Tag: Diagnosis

Smells like AI Spirit

Sep 6, 2023

By KIM BELLARD

There are so many exciting developments in artificial intelligence (AI) these days that one almost becomes numb to them. Then along comes something that makes me think, hmm, I didn’t see that coming.

For example, AI can now smell.

Strictly speaking, that’s not quite true, at least not in the way humans and other creatures smell. There’s no olfactory organ, like our nose or a snake’s tongue. What AI has been trained to do is to look at a molecular structure and predict what it would smell like.

If you’re wondering (as I certainly did when I heard AI could smell), AI has also started to crack taste as well, with food and beverage companies already using AI to help develop new flavors, among other things. AI can even reportedly “taste wine” with 95% accuracy. It seems human senses really aren’t as human-only as we’d thought.

The new research comes from the Monell Chemical Senses Center and Osmo, a Google spin-off. It’s a logical pairing since Monell’s mission is “to improve health and well-being by advancing the scientific understanding of taste, smell, and related senses,” and Osmo seeks to give “computers a sense of smell.” More importantly, Osmo’s goal in doing that is: “Digitizing smell to give everyone a goal at a better life.”

Osmo CEO Alex Wiltschko, PhD says: “Computers have been able to digitize vision and hearing, but not smell – our deepest and oldest sense.” It’s easy to understand how vision and hearing can be translated into electrical and, ultimately, digital signals; we’ve been doing that for some time. Smell (and taste) seem somehow different; they seem chemical, not electrical, much less digital. But the Osmo team believes: “In this new era, computers will generate smells like we generate images and sounds today.”

I’m not sure I can yet imagine what that would be like.

The research team used an industry dataset of 5,000 known odorants, and matched molecular structures to perceived scents, creating what Osmo calls the Principle Odor Map (POM). This model was then used to train the AI. Once trained, the AI outperformed humans in identifying new odors.

The model depends on the correlation between the molecules and the smells perceived by the study’s panelists, who were trained to recognize 55 odors. “Our confidence in this model can only be as good as our confidence in the data we used to test it,” said co-first author Emily Mayhew, PhD. Senior co-author Joel Mainland, PhD. admitted: “The tricky thing about talking about how the model is doing is we have no objective truth.”

The study resulted in a different way to think about smell. The Montell Center says:

The team surmises that the model map may be organized based on metabolism, which would be a fundamental shift in how scientists think about odors. In other words, odors that are close to each other on the map, or perceptually similar, are also more likely to be metabolically related. Sensory scientists currently organize molecules the way a chemist would, for example, asking does it have an ester or an aromatic ring?

“Our brains don’t organize odors in this way,” said Dr. Mainland. “Instead, this map suggests that our brains may organize odors according to the nutrients from which they derive.”

“This paper is a milestone in predicting scent from chemical structure of odorants,” Michael Schmuker, a professor of neural computation at the University of Hertfordshire who was not involved in the study, told IEEE Spectrum. It might, he says, lead to possibilities like sharing smells over the Internet.

Think about that.

“We hope this map will be useful to researchers in chemistry, olfactory neuroscience, and psychophysics as a new tool for investigating the nature of olfactory sensation,” said Dr. Mainland. He further noted: “The most surprising result, however, is that the model succeeded at olfactory tasks it was not trained to do. The eye-opener was that we never trained it to learn odor strength, but it could nonetheless make accurate predictions.”

Next up on the team’s agenda is to see if the AI can learn to recognize mixtures of odors, which exponentially increases the number of resulting smells. Osmo also wants to see if AI can predict smells from chemical sensor readings, rather than from molecular structures that have already been digitized. And, “can we digitize a scent in one place and time, and then faithfully replicate it in another?”

That’s a very ambitious agenda.

Dr. Wiltschko claims: “Our model performs over 3x better than the standard scent ingredient discovery process used by major fragrance houses, and is fully automated.” One can imagine how this would be useful to those houses. Osmo wants to work with the fragrance industry to create safer products: “If we can make the fragrances we use every day safer and more potent (so we use less of them), we’ll help the health of everyone, and also the environment.”

When I first read about the study, I immediately thought of how dogs can detect cancers by smell, and how exciting it might be if AI could improve on that. Frankly, I’m not much interesting in designing better fragrances; if we’re going to spend money on training AI to recognize molecules, I’d rather it be spent on designing new drugs than new fragrances.

Fortunately, Osmo has much the same idea. Dr. Wiltschko writes:

If we can build on our insights to develop systems capable of replicating what our nose, or what a dog’s nose can do (smell diseases!), we can spot disease early, prevent food waste, capture powerful memories, and more. If computers could do these kinds of things, people would live longer lives – full stop. Digitizing scent could catalyze the transformation of scent from something people see as ephemeral to enduring.

Now, that’s the kind of innovation that I’m hoping for.

Skeptics will say, well, AI isn’t really smelling anything, it’s just acting as though it does. E.g., there’s no perception, just prediction. One would make the same argument about AI taste, or vision, or hearing, not to mention thinking itself. But at some point, as the saying goes, if it looks like a duck, swims like a duck, and quacks like a duck, it’s probably a duck. At some point in the not-so-distant future, AI is going to have senses similar to and perhaps much better than our own.

As Dr. Wilkschko hopes: “If computers could do these kinds of things, people would live longer lives – full stop.”

Kim is a former emarketing exec at a major Blues plan, editor of the late & lamented Tincture.io, and now regular THCB contributor.

What’s a diagnosis about? COVID-19 and beyond

May 28, 2020• 2

By MICHEL ACCAD

Last month marked the 400th anniversary of the birth of John Graunt, commonly regarded as the father of epidemiology. His major published work, Natural and Political Observations Made upon the Bills of Mortality, called attention to the death statistics published weekly in London beginning in the late 16th century. Graunt was skeptical of how causes of death were ascribed, especially in times of plagues. Evidently, 400 years of scientific advances have done little to lessen his doubts!

A few days ago, Fox News reported that Colorado governor Jared Polis had “pushed back against recent coronavirus death counts, including those conducted by the Centers for Disease Control and Prevention.” The Centennial State had previously reported a COVID death count of 1,150 but then revised that number down to 878. That is but one of many reports raising questions about what counts as a COVID case or a COVID death. Beyond the raw numbers, many controversies also rage about derivative statistics such as “case fatality rates” and “infection fatality rates,” not just among the general public but between academics as well.

Of course, a large part of the wrangling is due not only to our unfamiliarity with this new disease but also to profound disagreements about how epidemics should be confronted. I don’t want to get into the weeds of those disputes here. Instead, I’d like to call attention to another problem, namely, the somewhat confused way in which we think about medical diagnosis in general, not just COVID diagnoses.

The way I see it, there are two concepts at play in how physicians view diagnoses and think about them in relation to medical practice. These two concepts—one more in line with the traditional role of the physician, the other adapted to modern healthcare demands—are at odds with one another even though they both shape the cognitive framework of doctors.

Grading the Federal Health IT Strategic Plan

Dec 22, 2014• 4

By JAMES SALWITZ, MD

It is a heart pounding, head spinning, edge of your seat page-turner; the sort of rare saga that takes your breath away as it changes you, forever. It hints at a radically different future, a completely new world a few years away, which will disrupt the lives of every man, woman and child. Available now, from the National Coordinator for Health Information Technology (ONC), Office of the Secretary, United States Department of Health and Human Services, is finally, without further ado; the Federal Health IT Strategic Plan 2015 – 2020.

You think I am kidding. A satirical dig at another monstrous, useless, governmental report? Absolutely not. The concepts outlined in this blueprint will transform healthcare. It is a tight, clear, document, which at only 28 pages, delivers almost as much change per word as the Declaration of Independence. This may be the most powerful application yet of computerized information technology.

If you want to know where healthcare and health IT are headed, The Plan is absolutely worth a read.

I have only one complaint; it is coated with too much sugar. Restricted by policy structure and jargon, the report does not go far enough.

Wait, Maybe Technology Won’t Replace Doctors After All!

Oct 2, 2014• 11

LISA SUENNEN

Such a good question from my friend David Shaywitz, MD, PhD, (and co-author with me of the book Tech Tonics). David has spoken and written about this this theme frequently, and most recently at the Health 2.0 conference held last week in Santa Clara, CA. He and I and 2000 of our closest friends were there to talk healthcare technology. Isn’t it ironic that it takes that level of human interaction to talk about the ways healthcare can disintermediate humans from healthcare?

What struck me so loudly at the conference was how easy it is for us all to forget how human the healthcare experience really is. I moderated and attended numerous sessions at the conference, each a twist on the theme of how technology can make healthcare delivery more accurate, more efficient, more effective than anything we have going today.

David participated in a session withMatthew Holt, Vinod Khosla and Dr. Jordan Shlain, who could not be farther part from each other on the topic of doctor vs. machine (David played the role of moderate guy in the middle), Mr. Khosla backed away or at least clarified his earlier statements about how 80% of doctors will be unnecessary in the coming new age of healthcare technology. His revision was that 80% of alldiagnosis will, in the future, be done by computers, not doctors, because computers are far better at seeing a holistic view of a patient and taking in all of the relevant data. He talked about how certain digital technologies can know everything about you, including when you are sleeping and when you are awake. It made me think that Santa Claus must be worried about being replaced by an app.

Diagnosis Is Not Therapy

Apr 22, 2014• 4

By Alan Spiro, MD

We all know “that patient” – the one we may dismissively label “noncompliant.”

The person with diabetes whose HA1C is consistently above normal limits – the one who swears, when confronted with the numbers (yet again) he’ll start eating right and using his insulin as prescribed.

And yet, month after month, the lab work tells a different story. We watch in helpless frustration as patients like these spiral downward, developing complication after complication.

I thought about “that patient” as I read a recent Wall Street Journal article describing Dr. Judith Hibbard’s Patient Activation Measure (PAM), which she and her colleagues at the University of Oregon developed some years ago.

First, let me say I greatly admire the research and work of Dr. Hibbard and her team; I believe that the PAM is a wonderful tool and a step forward in better understanding patients.

While the article, and Dr. Hibbard, argue that the use of the tool can better target the needs of patients – and I agree – I can’t help but worry that the entire premise that patients need to be “activated” misses a point.

Patients are people before they are patients.

We know that when people are sick, they are still part of their broader world of family, friends and finances. We also know that their social, spiritual and psychological selves are every bit as important, and as important to their “cure” as their activation as a patient.

I suspect that Dr. Hibbard would agree with me and even argue that the PAM reflects all of these factors.

PAM is accurately diagnosing the end state – how all these factors impact the patient and the patient’s ability to be involved in his or her own care.

I worry, however, that the PAM may be oversold by healthcare administrators who put it in place as a way of trying to address all the factors that affect patient activation.

Pathologizing the Human Condition

Sep 1, 2013• 9

By Dan Peters

The American Psychiatric Association recently published a new version of the Diagnostic and Statistical Manual (DSM). The DSM-5 is what medical, mental health, and chemical dependency professionals use to diagnose developmental, mental health, substance abuse and dependence, learning, and personality “disorders.” Now in its 5^th edition, the DSM was first published in 1952. At that time, the DSM was 129 pages containing 106 diagnoses.

Now, 61 years later, the DSM-5 consists of approximately 950 pages and roughly 375 diagnoses. The DSM-5, while researched far more than previous editions, is based on the medical model or the model of disease. Simply put, the medical model finds the causes of disease and illness and then prescribes a treatment to cure the disease or illness. This means a person has a pathology or pathogen that needs to be treated and cured.

The questions that eat at me during my day as a psychologist and at night as a person searching for answers are:

Is it possible to accurately identify mental health “issues,” “illness,” or “disorders?” versus extreme ranges within the sphere of the human condition?
Even if it is possible to identify these conditions, does it determine the course of “treatment” or “intervention?”
If so, is there a “treatment” for every identified “condition?”
Does it mean there is a treatment that works?
Do you need a diagnosis to get help?

Over the years, many have been critical of this approach to mental “health” issues. Referring to mental “health” is actually a newer name as people have historically been thought to have mental “illness.” This makes more sense for people who are unfortunately compromised by severe conditions termed schizophrenia, bi-polar (manic-depressive), and severe depression and anxiety. But does this make sense for children, adolescents, and adults who are challenged with some other, and possibly less severe, aspect of their functioning and development? Do all human problems warrant a medical or mental health diagnosis? When did a weakness become a “disorder” that requires “intervention” and/or “treatment?”

To be fair, the DSM provided structure and guidelines for approaching the complicated business of determining who had a “problem” that required help. However, it seems things have gone too far. Critics of the DSM believe that this latest edition has taken the business of diagnosing to a new level, one where approximately 50% of the population can be diagnosed with something. Critics also believe that this pathology finding approach supports the continued trend of medication prescribing as the number one mode of treatment, and continued trend of increased health care costs and premiums with increased utilization of individuals who need a “diagnosis” to meet “medical necessity” to receive services. What does that mean? It means if you don’t have a diagnosis, you don’t get help. It means you have to have a problem (pathology) to get help (treatment and intervention).

Without going into detail about some of the changes in the newest edition of the DSM, some diagnostic categories have been added and some diagnosis “thresholds” have been lowered. This means that you need fewer symptoms to “meet diagnostic criteria.” Here are some examples of concerns with the new DSM-5:

Temper tantrums will now be diagnosed as Disruptive Mood Dysregulation Disorder
Normal forgetting will now be diagnosed as Minor Neurocognitive Disorder
Gluttony will be diagnosed as Binge Eating Disorder
Grief will be diagnosed as Major Depression
First time substance users and college partiers will get a diagnosis of Substance Use Disorder
Everyday Worry will be diagnosed as Generalized Anxiety DisorderContinue reading…

Medicine in Denial: What Larry Weed Can Teach Us About Patient Empowerment

May 22, 2013• 9

By Leslie Kernisan, MD

[This post is the third and final part of a commentary on “Medicine in Denial,”(2011) by Dr. Lawrence Weed and Lincoln Weed. You can read Part 1 here and Part 2 here.]

It seems that Dr. Larry Weed is commonly referred to as the father of the SOAP note and of the problem list.

Having read his book, I’d say he should also be known as the father of orderly patient-centered care, and I’d encourage all those interested in patient empowerment and personalized care to learn more about his ideas. (Digital health enthusiasts, this means you too.)

Skeptical of this paternity claim? Consider this:

“The patient must have a copy of his own record. He must be involved with organizing and recording the variables so that the course of his own data on disease and treatment will slowly reveal to him what the best care for him should be.”

“Our job is to give the patient the tools and responsibility to organize the knowledge and slowly learn to integrate it. This can be done with modern guidance tools.”

These quotes of Dr. Weed’s were published in 1975, in a book titled “Your Health Care and How to Manage It.” The introduction to this older book is conveniently included as an appendix within “Medicine in Denial.” I highlighted it this section intensely, astounded at how forward-thinking and pragmatically patient-centered Dr. Weed’s ideas were back in 1975.

Thirty-eight years ago, Dr. Weed was encouraging patients to self-track and to participate in identifying the best course of medical management for themselves. Plus he thought they should have access to their records.

Zen and the Art of Charting

Mar 1, 2013• 23

By Leslie Kernisan, MD

One of the many challenges I face in my clinical work is keeping track of a patient’s multiple health issues, and staying on top of the plan for each issue.

As you might imagine, if I’m having trouble with this, then the patients and families probably are as well.

After all, I don’t just mean keeping up with the multiple recommendations that we clinicians easily generate during an encounter with an older patient.

I mean ensuring that we all keep up with *everything* on the medical problem list, so that symptoms are adequately managed, chronic diseases get followed up on correctly, appropriate preventive care is provided, and we close the loop on previous concerns raised.

This, I have found, is not so easy to do. In fact, I would say that the current norm is for health issues to frequently fall between the cracks, with only a small minority of PCPs able to consistently keep up with all health issues affecting a medically complex adult.

Doctors Make Mistakes. Can We Talk About That?

Nov 24, 2012• 4

[youtube width=”475″ height=”300″]http://www.youtube.com/watch?v=iUbfRzxNy20[/youtube]

The Last Well Child

Nov 8, 2012• 9

By Lawrence Rosen, MD

Q: “What is a well person?”
A: “A well person is a patient who has not been completely worked up.”

As I enter the exam room, a smiling 10-year-old boy greets me. Pete, my last patient of a long day, is here for his annual well visit. I chat with him about his life — home, school, nutrition, exercise, sleep, etc. — and I’m struck by something. Pete is really well. He’s well-fed (but not too much), active and well-rested, and, most importantly, he’s happy. He has not been to see me in an entire year, and only comes in for preventive health counseling. I think back on my entire day… and on my whole week. Pete is different from every other child I have seen this week. He is, in fact, the only truly “well” child I have seen in a long, long time. And I wonder — is he the last?

I’ve begun this post with a short riff on Dr. Clifton Meador’s satirical masterpiece, “The Last Well Person,” published in the New England Journal of Medicine in 1994. Meador profiles a 53-year-old man he imagines to be the last known truly “well” person in the U.S. in 1998. The patient is subjected to every known evaluation and found to be basically undiagnosable. I reflect on this story each day as I enter one examination room after another, visiting with patients (and their families) in my pediatric practice.

Sadly, the story of “Pete” is real. I no longer see many well kids even though I am a primary care pediatrician, dedicated to keeping kids healthy. Yes, I devote much of my time to counseling parents about lifestyle choices (e.g., nutrition, exercise, play, rest, sleep) to promote wellness and prevent disease. Still, each and every encounter must be “coded” with a numerical set of instructions based on diagnoses (associated with disease states) so that I can get reimbursed for the care I deliver. My ability to keep my office open (so that I can continue to try and help families keep their children healthy) is predicated on my skill in playing this diagnostic code game.