Coffee&Waffle Is Always Be With You

안녕하세요! Last time you heard from me I reported (literally, reported) on my homestay visit to Bonghwa and Gyeongju, which was the first weekend I had in South Korea. It's been two weeks since that that trip, and as I'm sure you can imagine, I have quite a bit to relay. This isn't aided by the fact that I lied to you all; I told you I would post a new blog later that week, and, well, yeah... that didn't happen. Ga Ram picked up the slack, but I apologize for any of you who were anxiously refreshing the page over the last 12 days in my absence, and any of you who feel personally insulted by this unfulfilled promise (I hope this doesn't apply to any of you, really). In the future, I think 10-12 days between posts is a more realistic goal, especially with two of us bloggerfolk instead of one.

Anyways. To the updates.

The Korean class is long finished, and the research is now almost two weeks old. The first week, my professor (Cho Hyung Joon) and both of his grad students were away at different laboratory elsewhere in the country for 3 days, and I was assigned to read several papers and a half a text book on MRI technique in order to get familiarized with the material. Then, this week, I was introduced to a matlab program that was written by one of his graduate students, Sohyun, and told that I would be running that program over and over to analyze some data. Yesterday, however, I was informed that my task would be different; instead, I would be writing my own code, which is a lot more involved... a LOT more involved. 

Allow me to explain (Warning: geek speak ahead!). There are three sets of experiments (organized by the day they were performed), and each of the experiments has an image associated with it taken by an MRI machine. The image is 64x64x64 voxel data, and a transverse slice (64x64x1) looks like a multicolored circle surrounded by blue. The circle is the (trabecular) bone, and the colors vary depending on the signal the MRI received over time. The original program that Sohyun wrote took one of these slices, created a noise filter by pulling up the image and manually selecting a region outside of the bone ('create mask'), then allowed you to either select a region within the bone and create a linear fit between frequency and time data. The slope of that line is (depending on which experiment was run) T2 or T2* data, and based on that data, we can determine which region is bone and which region is empty space (the very basics of MRI are covered here, if you're interested). I told you the original project was to run this program for all the experiments, but now it's a little different. I have been assigned (as of yesterday) the task of writing a code that will take the original voxel data, and create that linear fit for each individual voxel over time, giving me either T2 or T2*. Then, with the T2 or T2* data from the individual voxels, I will reconstruct the image using a black and white color scale corresponding to the value of those data. The end result will be a 3-dimensional voxel image of the trabecular bone sample, with dark regions corresponding to bone and white regions corresponding to the empty space occupied by water (or vice versa). Essentially, I'm building my own MRI image from the ground up.

This is a really exciting prospect... except for the fact that I really have no idea what I'm doing. I am a relative n00b when it comes to matlab, all the experience I have with it is only writing basic functions to graph things or fulfill stupid Engin101 projects. I expect to spend most of my time looking on the internet for examples of how to write proper loops, and asking my grad students to help me out with a lot of it. On the other hand, when I finish the project, I will have extensive experience in matlab, and I have a feeling this will come in handy with a lot of my future coursework, as well as any further research I do. Even if I don't finish the project entirely, I will have gained a lot of valuable experience being baptized by fire, so to speak. I have two weeks to finish that which I have no idea how to start. Wish me luck.

So you're probably wondering why the title of this (already somewhat lengthy) blog post is so peculiar. Well, you can read for yourselves, here:

This is, without a doubt, the best and most epic Engrish Fail I have seen thus far in South Korea.

That came from a restaurant called (you guessed it!) Coffee and Waffle, and I visited it in Daegu when I traveled there a weekend ago. Here's that story:

The pretty garden is there to distract you from the
pain your legs feel after all of those freaking stairs.

In the morning, Seong, Labiba, and I left Ulsan to Daegu via the KTX train. At the Daegu KTX station, we met Min Young, who is Seong's lab mate and Labiba's home stay person, and her friend (whose name I sadly forgot). From there, we went to the Daegu subway system, and then on to Woobangland, which is an amusement park in the city. Daegu is a notoriously hot city, due to the topography (it is surrounded by mountains on all sides, which through the magic of meteorology traps heat), and that Saturday was no exception. We miraculously got a student discount because we had our MCards on us, and Min Young was able to flirt with (or otherwise convince) the ticket man enough to let us go, despite our obvious foreign-ness. Once inside, we went to several amusements, including a descend-and-splash type ride (very nice considering the weather), a gigantic swing with a freely attached passenger compartment (so that we could go upside down when we went the full 2π radians) but only one true roller coaster: the Camel Back. It's called such because it has a few humps towards the end of the ride. I was able to sneak the flip cam on the ride, so if you're interested in seeing what a Korean roller coaster has to offer, see below (starring Min Young's hair! .. and a lot of wind noise):

125 meters in
125 pixels or less

The most interesting thing in the whole theme park was the "Sky Drop," a 130m descent from a tower. You are strapped into a harness, jump, and the first 100m or so of your descent are entirely freefall. It isn't a bungee jump, though; they merely slow you down at the end of your fall. It sounded like something I would have easily been game, save for the pricetag: 40,000₩ (30,000 if you went in pairs). That translates to about $40 to... fall down. I decided I'd save my money for better things.

If and when you see me wearing this shirt,
I encourage you to follow the instructions.

Like t-shirts! After we left Woobangland, we went to a local restaurant and feasted upon a gigantic collective plate of chicken, noodles, and spicy sauce (the name of which again evades me), as well as a visit to Coffee&Waffle, where we had Patbingsu, a Korean dessert with ice cream, fruits, sweet beans, and ice. After that, we debated on whether or not to go to a movie, but decided instead to go to a department store called Debec, a 12-story mega-shop in the middle of downtown Daegu. The first few floors were all girly things (jewelry, shoes, women's formal and casual, etc), then there were some sports apparel and casual wear on the next few floors, and men's dress clothes and shoes were on the top few floors, with a food court on the 12th floor. While I didn't find a crazy Korean hoodie that I have sought since my arrival, I did find this awesome t-shirt, which, I think, is an accurate representation of my personality and lifestyle.

It had been a long day already, so I grabbed a coffee before we went to a bar for a little while, so we could order a cheap pitcher of beer in anticipation of our final stop for the night: the club. After a little while at the bar, we walked around the corner to a night club called Gypsy (I think). Being in American clubs, I was hoping that the music that they played in Korean clubs was a little more techno and a little less Lil' Jon, and fortunately, I got exactly that. From 12-4am, we danced to some really good local DJ talent who spun nothing but infectious electro and deep house music – and yes, we did dance until 4am. There were 4 DJ's who were on stage while I was there, but my favorite was the last one, DJ Jini, holding down the spot for the late-night ragers. By the time she was halfway through with her set, however, we were all exhausted and decided to call it a night. We stayed the night at a "jimjiljang," which is a 24-hour establishment where you can walk in, shower up, keep your belongings in a locker, and spend the night on the floor. It was nice to be able to rinse off after drenching myself in sweat, but I'm still not accustomed to the whole sleeping on the floor thing, so when we woke up at 9:30 on Sunday to head back to Ulsan, I had probably gotten a total of 3 hours in between all of the the waking up and falling back asleep I had done. I was grateful to get back to the dorm room and get some real sleep for a little while.

I could discuss more, but I won't, because I've already taken up enough time and verbiage and it's ungodly early in the morning here (I should have been in bed like 3 hours ago... whoops). On my next installment, I'll give you that long-promised walking tour of UNIST, take you through a tour of the Hyundai Heavy Industries shipbuilding yard, and report on what other adventures I'll be taking in the... 2 weeks I have left. Wow. Time's been flying. Talk to you next time!