For millions of people with epilepsy, life comes with too many
restrictions. If they just had a reliable way to predict when their next
seizure will come, they could have a chance at leading more independent
and productive lives.
That’s why it is so encouraging to hear that researchers have
developed a new algorithm that can predict the onset of a seizure
correctly 82 percent of the time. Until recently, the best algorithm was
hardly better than flipping a coin, leading some to speculate that
seizures are random neurological events that can’t be predicted at all.
But the latest leap forward shows that seizures certainly can be
predicted, and our research efforts are headed in the right direction to
make them even more predictable. The other big news is how this new
algorithm was developed: it’s the product of a crowdsourcing
The teams analyzed a huge data set detailing the electrical activity in the brains of people while under evaluation for surgery to treat their epilepsy. They also had an even larger data set from studies with dogs, whose epilepsy closely resembles that seen in people.
The competition shows that sharing data to collaborate on complex
problems can yield clever and unexpected solutions. Also noteworthy is
that none of the winners were clinicians. This suggests that if we can
find a platform to engage players from other seemingly distant
disciplines, the chances are greater to find innovative, out-of-the-box
solutions to current challenges..
That’s why NINDS has launched iEEG.org to catalyze collaboration and
sharing of datasets, algorithms, and research tools for BigData studies
of epilepsy. On this website are almost 2,000 datasets freely available
for analysis, and there is already a user base of more than 670 active
collaborators. It will be exciting to see how long it takes to go even
beyond that 82 percent predictability..
This is what big data, ML and AI in general is all about. Integrating knowledge from data and wisdom of the crowds to bring in changes to benefit all species. Please don't confuse this with utopia and nirvana but think of it as a small step towards giving every living being a chance to live and enjoy the time on this blue planet pleasurable.
In 1958, New York’s modern master planner Robert Moses proposed to blast a highway through Greenwich Village, scattering its communities in order to make room for the inevitable technology of its day, the automobile. Moses had already built a highway through the Bronx, which never recovered from it. His plan for the Village was defeated by an alliance of local residents, including the urban philosopher Jane Jacobs, who articulated what would be lost in unforgettable terms: “the sidewalk ballet”, the dense web of glances, handshakes and hellos that constitutes city life at its most creative and fulfilling.
With digital technology today we are roughly at the stage we were with the car in the 1950s – dazzled by its possibilities and unwilling to think seriously about its costs, which is another way of saying we haven’t thought about how to maximise its benefits. Tools, whether they are made of flint or silicon, should be deployed to extend our potential, not erase it. Hunter-gathering has been revolutionised many times over but we still have the job of being human. It’s up to us to define the scope of work.
I had a quite personal reason to set out for these answers. One of my earliest memories as a kid is trying to wake up one of my relatives, and not being able to. Ever since then, I have been turning over the essential mystery of addiction in my mind -- what causes some people to become fixated on a drug or a behavior until they can't stop? How do we help those people to come back to us? As I got older, another of my close relatives developed a cocaine addiction, and I fell into a relationship with a heroin addict. I guess addiction felt like home to me. If you had asked me what causes drug addiction at the start, I would have looked at you as if you were an idiot, and said: "Drugs. Duh." It's not difficult to grasp. I thought I had seen it in my own life. We can all explain it. Imagine if you and I and the next twenty people to pass us on the street take a really potent drug for twenty days. There are strong chemical hooks in these drugs, so if we stopped on day twenty-one, our bodies would need the chemical. We would have a ferocious craving. We would be addicted. That's what addiction means. [---]
After the first phase of Rat Park, Professor Alexander then took this test further. He reran the early experiments, where the rats were left alone, and became compulsive users of the drug. He let them use for fifty-seven days -- if anything can hook you, it's that. Then he took them out of isolation, and placed them in Rat Park. He wanted to know, if you fall into that state of addiction, is your brain hijacked, so you can't recover? Do the drugs take you over? What happened is -- again -- striking. The rats seemed to have a few twitches of withdrawal, but they soon stopped their heavy use, and went back to having a normal life. The good cage saved them. (The full references to all the studies I am discussing are in the book.) When I first learned about this, I was puzzled. How can this be? This new theory is such a radical assault on what we have been told that it felt like it could not be true. But the more scientists I interviewed, and the more I looked at their studies, the more I discovered things that don't seem to make sense -- unless you take account of this new approach. [---]
This gives us an insight that goes much deeper than the need to understand addicts. Professor Peter Cohen argues that human beings have a deep need to bond and form connections. It's how we get our satisfaction. If we can't connect with each other, we will connect with anything we can find -- the whirr of a roulette wheel or the prick of a syringe. He says we should stop talking about 'addiction' altogether, and instead call it 'bonding.' A heroin addict has bonded with heroin because she couldn't bond as fully with anything else. So the opposite of addiction is not sobriety. It is human connection. When I learned all this, I found it slowly persuading me, but I still couldn't shake off a nagging doubt. Are these scientists saying chemical hooks make no difference? It was explained to me -- you can become addicted to gambling, and nobody thinks you inject a pack of cards into your veins. You can have all the addiction, and none of the chemical hooks. I went to a Gamblers' Anonymous meeting in Las Vegas (with the permission of everyone present, who knew I was there to observe) and they were as plainly addicted as the cocaine and heroin addicts I have known in my life. Yet there are no chemical hooks on a craps table. But still, surely, I asked, there is some role for the chemicals? It turns out there is an experiment which gives us the answer to this in quite precise terms, which I learned about in Richard DeGrandpre's book The Cult of Pharmacology.
The writer George Monbiot has called this "the age of loneliness." We have created human societies where it is easier for people to become cut off from all human connections than ever before. Bruce Alexander -- the creator of Rat Park -- told me that for too long, we have talked exclusively about individual recovery from addiction. We need now to talk about social recovery -- how we all recover, together, from the sickness of isolation that is sinking on us like a thick fog. But this new evidence isn't just a challenge to us politically. It doesn't just force us to change our minds. It forces us to change our hearts. Loving an addict is really hard. When I looked at the addicts I love, it was always tempting to follow the tough love advice doled out by reality shows like Intervention -- tell the addict to shape up, or cut them off. Their message is that an addict who won't stop should be shunned. It's the logic of the drug war, imported into our private lives. But in fact, I learned, that will only deepen their addiction -- and you may lose them altogether. I came home determined to tie the addicts in my life closer to me than ever -- to let them know I love them unconditionally, whether they stop, or whether they can't.
Writer and intellectual Susan Sontag, in her book Illness as Metaphor, wrote of this obligation to be sick in our lives. And she also wrote that to decorate our illness with metaphors and melodramas was to make matters worse. "Illness is not a metaphor," she wrote. "The most truthful way of regarding illness—and the healthiest way of being ill—is one most purified of, most resistant to, metaphoric thinking." For her, stripping illness of its storytelling power was a treatment. [--] The first treatment is one we administer on our own and continue to administer throughout illness. A symptom arises, and then we treat ourselves by deciding what we think about it. "It's nothing," we might say. Is a pain that we call "nothing" a metaphor? Sometimes this is enough, because sometimes—usually—the symptoms do indeed go away. The symptoms are symptoms of "nothing." But the first treatment is like standing on the tip of a pyramid. Hope, delusion, and reasoned expectation all meet in a point. Which of the three faces will you tumble down when you lose your balance? Of the three faces of that first treatment, hope is the most frightening. Delusion finds its end, somewhere. A doctor, a new symptom, a new pain. When you treat your illness with hope, you might never stop falling. My mother tried hope—which she later discovered was not actually hope at all—for a long time. Too long, some people have said to me. She "should" have gone to the doctor sooner. [---] When you're sick, you know it because your attention shows up, unbidden, somewhere strange on your body. In your aching forehead. In your laboring lungs. In a tumbling discomfort just below your navel. But many people who end up with a serious diagnosis had no symptoms. People fear seeing their doctor because they fear an unexpected pronouncement. The word origin of "diagnosis" is "to recognize." What is being recognized is the hope or curse instilled by a way of looking and thinking. When you find out you have or may have a secret asymptomatic disease, suddenly your attention shows up in the diagnosed location. You touch that part of your body. You look at it in the mirror. You think into it.
A question that is bound up in illness for us: Who's to blame? If the person who chooses to pray as treatment dies of cancer, is it their fault? If so, isn't the same true for someone who chooses chemotherapy for cancer and dies of cancer?
People will be quick to tell you that some attitudes toward health are "dangerous." This is true. They're all dangerous.
Between two cancers, my mother used a hormone cream to help her have sex more easily. Later, some people in my family suspected that this resurrected the first cancer. I have no thoughts either way about this. My mother was also depressed, she was constantly having dental work done, she didn't exercise often, and she ate a lot of sugar. These are all "reasons" why some people say she might have gotten cancer. Responsibility and its harsh twin, blame, are treatment for anxiety.
But what if we eat raw food? What if we drink enough water, if we take vitamins, if we sleep well, if we exercise, if we meditate, if we go on "retreats," if we take psychedelic plants, if we get massages, if we become vegetarians, if we eat more organ meats, if we force ourselves to laugh, if we take morning walks?
We try to avoid illness and treatment, and in avoiding it create a constant state of illness and treatment.
- More beautiful and insightful words from Susan Sontag's daughter Conner Habib here
The time will come when diligent research over long periods will bring to light things which now lie hidden. A single lifetime, even though entirely devoted to the sky, would not be enough for the investigation of so vast a subject... And so this knowledge will be unfolded only through long successive ages. There will come a time when our descendants will be amazed that we did not know things that are so plain to them... Many discoveries are reserved for ages still to come, when memory of us will have been effaced.
Take the case of São Paulo, where I've been working. It's gone from being Brazil's most dangerous city to one of its safest, and it did this by doubling down on information collection, hot spot mapping, and police reform, and in the process, it dropped homicide by 70 percent in just over 10 years. We also got to focus on those hot people. It's tragic, but being young, unemployed, uneducated, male, increases the risks of being killed and killing.
We have to break this cycle of violence and get in there early with our children, our youngest children, and valorize them, not stigmatize them. There's wonderful work that's happening that I've been involved with in Kingston, Jamaica and right here in Rio, which is putting education, employment, recreation up front for these high-risk groups, and as a result, we're seeing violence going down in their communities.We've also got to make our cities safer, more inclusive, and livable for all. The fact is, social cohesion matters. Mobility matters in our cities.
Combrisson and Jerbi note that this problem is well known to statisticians and computer scientists. However, they say, it is often overlooked in neuroscience, especially among researchers using neuroimaging methods such as fMRI, EEG and MEG.
So how serious is this problem? To find out, the authors generated samples of random ‘brain activity’ data, arbitrarily split the samples into two ‘classes’, and used three popular machine learning tools to try to decode the classification. The methods were Linear Discriminant Analysis (LDA), Naive Bayes (NB) classifier, and the Support Vector Machine (SVM). The MATLAB scripts for this is made available here.
By design, there was no real signal in these data. It was all just noise – so the classifiers were working at chance performance.
However, Combrisson and Jerbi show that the observed chance performance regularly exceeds the theoretical level of 50%, when the sample size is small. Essentially, the variability (standard deviation) of the observed correct classification rate is inversely proportion to the sample size. Therefore, with smaller sample sizes, the chance that the chance performance level is (by chance) high, increases. This was true of LDA, NB and SVM alike, and regardless of the type of cross-validation performed.
The only solution, Combrisson and Jerbi say, is to forget theoretical chance performance, and instead evaluate machine learning results for statistical significance against sample-size specific thresholds.