False Positives and Unintended Consequences

After an appendectomy, the removed appendix is examined to determine whether appendicitis (or some other problem) was present. In a noticeable proportion of people, it is not. This is not a problem, for a good reason: the consequences of untreated appendicitis can be swift and catastrophic. The consequences of appendectomy are, in general, mild by comparison, at least in societies with good sanitation. (There is always risk from surgery.)

This is being revised after trials with intravenous antibiotics have shown very high rates of survival, but appendectomy was introduced before the antibiotic era, many organisms are involved, and we’re losing antibiotic efficacy – plus appendicitis can still progress after IV therapy – so surgery still needs to be in the toolkit. Still, knowing that there are options well worth trying, and with a good track record, is good news for situations where surgery may not be available, practical, or advisable.

[T]ypes of patients in whom appendectomy might be avoided:

  • Patients with an appendiceal abscess, who would be better treated with percutaneous drainage;
  • Patients who have had a recent myocardial infarct;
  • Patients with severe lung disease;
  • Women in the first trimester of pregnancy; and
  • Persons in a remote environment such as Antarctica or on a mission to Mars.

From Evaluating Acute Appendicitis: Does Everyone Need an Operation? (subscription required)

(Mars – or the Antarctic – is hardly the only remote environment of interest; one of the studies was in Navy personnel on a submarine.)

Appendectomy is a classic teaching case on the value of ending up with a few false positives. It’s better to perform a low-complication procedure a few extra times than to have people walking out of the ER and dying when they get home. But as technology changes – in this case, antibiotic treatment options and CT-scan evaluation – it’s always good to re-evaluate even the obvious “tried and true” approaches, partly to see if there’s a better way across the board and also to address the situations where the outcome of surgery was likely to be bad. On the way, we seem to have discovered some properties of the appendix that can help us understand the environment for other GI disease – a new teaching case for this much-maligned “vestigial” organ!


From The Awkward Yeti

We are probably right to remain suspicious, though. The appendix may well have more tricks tucked into its submucosa.

Spy Society

Computers were invented as crypto-ware and spy-ware and control-ware. That’s what Alan Turing was all about. That’s where computing came from, that’s the scene’s original sin, and also its poisoned apple. —Bruce Sterling, The Ecuadorian Library or, The Blast Shack After Three Years

Update August 6: Danny O’Brien comments on Sterling’s position, with some background.

Trying Out the Nike+ FuelBand

I am using the Nike+ FuelBand this week as part of an activity for a conference I’m attending. I like gadgets and already wear a pedometer, and I’ve been curious about the FuelBand. It’s been an interesting few days.

Nike hints that the FuelBand is something special, with lots of different ways to evaluate your activity, but it’s mainly a wrist-based pedometer. As a consequence, it can’t always capture your steps and records “steps” for rhythmic arm motions. (I have worn it overnight a few times and “learned” that I took a couple of dozen steps in my sleep, which I definitely didn’t, although I do move around a bit.)

Nonwalking activities are represented as “steps,” generally yielding low activity estimates. Rowing on a machine or riding a bicycle shows strokes as steps, and it acknowledges stroke rate – the bike is more ‘caloriffic’ than the rowing machine – but can’t understand resistance. (Nike claims its logic is more than just pedometry, but I have kept wearing a pedometer, too, and it looks like pedometry to me.) Of course, those activities are better measured with other devices, but it puts the active user in the position of having having irrelevant information when thinking about their overall activity for the day.

The display is simple and narrow, with a few measures (steps, calories, and “Nike Fuel” – a weighted calculation that seeks to provide a common measurement unit across individuals and activities). It also serves as a watch. The display includes a line of tiny dots that light up as you progress toward your goal for the day, with the first dots in red, the middle dots in yellow, and the final dots in green. The device itself is simple and attractive, and sits unobtrusively on the wrist. I placed it on my ankle for bike rides. It is too small to clasp over my ankle but the shape is sufficient to hold it there as a cuff-style bracelet.

FuelBand offers no way to designate a start and stop for a specific activity, has no stopwatch, and displays only a few built-in measures. The web interface and iPhone app show an hour-by-hour breakout of your activity level on a graph, but they don’t show start and stop times or splits. In general, though, the app and web interface are pleasant to use, with stylish visualizations and cute reward sequences for meeting goals. You can also sync to your iOS device anytime throughout the day, which makes the device more engaging and interactive than a simple pedometer.

On setup, FuelBand invites you to set a daily goal, but changing it only takes effect the following day. I first selected 3000 points as a goal (defined by the system as an active day). I swamped that on a normal work day and selected 5000 points the next day, exceeding that as well, but when I decided to have a slow day, I was out of luck; I’m now wearing a device that is chiding me with little red dots for being lazy all day just because I didn’t plan my laziness last night. This merely annoyed me as a motivated, athletic person, but I wonder if it would make a less combative person take the FuelBand off and put it in a drawer.* Battery life is rated at about 4 days, but I find the battery is down to around 20% at the end of single day. In short, I’m getting lots of cues that this device was definitely not designed with me in mind.

I’m glad I had the opportunity to test the FuelBand. Nike has promoted it heavily with its Nike aesthetic, which is much higher energy than the apparent target audience, and the promotion definitely activated my covet drive. FuelBand is a good choice for someone who is already using iOS (the only OS the app supports for now), wants a set-it-and-forget-it tracking device, and just wants some reinforcement for moving around more during the day. It’s particularly bad news for the Jawbone UP, which suffers from much less name recognition combined with a recent episode of major manufacturing trouble. If you’re an active athlete, you’re better off sticking with your sport-specific devices. But it’s a nice gateway device for people looking to get more movement into their daily lives, just as walking is a good gateway activity.

*There are lot of good reasons to decide on the fly to have a slow day – illness, injury, or a dramatic change in plan, for example. The user can add a brief note and choose a mood icon in the web interface, but the purpose of the red dots on the progress display is to remind you to shake a leg. If the developers are already counting on the user to be susceptible to being nudged, they can expect the user to feel criticized, too – and not best pleased if it feels unfair.

Poppin’ Fresh Android Phone

Straight from the uncanny valley to you!

Like the full-sized Telenoid robot, the Elfoid handset is designed to add an element of realism to long-distance communication by recreating the physical presence of a remote user.

More

Gawker Redesign

I don’t have much to say about the visual design of the Gawker sites. I am a regular visitor of only one of them, and I usually visit from a desktop computer. I don’t have any trouble getting around the new layout. But Gawker made one big error, and that’s in the functionality for directing visitors who link to specific articles on a mobile device (at least, on iPad and iPhone): visitors ends up at a listing of headlines, which may or may not contain the headline that interests them. If they even know what that headline is, since they may have arrived from a shortened link in a Twitter message, introducing the article with a cryptic remark.

It’s fine to say, “Keep your hair on. They’re working on a fix.” or even “Sounds like you follow faux-clever jerks on Twitter.” You’re entitled to that opinion. But a basic principle of sound Web design is to make sure the user always has a “scent of information” to follow. If users find themselves someplace unexpected, a good design will help them on their way. And that’s just for people navigating the site. If they’re following links to specific pages, getting them there should be a no-brainer.

If a person follows a link to a specific page in your site, it’s just silly to think it’s perfectly fine to send them anywhere else. If your developer knows enough about the device making the request to shunt it to a different layout of the site, the site should be capturing enough about the link the user selected to get all the way there. If it dumps the user on a TOC page, your developer simply didn’t complete the job. And if hash-bangs, or whatever the new hotness is, don’t work well enough or consistently enough with the major pathways into your site, then maybe you should resist the temptation. Who knows? If an iPhone can’t find your page with your newfangled whatsit, maybe Google can’t, either.

Those of us who have been using mobile for a long time are familiar with this half-assed approach. We’ve been seeing it on television and newspaper websites for years, going back long enough that some of us could kind of understand why a Web team’s use cases didn’t capture us. But that’s not the situation today, even for those legacy outlets. So why would a new-media darling, which surely has a massive base of users on the current It Device, whatever that may be, repeat such a classic old-media mistake? Engaged audiences already greet redesigns with suspicion—why not take the time to make sure the functionality is solid?

When the ergot hits the rye

At some point in my adolescence, I was brought along to some kind of poetry group and someone read a poem with on the theme of ergot poisoning from grain. I have no other memories of the group at all, but I was reminded of it by this photo:

So, there it is. That’s what it looks like.

That, and, maybe, the Salem witch trials.

Beckoning Skynet

Beyond robots that think about what they are thinking, Lipson and his colleagues are also exploring if robots can model what others are thinking, a property that psychologists call “theory of mind”. For instance, the team had one robot observe another wheeling about in an erratic spiraling manner toward a light. Over time, the observer could predict the other’s movements well enough to know where to lay a “trap” for it on the ground. “It’s basically mind reading,” Lipson says.

What could go wrong?

Read the rest of the article

I Hate This Thing

This is MOWGLI, in a video from a few years ago. I don’t know the state of the art in this technology, but I am sure this represents something great in engineering. And my highly visceral response to this thing is NO. NO NO NO NO NO.

Cat jumps in my lap? I love it, wooza good kitty, how about a treat. But I think I might be one of those people who want robots to stay where I can keep an eye on ‘em. I mean, look at that thing. You know it won’t be content to require getting power via a cord forever. It just wants to sit in your lap now, but next thing you know, it’ll be leaping onto your bed, injecting you with some kind of paralytic compound, and using your body to charge its battery.

Read more

I have mixed feelings about the Boston Dynamics Big Dog robot, too, but about 43 seconds into this video, someone tries to kick it over onto its side, and the way its little legs buckle to keep its balance inspires sympathy in me.

Produce MRIs

MRI machines have to be tested and calibrated, and Andy Ellison puts produce into the machine he works with to use in those tests. The results are amazing and wonderful, especially when they make you realize how little you’ve thought about familiar foods. It’s a no-brainer that onion and artichoke will be lovely and smooth and somewhat predictable, but what about watermelon [3MB image]?

Many more at Inside insides.

Watson Talks Trash

What is it like to have a physical body? I imagine it would be cumbersome. Useful for transport, I guess. I usually get wheeled around on this rolling desk. Of course my 2-ton megaprocessor is in tow somewhere as well, but I regard that as a nonessential appendage, like a tail. You don’t have a tail, do you? Oh that’s right, you lost your tail several hundred million years ago when you began walking upright and acquired that large frontal lobe. This reminds me of an amusing fact I observed the other day. Did you know the existence of the human race is the product of an evolutionary toss of the dice? Not of years of award-winning engineering and painstaking assembly, but of chance, completely fleeting and random. A blip on the screen. At least, on my old screen. My new monitor has lossless rendering and over two hundred thousand dpi. —Watson Sizes Up One of His Opponents before the Show

What do you say to a thing like that, except maybe “I know where your off switch is.”

Oh, here we go:

I don’t get this urge to taunt Watson. Didn’t any of these people see The Demon Seed?

Except Watson didn’t have to deal with that, and won. Now it’s off to medical school!