Category Archives: science

An Inside Job. [Weekly Head Voices #33]

Posted on November 21, 2010 by cpbotha| 6 Comments

Hiatus: temporarily over.

My sleeping patterns are not quite what they used to be, mostly due to the latest manifestation of our little gene pool over here. Added to that, I’ve been really busy. Added to that, work has been throwing unnecessary curve-balls that have done their part in keeping me (pre)-occupied. In the end, a dash of perspective, several extremely wise friends and a generally sunny predisposition go a really long way, so here I am. :)

In this post, I’m going to touch upon the highlights of the past seven weeks (40 to 46) and I’m going to do so with bullets, seeing as I’m in a bullety-kind of mood.

  • I spent a week in Stellenbosch teaching Information Visualisation! Read all about it in this special blog post.
  • The week after that my mom came to visit us. She really has the best genes.
  • I met my new work laptop: A Dell Latitude E6410 with Core i5 M540 2.53GHz CPU, 4G ram, 500G HDD, NVidia NVS 3100m GPU with 512 MB video mem. It’s a 14″ latop (15.4″ is a completely useless format, flame me in the comments), but due to the materials used quite heavy. I like it!
  • Another one of our STW NIG research proposals has been granted. I conceived and developed this one together with colleagues from the LUMC Departments of Anatomy, of Surgery and of Orthopaedics. The title is High-definition Atlas-based surgical planning for Pelvic Surgery and we get to combine high-resolution human histology, mega-volume rendering, , biomechanical modelling and surgical simulation. With this grant we can appoint two more Ph.D. students (one in my group, one at the LUMC) for four years to work on this!!
  • The week after that I went to Salt Lake City (Utah, US and A) to attend the conference previously known as IEEE Visualization. It was AWESOME! (blog post half-written, will soon publish).
  • The week after that I had a really cool dream: I was playing high-speed catch with a bunch of people outside. The ball was randomly changing shape between rock, papers and scissors. If you wanted to catch it, you didn’t only have to be at the right place at the right time, but your hand also had to be in the right complementary configuration to catch the shape-shifting ball.
  • Now that we’re talking about dreams: I finally saw Inception and was completely blown away. It’s not about being complicated, it’s about being a well-told story and a fantastically filmed movie. What I positively adored, is the fact that Nolan doesn’t require much: He only needs you to believe that dreams can be shared, without giving away too many of the details of the mechanism, and then builds a marvellous story on that canvas. My absolute favourite scene was the waking up on the plane, just before the landing, almost at the end of the film. I loved the way in which the characters looked at each other, and the possible interpretations of their expressions.
  • On the topic of the Underworld gig in the Heineken Music Hall: I hope I have my voice back before my morning lecture tomorrow morning. Thank you Fantastic Friends!

That’s it kids. If you’ve come this far, you’re now mostly up to date. Please leave me a comment, it’s good for my serotonin! I hope you have a great week. To get you off to a  good start, here’s a music video showing what an infectious idea could look like…

→ 6 Comments

Posted in life, science, weekly head voices, work

Tagged , , , ,

Teaching InfoVis in Stellenbosch

Posted on November 21, 2010 by cpbotha| 1 Comment

In week 40 (that’s Monday October 4 to Friday October 8 for those of you not so much into week numbers) I had the privilege of giving a week-long Information Visualisation course to a group of post-graduate students (a mix of B.A. Honours in either Socio-informatics or in Decision-making and Values Studies) at the Centre for KDD of the University of Stellenbosch in the building previously-known-as “The BJ”.

With this post, I want to summarise, extremely compactly, my impressions:

  • In spite of the odd publication at an InfoVis venue, and in spite of putting significant effort into broadening my horizons also into InfoVis, I’m strictly speaking still a Scientific Visualisation (SciVis) guy. In this context, developing and giving an InfoVis course was a fantastically educational experience. I aspire to be equally fluent in SciVis and InfoVis when I grow up.
  • I designed a course based on a mix of lecturing, paper reading and discussion and hands-on exercises. I can now highly recommend this combination. Next time I will put more effort into involving everyone more actively in the discussion. What also seemed to work well was the course website that students could add their work to during class.
  • I sourced almost all of my material from generous InfoVis colleagues, primarily Dr. Tamara Munzner, but also Dr. John Stasko, Prof.dr. Jarke van Wijk and Dr. Maneesh Agrawala. Thank you!
  • Having spent 7 years at the Stellenbosch Engineering Faculty (1500 guys all wearing t-shirts in jeans; furthermore, social skills are frowned upon) and the past 10 years at the TU Delft, a kind of Ultra-Engineering-Faculty, spending the week at The BJ interacting with socially adept and skilled communicators was a truly interesting experience.
  • I was otherwise also impressed with the level of cooperation between the students and their willingness to work hard (and long).
  • Laptop-use during lectures in a relatively small class: It’s an interesting phenomenon, having to deal with your own little communication backchannel. Banning laptop use is obviously not an option as there are too many possible advantages, but I’m not sure yet what a good solution would be towards making sure that the laptops are being used to augment learning.
  • Most students chose ProtoVis (from a list of 8 possibilities, and they were free to use anything else too) to implement their final mini-projects with. This was probably due to their previous web-programming experience, and also that ProtoVis requires more or less zero setup.

Conclusion: Much learned, hopefully the students did too. :) All in all a positive and energising experience!

→ 1 Comment

Posted in science, work

Tagged , , , , ,

An indecent proposal. [Weekly Head Voices #31]

Posted on September 20, 2010 by cpbotha| 12 Comments

Dear readers,

I would like to introduce you to my new friend:

Pretty Leifheit kitchen timer, ideal for the Pomodoro Technique. Also the first time I use the macro setting on my Canon. Doh.

It is old-fashioned and mechanical. It makes an extremely comforting ticking sound, and then after the 25 minutes of focus-time are over, starts ringing. The ticking is not too loud, and not too soft. The ringing is just the right length. You don’t take my word for it, as I’ve made you a short movie clip:

YouTube Preview Image

Comforting, no?

In spite of a number of deadlines having been successfully met, I’m still in the lamentable situation where every second counts, so I’m going back to bullet mode for the rest of this post:

  • The FNSF and I managed to finish that Indecent Research Proposal and submit it at 11:57 on Wedensday, 3 minutes before the Deadline for Indecent Proposals at the Dutch Science Foundation (NWO). We almost missed the deadline due to the extremely primitive web-submission system in use there. If you happen to be invited to review this proposal, please approve it. You Can Trust Us. :)
  • The real-time collaboration functionality in the new Google Docs was a life-saver during the writing of this proposal. Being able to see FNSF’s cursor move and edits happen in real-time was really great for coordinating. Now if they could only improve: 1) The styling possibilities (there are none at the moment, so you have to manually change typesetting if you don’t agree with their default style) and 2) The PDF export (it inserted page-breaks within tables and generally screwed up. Eventually I exported to MS Word, fixed typesetting and page-breaks there, then exported to PDF); then I would be happy.
  • On Friday I had one of the most hectic oral defences of my life. There’s a high probability I’ll have to try again in a year or two’s time. Or never. I’ll keep you posted. Or not.
  • Partly due to the events of Friday, but also to continuous general self-reflection (read: voices in my head that never sleep), I will gradually revise my publication strategy in the coming years. Bottom-line: I have to focus more on generalising until our applications start looking like pure theory. :)
  • I’m slowly learning how to say no. It’s hard, but at least there is some professional help available:
YouTube Preview Image

In a few weeks, I hope to have the time to write you a proper post. Until that glorious occasion, I’ll do my best to keep you up to date with these healthy and low-fat WHV light editions!

→ 12 Comments

Posted in life, science, weekly head voices, work

Tagged , , , , , ,

The Human Animal Post

Posted on August 1, 2010 by cpbotha| 16 Comments

(This post has a point. A very important point if I might say so myself and I’m even skipping the Weekly Head Voices because of it. Please read it, in sessions if you have to, from start to finish. It has a WHV Nerd Index of 0/5 and a Backyard Philosophy Index of 5/5. You can get back at me in the comments.)

It turns out that when any normal human being is faced with observations or evidence that oppose their already formed opinions, they tend to ignore or downplay the value of those observations. Conversely, any scrap of evidence that seems to confirm the opinion in question is considered to be good and trustworthy evidence. This is called confirmation bias: You and I both suffer from it, and it can be a dangerous phenomenon. Ideally, we would be able to judge the evidence and come to a reasonable decision, but this turns out to be exceptionally hard.

A photograph I recently took of a confirmation bias. Note that observations from below confirming preconceived ideas are favoured more than observations from above, resulting in the tell-tale yellow tinge and staircase edge at the bottom of the core.

Another interesting one is the planning fallacy: We are apparently hard-wired to underestimate the time we’ll need to complete some or other task. In other words, you always think you’ll need far less time to complete that project than you’ll end up using. Most of you have experienced this first-hand, or indirectly, when some huge IT or building project falls way behind schedule (and budget).

There is a whole list of such cognitive biases. As I’ve mentioned before, we’re basically walking jugs filled to the brim with misunderstandings, and mostly we’re unaware of it. Now you might think to yourself: But surely I’m better and more logical than the rest! I’m sorry to have to disappoint you, but you’re even biased about your biases. It’s called the bias blind spot, a meta-bias that means you’ll always estimate your own unbiasedness more highly than that of your neighbour.

I hope you didn’t nod off right there, because the next topic I’d like to touch on is that of sleep (please excuse the lame joke, I needed it for the continuity). Ever thought of exactly why you get sleepy at night? Or why your teenager (or teenage sibling, or yourself; substitute whatever’s more relevant) is not able to go to bed or wake up on time? Mostly we just go through the (sleep) motions without asking why or how. When you’re exposed to light in the morning, your eyes talk to your supra-chiasmatic nucleus (SCN), a  cluster of brain-cells about the size of a grain of rice in your already small hypothalamus, which then, minute as it, starts raising your body temperature, releases the cortisol hormone (this gives you a boost) and stops the release of melatonin. When it gets dark, the SCN throws a switch which causes the pineal gland (itself just pea-sized) to start producing melatonin, another hormone, but one that makes you feel generally less alert and altogether sleepy, so you start thinking about that wonderfully fluffy and soft bed of yours, and it gets harder and harder to stay awake.

I find that really fascinating: There it is, the SCN, a rice grain sized clock that orchestrates the daily rhythm of your whole super-complex body! This helps us to explain to our toddlers why they should go to bed earlier: Because your melatonin is activated much earlier honey! (There’s no arguing with that, even when you’re 4 years old.) It also explains teenagers’ sleeping habits: Adolescent hormones and life-style seem to interfere with melatonin production, and so teenagers get their melatonin kick hours later than adults.

The hypothalamus consists of more interesting nuclei than just the SCN. The arcuate nucleus also lives there, and it’s one of the spots in your brain that produces dopamine. Dopamine rules your life. It plays an important role in just about every bit of your humanity that you care about: Motivation, punishment and reward, sleep, mood, attention, sexual gratification, learning, and so on. Dopamine is instrumental in drug addiction, and, perhaps unsurprisingly, also in ensuring that one human becomes completely and irrevocably obsessed with another when falling in love. Serotonin and norepinephrine are also involved, but dopamine is the one that makes you go completely crazy. Once you get past the crazies and you have the good fortune of finding yourself in a longer term relationship, guess what? Once again, there’s not really that much magic or even that much mystery: Oxytocin and vasopressin, two hormones that also act as neurotransmitters, help to ensure that you experience the warm-fuzzies when with your partner and hence facilitate the pair-bonding experience. Oxytocin also plays a crucial role in breastfeeding (our friend the hypothalamus notices sucking at the breast and kicks the oxytocin production into gear that gets the milk going to where it needs to be). Conveniently, it probably also plays a role in mother-infant bonding.

I could harp on for a long time about how exactly all of these (and many other) little tidbits determine your life, but by now, I hope that I’ve said enough and that you’re asking: What’s your point Vanessa?!.

My point is that you’re an animal. A cool kind of animal called the human animal, but an animal nonetheless. You’re a relatively complex machine, but a machine that can be studied and that can be understood. We humans have become incredibly skilled at picking apart the human machine. We have tools that can see which part of your brain is active during what activity, we even have tools to make the cells in your body glow in the dark when they’re being naughty. We still have much to figure out, but we know an astronomical amount more than we knew 50 years go. The points above are a microscopic example of recently acquired knowledge. Every day, the pile of Stuff We Know is growing more quickly, and the pile of Stuff We Don’t Know is shrinking more quickly. It’s a terribly exciting time to live in.

My actual point is: You’ve been given one of these human machines to drive. For life. You’ll be getting only this one, and you’ll be taking it to some strange places, even off-road at times. You’ll make a number of important decisions that deeply affect you and your machine and even the other machines around you.

Don’t you think that you should read the manual?

→ 16 Comments

Posted in life, science

Tagged , , , , ,

VCBM 2010 [Weekly Head Voices #26]

Posted on July 8, 2010 by cpbotha| 1 Comment

(This post is a slightly longer than average report detailing our trip to the EG VCBM 2010 conference. It’s of course super-entertaining, but if you still do wish to skim through it, I’ve bolded the per-paragraph themes. If you’re not sure what these danged conferences are about, see my recent EuroVis 2010 post for a general introduction.)

Last week, I accompanied Peter Schaafsma (he of the orbital fat mobility paper), Bastijn Vissers and André van Dixhoorn (they of the resting state fMRI brain connectivity paper) to Leipzig, where they had been selected to present their work at the second Eurographics Workshop on Visual Computing for Biology and Medicine (VCBM).

Pretty VCBM logo.

Things got off to a great start when, as we were travelling there by speeding bullet (okay, it was just a brand-new rental Opel, but the Autobahn turns any car into a Speeding Bullet!), we managed to strike, at high speed, a high-quality German plastic bucket that had suddenly appeared right in the middle of the road. After a few more kilometres of noticing that our speeding bullet was not able to pass the very slowly accelerating bullet stage and was making strange disconcerting noises to boot (excuse the automotive pun), we stopped to investigate, noticing to our shock that the bucket, having been very badly burnt, Mustafa-style, was still lodged under the car.

After carefully dislodging the remaining half of the bucket, we were even more shocked to notice that the car was dripping concerning amounts of liquid more or less from the spot where the bucket had been stuck. Further on-site and online investigation by the crew brought to light the following observations: 1) The liquid was not hot, and so probably did not come from the engine. 2) The liquid was tasteless (don’t ask). 3) The air-conditioner had been running all the time, and certainly would have to get rid of water condensate from the cooled air. This latter observation was to us not immediately obvious, but now is, and hopefully to you as well: Automobile air-conditioners often get rid of condensed water through outlets under the car. Phew.

Our speeding bullet tore through the remaining 200 kilometres in record time, where my man Helmut from Vienna and these wonderful brain-boosters were waiting for us on the Nikolaikirchhof:

Leipzig's famous 1 litre brain boosters. Image deliberately deFaced. The arms in the photo might or might not belong to anyone that you know.

The conference kicked off the next morning bright and early with a brilliant keynote by Prof. Anders Ynnerman.  This was related to the great talk he gave in Delft the week before, but even better, as, amongst other things, he had had the Virtual Autopsy multi-touch table shipped all the way to Leipzig to be able to demo live during his presentation. He even managed to undo the evil our projector in Delft had wrought on his laptop’s colour profiles and finally showed his visualisations in their full glory on the projector in the Mediencampus Villa Ida.

The second keynote, given by Dr Roland Bammer, focused on their work on eliminating motion artefacts in brain MRI, both by image post-processing but also, and this is the really cool bit, by mounting a special marker on the forehead of the subject that allows real-time 3D motion tracking and linked real-time low-level correction of the MRI acquisition process. They are currently working hard on getting their tech into users’ hands. Help them by bugging your local Radiology Department about this! :)

The multi-touch virtual autopsy table was available throughout the conference to try out. Two observations: 1. Sometimes the glass surface of the multitouch.fi table is harder to work with than the MS Surface's matt finish. 2. It's hard to use the table in environments with too much lighting.

The rest of the scientific program was perfectly varied, consisting of paper presentations, an invited talks session and a posters session, including a plenary fast-forward where each poster author was given three minutes to promote their work.  What really shone throughout the two days, was the superb organisation in Leipzig: Thanks to Dr Alex Wiebel and colleagues, the conference was a text-book example of conference organisation, with each event occurring at exactly the right time, not a moment too soon or a moment too late.  The Mediencampus Ida Villa was the ideal location for the conference, not in the least due to its air-conditioned auditorium shielding us from the more than 30 degrees Celsius outside temperatures. Another important manifestation of the superb organisation were the magical coffee breaks, always taking place exactly when you needed them, with copious amounts of cookies to boot!

Erik Pernod and Herve Delingette won the VCBM 2010 Best Paper award with their paper titled Interactive real time simulation of cardiac radio-frequency ablation, really great work combining elements of simulation, visualisation and a clear clinical application.  The best paper committee had a relatively easy choice, as the reviews (3 to 4 per submission) were also unanimous about this paper’s ranking. As an added bonus, the work is available within the open-source SOFA framework.

My personal and completely biased Head Voices VCBM 2010 Best Talk Award however, goes to the invited talk by Marc Streit and Alexander Lex on Caleydo: Visual Analysis of Biomolecular Data. The presentation was an extremely entertaining show by two skilled speakers, striking just the right balance between focus and variety as they demonstrated several aspects of their Caleydo visual analysis software.  At several points, for example the explicit visual linking of different heterogeneous data sources (paper here, youtube video here, see here for other publications.), I had the typical reaction to good scientific contributions: Now why didn’t I think of that?!

Three to four of the VCBM 2010 papers (this includes the best paper of course) will soon be selected to submit an extended version to the forthcoming special issue in the Computers and Graphics journal on Visual Computing for Biology and Medicine.  Very importantly: Everyone is welcome to submit an aptly themed paper to this special issue! See the C&G special issue on VCBM call for papers for more information.

On a more personal note: This was the second VCBM workshop, with the first having been held in Delft in 2008, organised by yours truly. It was really great in Leipzig experiencing the growing community around this event and especially connecting with some of my favourite people. I’m greatly looking forward to future VCBM workshops, which makes it doubly cool that  VCBM 2012 will be hosted by Prof. Anders Ynnerman in Linköping (Sweden) and that after that, it will become a yearly phenomenon.

Will you join us?

→ 1 Comment

Posted in science, weekly head voices, work

Tagged , , , ,