, CARE, and EAT. The wish we have for ourselves and for others is to be well. “Be Well” is our salutation and our call to actions.

How far out are we looking?

The future is now. ACE is here – together with real time measuring and monitoring. ACE is our pathway to greater and greater levels of personal well-being.

ACE measuring and monitoring will be supported by all elements of the quantified self movement. FitBit, Apple Watch, and so many other new monitoring devices will allow us to to bring personal well-being into a real-time modality.

ACE represents three pillars, each deceptively simple:

A – ACT: ACT is short for activity. The call to action is “stay active”. Well-being activity has physical activity at its center, but the pillar also embraces social activity, and activities of the mind. Staying active is a critical element of being well.
C – CARE: CARE is short for well-being care. The call to action is “care for yourself” and “care for others.”Well-being care of course has health care at its center, but there is so much more. e.g. genomics, massage, essential oils, acupuncture, etc. “Caring for myself” and “Caring for others” are elements of this pillar. “Preventive care” regular check-ups, colonoscopies after age 50, mammograms, pre-natal care for expecting mothers, etc.
E – EAT: EAT is short for eating and drinking. The call to action is “Eat well.” Well-being eating is the exploration of how what we eat and drink contributes to our well-being.

As simple as these pillars appear, each is complex: deep enough for a life-time of focus. Each represents bodies of research, skills, capabilities, and areas of professional endeavor. All together, these pillars represent pathway that each of us will follow as we attain greater and greater levels of personal well-being.

Discussion:

ACT

A – ACT (walking, running, calories burned etc)

Staying active is a critical element of being well. Well-being activity has physical activity at its center: sports, walking, lifting, climbing, yoga, and all of the other activities that light up a FitBit. The pillar also embraces activity of other kinds, e.g. social activity, and activities of the mind.

CARE

Well-being care is all about promoting health. Of course, it has health care at its center, but there is so much more. e.g. mental health, addictive behaviors, massage, genomics, essential oils, acupuncture, etc.

“Caring for myself” and “Caring for others” are elements of this pillar. “Preventive care”, eldercare and aging, palliative care are included, but so are regular check-ups, colonoscopies after age 50, mammograms, pre-natal care for expecting mothers, etc.

The ability to routinely monitor vital signs at home or at the office will be a part of this pillar. Lab work – including saliva, blood, and stool samples, will be more real time, more regular and less expensive. These trends will be one of the keys to progress in the care pillar. On the innovation side of this pillar will be many technologies, but breakthroughs in genomics will certainly be high on the list. Telemedicine is another innovation that will alter access to well-being care.

Predictive modeling will be more relevant than never. Am I headed for pre-diabetes? If so, what evidence shows me a path to avoid that condition?

CARE-MMEDS (what MEDS I take, what compliance I have, etc)

CARE-RResting Metabolic Rate (calories burned at rest)

CARE-VVITALS (pulse, BP, etc)

CARE-LLABS (blood testing, etc)

CARE-SSleep (duration, deep sleep, etc)

EAT

EAT is short for eating and drinking. The call to action is “Eat well.”

Well-being eating is the exploration of how what we eat and drink contributes to our well-being. Naturally, there is a social element, where eating and drinking together makes the experience more fulfilling. There is a physiological element, having to do with ingestion, osmosis, calories, glucose and glycogen, enzymes, etc. There is a psychological element, related to the feelings of satiety, or hunger, or thirst, and their related cravings. There is a sensory element, where sweet and sour contrasts, aromas, and their related metaphorical associations, play a part.

Eating delicious food and drink with friends is certainly a component. But achieving a balanced diet, with moderation as a central tenant,

On the one hand, this pillar is ancient. For thousands of years, elders have taught daughters and sons how to cook well. and cooking techniques have evolved

On the other hand, this pillar is ripe for innovation. The new breakthrough science related to the micro-biome is a part.

EATS (what I eat and drink, especially calories)

Implications

Monitoring all components of ACE (MEDS, Activity, Resting Metabolism,VITALS, EATS, LABS, Sleep) is now going to accelerate at an exponential rate.

There will be three settings where ACE monitoring will accelerate:

Employees in Workplaces: Employers will offer employees routine monitoring as part of employee benefits and/or health insurance.
Residents in Communities: Communities will offer residents routine monitoring as one of their amenities. Wellbeing facilities and programs will become as important as golf courses and swimming pools. Look for HOA’s,Condo and Coop associations, and subdivision developers to increasingly view MARVELS as critical to “place-making”.
Clients of service-providers: Hotels, spas, assisted-living centers, nursing homes, and many others will increasingly offer MARVELS monitoring as one of their base services.

The Privacy Imperative will be the critical success factor for all of these pushes into the future. It is foundational.

Without it, there will be no progress.

With it, personalized, real-time care will flourish. Each individual will be able to opt-in to his care-coaching community (and to opt-out whenever they choose), and get the extraordinary benefits that such a community can provide.

Want to talk to your well-being coach? FaceTime them, and they – with your permission – will help you sort out what’s going on with you.

Feel like you might need a check-in with a doctor? Send them an email – with your ACE history embedded in it, or get them on the phone or FaceTime, and see if they need you to come in.

The future is now.

BEWELL Centers will be everywhere. Look for:

DWELL CENTERS (part of BEWELL Centers) – for community ACE measuring and monitoring support. Target population is neighbors in the community.

Employee BEWELL CENTERS (part of BEWELL Centers) – for employees in workplaces ACE measuring and monitoring support. Target population is employees in the workplace.

CLIENT BEWELL CENTERS (Part of BEWELL Centers – for service-providers ACE measuring and monitoring support.Target population is clients of the service provider.
(Walgreens and CVS are already moving aggressively in this direction>

References:
The Privacy Imperative
LABS revolution
LABS By Disease
Quantified Self Movement

Amazon, BH, JPMorgan

With 1.2 million employees, Amazon, Berkshire Hathaway, and JP Morgan have decided to venture together into health care for their employees.

Following in the grand tradition of Henry Ford, who set up Henry Ford Hospital in Detroit, these three giants are stepping in too.

They have no illusions about how difficult it will be. But with premiums rising 19% per year, its clear that Congress is doing nothing, and someone has to do something.

“Planning for the new company is being led by Marvelle Sullivan Berchtold, a JPMorgan managing director who was previously head of the Swiss drugmaker Novartis’s mergers and acquisitions strategy; Mr. Combs; and Beth Galetti, a senior vice president at Amazon.”

The article points out that there are others working on this.

“Robert Andrews, chief executive of the Healthcare Transformation Alliance, a group of 46 companies, including Coca-Cola and American Express, that have banded together to lower health care costs.”

“Walmart contracted with groups like the Cleveland Clinic, Mayo and Geisinger, among others, to take care of employees who need organ transplants and heart and spine care.”

“Caterpillar, the construction equipment manufacturer, sets its own rules for drug coverage, which it has said saves it millions of dollars per year, even though it still uses a pharmacy benefit manager to process its claims.”

Suzanne Delbanco, the executive director for the Catalyst for Payment Reform, a nonprofit group that mainly represents employers”

=================
CREDIT: https://www.nytimes.com/2018/01/30/technology/amazon-berkshire-hathaway-jpmorgan-health-care.html?smid=nytcore-ipad-share&smprod=nytcore-ipad

TECHNOLOGY
Amazon, Berkshire Hathaway and JPMorgan Team Up to Try to Disrupt Health Care

By NICK WINGFIELD, KATIE THOMAS and REED ABELSON
JAN. 30, 2018
SEATTLE — Three corporate behemoths — Amazon, Berkshire Hathaway and JPMorgan Chase — announced on Tuesday that they would form an independent health care company for their employees in the United States.

The alliance was a sign of just how frustrated American businesses are with the state of the nation’s health care system and the rapidly spiraling cost of medical treatment. It also caused further turmoil in an industry reeling from attempts by new players to attack a notoriously inefficient, intractable web of doctors, hospitals, insurers and pharmaceutical companies.
It was unclear how extensively the three partners would overhaul their employees’ existing health coverage — whether they would simply help workers find a local doctor, steer employees to online medical advice or use their muscle to negotiate lower prices for drugs and procedures. While the alliance will apply only to their employees, these corporations are so closely watched that whatever successes they have could become models for other businesses.

Major employers, from Walmart to Caterpillar, have tried for years to tackle the high costs and complexity of health care, and have grown increasingly frustrated as Congress has deadlocked over the issue, leaving many of the thorniest issues to private industry. About 151 million Americans get their health insurance from an employer.
(Why will health care be so difficult for these companies to untangle? Analysis from The Upshot.)
But Tuesday’s announcement landed like a thunderclap — sending stocks for insurers and other major health companies tumbling. Shares of health care companies like UnitedHealth Group and Anthem plunged on Tuesday, dragging down the broader stock market.

That weakness reflects the strength of the new entrants. The partnership brings together Amazon, the online retail giant known for disrupting major industries; Berkshire Hathaway, the holding company led by the billionaire investor Warren E. Buffett; and JPMorgan Chase, the largest bank in the United States by assets.

They are moving into an industry where the lines between traditionally distinct areas, such as pharmacies, insurers and providers, are increasingly blurry. CVS Health’s deal last month to buy the health insurer Aetna for about $69 billion is just one example of the changes underway. Separately, Amazon’s potential entry into the pharmacy business continues to rattle major drug companies and distributors.
(Here’s a look at how the even the threat of Amazon’s entry into an industry can rattle stocks.)

The companies said the initiative, which is in its early stages, would be “free from profit-making incentives and constraints,” but did not specify whether that meant they would create a nonprofit organization. The tax implications were also unclear because so few details were released.
Jamie Dimon, the chief executive of JPMorgan Chase, said in a statement that the effort could eventually be expanded to benefit all Americans.

“The health care system is complex, and we enter into this challenge open-eyed about the degree of difficulty,” Jeff Bezos, Amazon’s founder and chief executive, said in a statement. “Hard as it might be, reducing health care’s burden on the economy while improving outcomes for employees and their families would be worth the effort.”

The announcement touched off a wave of speculation about what the new company might do, especially given Amazon’s extensive reach into the daily lives of Americans — from where they buy their paper towels to what they watch on television. It follows speculation that the company, which recently purchased the grocery chain Whole Foods, might use its stores as locations for pharmacies or clinics.
(We asked health care experts to imagine what the three corporations might do.)

“It could be big,” Ed Kaplan, who negotiates health coverage on behalf of large employers as the national health practice leader for the Segal Group, said of the announcement. “Those are three big players, and I think if they get into health care insurance or the health care coverage space, they are going to make a big impact.”

TAKING ON ‘THE HUNGRY TAPEWORM’
A look at the three companies that announced a joint health care initiative on Tuesday.

Total employees: 1.2 million 
Amazon: 540,000 
Berkshire Hathaway: 367,000
JPMorgan Chase: 252,000.
Individual strengths 
Amazon: logistics and technology
Berkshire Hathaway: insurance
JPMorgan Chase: finance.

Jeff Bezos of Amazon:
“The healthcare system is complex, and we enter into this challenge open-eyed about the degree of difficulty.”
Warren E. Buffett of Berkshire Hathaway:
“The ballooning costs of healthcare act as a hungry tapeworm on the American economy. Our group does not come to this problem with answers. But we also do not accept it as inevitable.”
Jamie Dimon of JPMorgan Chase:
“The three of our companies have extraordinary resources, and our goal is to create solutions that benefit our U.S. employees, their families and, potentially, all Americans.”

But others were less sure, noting that the three companies — which, combined, employ more than one million people — might still hold little sway over the largest insurers and pharmacy benefit managers, who oversee the benefits of tens of millions of Americans.

“This is not news in terms of jumbo employers being frustrated with what they can get through the traditional system,” said Sam Glick of the management consulting firm Oliver Wyman in San Francisco. He played down the notion that the three partners would have more success getting lower prices from hospitals and doctors. “The idea that they could have any sort of negotiation leverage with unit cost is a pretty far stretch.”

Even the three companies don’t seem to be sure of how to shake up health care. People briefed on the plan, who asked for anonymity because the discussions were private, said the executives decided to announce the initiative while still a concept in part so they can begin hiring staff for the new company.

Three people familiar with the partnership said it took shape as Mr. Bezos, Mr. Buffett, and Mr. Dimon, who are friends, discussed the challenges of providing insurance to their employees. They decided their combined access to data about how consumers make choices, along with an understanding of the intricacies of health insurance, would inevitably lead to some kind of new efficiency — whatever it might turn out to be.

“The ballooning costs of health care act as a hungry tapeworm on the American economy,” Mr. Buffett said in the statement. “Our group does not come to this problem with answers. But we also do not accept it as inevitable.”

Over the past several months, the three had met formally — along with Todd Combs, an investment officer at Berkshire Hathaway who is also on JPMorgan’s board — to discuss the idea, according to a person familiar with Mr. Buffett’s thinking.

The three chief executives saw one another at the Alfalfa Club dinner in Washington on Saturday, but by then each had already had dozens of conversations with the small in-house teams they had assembled. The plan was set.

Mr. Buffett’s motivation stems in part from conversations he has had with two people close to him who have been diagnosed with multiple sclerosis, according to the person. Mr. Buffett, the person said, believes the condition of the country’s health care system is a root cause of economic inequality, with wealthier people enjoying better, longer lives because they can afford good coverage As Mr. Buffett himself has aged — he is 87 — the contrast between his moneyed friends and others has grown starker, the person said.

The companies said they would initially focus on using technology to simplify care, but did not elaborate on how they intended to do that or bring down costs. One of the people briefed on the alliance said the new company wouldn’t replace existing health insurers or hospitals.

Planning for the new company is being led by Marvelle Sullivan Berchtold, a JPMorgan managing director who was previously head of the Swiss drugmaker Novartis’s mergers and acquisitions strategy; Mr. Combs; and Beth Galetti, a senior vice president at Amazon.

One potential avenue for the partnership might be an online health care dashboard that connects employees with the closest and best doctor specializing in whatever ailment they select from a drop-down menu. Perhaps the companies would strike deals to offer employee discounts with service providers like medical testing facilities.

“Each of those companies has extensive experience using transformative technology in their own businesses,” said John Sculley, the former chief executive of Apple who is now chairman of a health care start-up, RxAdvance. “I think it’s a great counterweight to what government leadership hasn’t done, which is to focus on how do we make this health care system sustainable.”

How Amazon Rattles Other Companies
The e-commerce giant’s actions – some big, like buying Whole Foods Markets; some smaller, like Amazon meal kits – have led to stock sell-offs for a wide range of businesses.

Erik Gordon, a professor at the University of Michigan’s Ross School of Business, predicted that the companies would attempt to modernize the cumbersome process of doctor appointments by making it more like booking a restaurant reservation on OpenTable, while eliminating the need to regularly fill out paper forms on clipboards.

“I think they will bring the customer-facing, patient-facing thing into your smartphone,” he said.

Amazon has long been mentioned by health care analysts and industry executives as a potential new player in the sector. While the company has remained quiet about its plans, some analysts noted that companies often use their own employees as a testing ground for future initiatives.

The entry of Amazon and its partners adds to the upheaval in an industry where much is changing, from government programs after the overhaul of the tax law to the uncertain future of the Affordable Care Act. All the while, medical costs have persistently been on the rise.

Nationwide, average premiums for family coverage for employees rose to $18,764 last year, an increase of 19 percent since 2012, according to the Kaiser Family Foundation. Workers are increasingly paying a greater share of those costs — they now pay 30 percent of the premium, in addition to high deductibles and growing co-payments.
“Our members’ balance sheets speak for themselves — health care is a growing cost at a time when other costs are either not rising or falling,” said Robert Andrews, chief executive of the Healthcare Transformation Alliance, a group of 46 companies, including Coca-Cola and American Express, that have banded together to lower health care costs.

Other major employers have also sought more direct control over their employees’ health care. Walmart contracted with groups like the Cleveland Clinic, Mayo and Geisinger, among others, to take care of employees who need organ transplants and heart and spine care. Caterpillar, the construction equipment manufacturer, sets its own rules for drug coverage, which it has said saves it millions of dollars per year, even though it still uses a pharmacy benefit manager to process its claims.

Suzanne Delbanco, the executive director for the Catalyst for Payment Reform, a nonprofit group that mainly represents employers, said controlling rising prices is especially hard in markets where a local hospital or medical group dominates. While some have tried to tackle the issue in different ways, like sending employees with heart conditions to a specific group, “it’s piecemeal,” she said.

She added, “There are so many opportunities to do this better.”

The issue is not solely a 21st-century concern: In 1915, Henry Ford became increasingly worried about the quality of health care available to his growing work force in Detroit, so he opened the Henry Ford Hospital. It is still in existence today.

Nick Wingfield reported from Seattle, Katie Thomas from Chicago and Reed Abelson from San Francisco. Michael J. de la Merced contributed reporting from London, and Emily Flitter from New York.

A version of this article appears in print on January 31, 2018, on Page A1 of the New York edition with the headline: 3 Giants Form Health Alliance, Rocking Insurers. Order Reprints| Today’s Paper|Subscribe