Finding Laura Ingalls

A Modern Prairie Girl

Lake Pepin: A House in the Big Woods

My first Little House site of the summer was Pepin, Wisconsin where Laura Ingalls Wilder was born. Seven miles outside of town sits the Laura Ingalls Wayside where there is a mock up of the Ingalls cabin located on the exact site where the Ingalls property was. There’s a lot less trees than when it was the Big Woods. The door on the cabin is locked with a latch string, as described in Little House on the Prairie. Inside the cabin there is a large picnic table hewed out of wood that would be good for a picnic lunch if I ever returned. There is a board along the bedroom wall with information on the lineage of both Charles (Pa) and Caroline (Ma). Half of the house is split into two rooms: the bedroom and the pantry.

The town of Pepin is not very big. There is a three room museum on Laura and the founding of Pepin. There are a few restaurants in town to eat at and the view of the lake is great on a blue day. If anyone ever comes to visit, I would not recommend more than a morning in Pepin unless you have a boat. There are multiple towns around Lake Pepin and those could surely turn it into a day. I would recommend visiting Stockholm, which is just the town over. They have a saying there: “Make Stockholm weird again.” There are plenty of little artsy places, fun shoppes, and a few great places to get baked goods. If you’re in the area you definitely need to stop in this cute down and explore!

Walnut Grove: On the Banks of Plum Creek

The trip to Walnut Grove has been a long awaited adventure. Even before I arrived in Minnesota, I was planning on coming here one weekend with James’ family. Some of James’ mom’s side of the family is from here, which made the trip even more interesting. The town is even smaller than Pepin from what I saw of it. Yet, every year the town puts on the Little House Pageant and attracts between 500-1000 spectators for every show.

The Little House museum here is much larger than the one in Pepin and is spread out between multiple buildings. You start in the gift shop where you can buy all sorts of pioneer paraphernalia, Little House books, and of course bonnets and dresses. I got the Little House Cookbook to use in the future. There is one building in the museum that is strictly dedicated to the Ingalls family. Half of the building explains their journeys and really gets into the history behind Laura’s time in Walnut Grove. They even have some of Laura’s old things there! The other half of the building is for the fans of the well-known Little House on the Prairie TV show. There is all sorts of collectors items, autographs from the cast, and pictures from the set.

There is a mock chapel that would have been similar to the one Pa helped buy the bell for. Inside of it there are a few pews and pump organs, one of which even works. There is a building that has a doll collection, a military history collection, and a collection of farm and household equipment from Laura’s time. Another building has multiple different scenes set up from Laura’s time in Walnut Grove. There’s a post office, a general store, a livery, and a Prairie Schooner wagon at the ready. There is a recreated prairie sod house complete with spiders! One would surely not notice a cow grazing overhead while in there. The final building of the museum is a recreation of the first floor of the Ingalls’ wood frame house fully furnished.

In the evening, we went to the site of the Dugout on the banks of Plum Creek. The Creek was too high to walk across because of all the rain, but thankfully there was a nice bridge to take to the site of the Dugout. The Dugout itself has collapsed in, but they were able to locate the site of it using old records. There were many people visiting the Dugout, some in full-out Prairie era dress. I of course had my bonnet on. 🙂 I think we should make bonnets a things again; they are so practical. There was also a very nice prairie walk behind the Dugout site on a restored prairie. We saw lots of butterflies and flowers.

That night we went to the pageant, which was at an outdoor theater just outside of town. The pageant was the highlight of the trip. The costumes were great and the actors all carried themselves accordingly. There were professional special effects (including fire!) and the sets were incredible. They pulled out full houses on wheels and assembled a church right before our eyes. If you ever find yourself in southwestern Minnesota in July I would recommend making a trip to Walnut Grove to watch the pageant for yourself, no matter how buggy it is!

A week of Haikus

This idea was inspired by an article I read on TED. For an entire week I will be writing a haiku every single day.

Day 1: 7/11

Lunch with friends
Green tea lemonade
Our dresses swing

Day 2: 7/12

An EEG Demo
My head is all gelled
No sounds play

Day 3: 7/13

It’s right there
The train speeds away
Time to wait

Day 4: 7/14

Day of rest
To catch up on life
A blessing

Day 5: 7/15

Lunch with royalty
I met Plack

Day 6: 7/16

Pass the time
Coding helps a friend
Lab’s empty

Day 7: 7/17

Birthday time
Can’t keep it secret
Happy birthday

Well after a week of haiku keeping, I can say it’s an interesting way to keep track of what happened in that day. You can see what was deemed important enough at the time to write a haiku, what brought some version of emotion to me. Haikus are pretty short and easy to write. They can be written while taking the train somewhere, during a five-minute break at work, or while watching Netflix in the evening. I often have difficulty keeping up a journal because I feel the pressure to write pages and pages about my entire day. A haiku is short and simple, even those that don’t write much and don’t like blocks of text can adapt to it. 10/10 recommend.

What am I doing with pitch and brains?

A Bit of Background

Pitch is a perceptual quantity relating to the periodicity of a sound. Another definition for pitch is the “attribute of auditory sensation by which sounds are ordered on the scale used for melody in music” (Oxenham 2017). A low pitch sound will have a long wavelength and sound deep like a base drum. A high pitch sound will have a short wavelength and be like a bird call. Wavelength is the distance from one peak to the next.

Low pitch sounds have a longer wavelength and are”deeper”. High pitch sounds have a shorter wavelength and are “higher”. A man’s voice is lower pitch than a woman’s voice.

Laboratory sounds are often categorized based on whether they are pure tones or complex tones. A pure tone consists of only one specific frequency and the waveform looks like the familiar sine wave. Pure tones are the most basic unit of sound and are not found in nature. If you have ever had a hearing test, then you have heard the pure tone “beeps” that an audiologist uses to screen your hearing. A complex tone is made by adding multiple pure tones together. Complex tones consist of multiple frequencies and are what we would call natural sounds. For more information on pure tones vs. complex tones look here.

Harmonic tones are those that share a common frequency called the central frequency or fundamental frequency (f0). Harmonic tones are considered pitch-evoking tones, meaning that one could pick out the pitch even when the fundamental frequency is not there. Inharmonic tones have no common frequency and also have a low perception of pitch.

In the inner ear there is a structure called the cochlea that looks like a snail shell. The cochlea is the organ of hearing and allows for the conversion of mechanical sound waves to electrical signals that can be sent to the brain. Within the cochlea sits the cochlear filter, or basilar membrane, which sorts the vibrations based on their frequency (whether they are high-pitched or low-pitched). Here is a video that demonstrates how the basilar membrane works.

Harmonic tones are specially coded in the basilar membrane. Lower harmonics (up to 8-10th) are distinguishable and put in it’s own “box”; these are resolved tones. Resolved tones are spaced differentially enough on the basilar membrane to fall into their own cochlear filter. Higher harmonics (above the 10th) are not as distinct and can sometimes be confused for one another. These higher harmonics are known as unresolved harmonics which means that multiple harmonics fall within the same cochlear band.

Harmonics that fall within their own distinct cochlear band are called resolved harmonics. Higher harmonics that incorporate multiple harmonics into the same cochlear band are called unresolved harmonics.

What are we trying to do?

Now that you have a bit of background on all the terminology, let me explain my current project in a bit more detail. Multiple studies have found that there are so-called “pitch regions” of the brain, regions especially skilled in perceiving pitch. They have found the antero-lateral (antero- to the front, later-to the side) region of Heschl’s gyrus is the primary location of these regions. Heschl’s gyrus is in the temporal lobe of the brain and is where the Primary Auditory Cortex is located. These studies have used many different types of pitch-evoking stimuli and control stimuli ranging from noise to harmonic tones. Many of these studies have used resolved (low-frequency) harmonic tones as pitch-evoking stimuli and compared them with unresolved (high-frequency) harmonic tones and frequency-matched noise that has no detectable pitch.

The red dot area where Heschl’s gyrus is located in the brain.

Few studies have compared the neural activation of inharmonic tones, which have an indiscernible pitch, to harmonic tones. The present study aims to determine if there are differences in neural activation for harmonic and inharmonic sounds and if these differences correspond to the previously reported “pitch-sensitive” regions. The test presents harmonic and inharmonic tones that are both high- and low-frequency to subjects while measuring brain activation using fMRI. We hypothesize that harmonic tones will have stronger representations in these “pitch-sensitive regions” of the brain compared to the inharmonic tones.

The designing of a test

What does my day actually look like you ask? I spend most of my day reading papers, doing documentation, and writing code on MATLAB. Plus I often shadow people in the lab as they do different things and run their own tests. My first couple weeks I spent most of the days building different iterations of the same test to carry out the task described above.

First I had to build the test stimuli which are shown in the figure below. There are six different conditions presented during this test and everything about them is mostly randomized to some degree. The six conditions are harmonic resolved tone complexes, inharmonic resolved tone complexes, harmonic unresolved tone complexes, inharmonic unresolved tone complexes, low-frequency noise, and high-frequency noise. The noise conditions are acting as our control conditions. The low-frequency noise is the control for the resolved tones and the high-frequency noise the control for the unresolved tones.

A frequency spectra of the six conditions. The low frequency conditions are shown in the first row and the second row depicts the high-frequency conditions. The first column corresponds to the harmonic tones, the second column the inharmonic tones, and the third column shows the spectra for the noise control.

Next, we had to build an actual behavioral test which the subjects could do while they are in the fMRI (functional Magnetic Resonance Imaging) machine. The test is all run off a MATLAB code that I have built and tweaked many times. 🙂 The subjects don’t see any of the inner workings; all they see is a window that displays messages about their progress and their score. For example it might display: “The score was 92%. The next block is 5.” There is a block that turns yellow to indicate that the sound is on and another block that turns green when a button is pressed.

Guided User Interface (GUI) for the Harmonic-Inharmonic Test. The white box displays text to prompt the participant and notify them of their progress. The sound status block turns yellow to indicate that the sound is on. The button press block turns green to indicate that a button was pressed.

Each block consists of only one condition. Within each block there are 25 tones. Each tone is made up of three 200-ms pulses of identical stimuli. Each tone has a different central frequency than the ones before and after it, unless it is a repeater. To keep the participant engaged with the sound and to measure their attentiveness, we have built in a one-back task. When the participant hears a tone that is identical to the preceding tone they press the button.

The stimulus presentation for a single block. The f0 of a tone is represented by the color of the block. Each tone consists of three pulses which is represented by the boxes in the square. The gaps between each tone represents the 200 ms of silence between each tone. The blue blocks represent the one-back detection task where the participant would identify when the same stimulus was presented twice in succession. 

Each run consists of two blocks (20-s duration) of each condition, as well as five 12-s “rest” blocks. A rest block follows and tails each run with the remaining three randomly dispersed between the active task blocks. Between each active task there is four seconds of rest built in. These rest blocks are just periods of silence.

The stimulus presentation  for a single run where the blocks of silence are represented by a solid block  and the active blocks are striped. The gaps between the blocks represent the four seconds of silence between blocks.

The figure below shows what multiple runs over the course of one session could look like. All of the blocks are randomly dispersed throughout the run and never allowed to be in succession. Each run lasts about five and a half minutes, which lets us get a lot of data in a short session.

Multiple runs in succession. The rest periods of silence are represented by the black blocks. The harmonic resolved tones (HR), inharmonic resolved tones (IR), harmonic unresolved tones (HU), inharmonic unresolved tones (IU), low-frequency noise (LN), and high-frequency noise (HN) blocks are pseudorandomly presented.

Now we are preparing our protocol for the MRI. Soon we will be scanning our first pilot subject and calibrating the system. Onwards to more science!

Week 6: Week of 7/8

This week brought a big meeting on Tuesday that to my mind determined the rest of the project. Really, it just decided whether or not I would get to see my test run in an MRI system before I left. I really want those brains on my final poster if I am going to be honest. The meeting went great and my code got approved with just a minor tweak, which meant I could move on to testing subjects in a pilot test (that of course I had already scheduled for the next day!).

What the “pretty cover”, or Guided User Interface (GUI) of my test looks like.

The W in Wednesday stands for whirlwind right? Wednesday brought a 12 hour work day consisting of two seminars and running an entire pilot study. I started with a weekly seminar; this week covered the ins and outs of abstract writing and poster designing. It ended with a two-hour seminar where we dissected graduate school personal statements and talked about what strengthened a personal statement and what killed one.

The pilot study was it’s own pit of craziness. As soon as I returned from morning seminar I had my first subject and I definitely hadn’t reviewed the testing protocol well enough! Five subjects turned into four subjects when one subject had to be ruled out of the study for not meeting the tough qualifications our lab had set. The protocol consisted of paperwork, a hearing test, and seven runs that lasted five and a half minutes. That doesn’t sound like a lot till you realize you still have to get work done while running subjects.

Every second has to be down to the clockwork. I found the perfect use for my new Fitbit: letting me know when 5.5 mins had past. As soon as the timer went off I ran to the booth with my notebook to jot down some preliminary data and set the subject up for the next run. During the entire proceedings I was building a code to analyze the data and then analyzing the data in real time, while working on a piece of documentation for my team. When I was running the last subject, my subject knocked on the window of the booth to get my attention …. I had fallen asleep in the chair next to the booth. Gotta love the research life!

The kids today successfully put on the EEG cap and were surprisingly good at gelling the electrodes in.

Don’t worry work isn’t always all science. This week we’ve had two different group of campers visiting the lab to see the space and learn a bit about sensory neuroscience. The first demo we showed them was a haunted house sensory experience that someone had designed. This took place in one of the booths that we have set up with a sound localization system and three walls of screens. Then a bunch of short scary clips come up with Goldy the Gopher edited into them and they have to try to find Goldy. After this, we showed them a program that “gauges” their emotions based on their facial features.

The emotion display system that uses facial recognition to categorize emotions. A positive spike indicates that emotion is present.

Next my involvement came more into play, we explain a bit more about the work we do and the auditory system. Then we showed them the EEG system and they got to set up a “subject” (a.k.a. me). My head got all jelled up, with varied success, and they got to button all the electrodes to the swim-cap-like EEG cap. The last demo of the day is probably one of my favorites because it’s more of a biomechanics thing than our lab, but technically we do have a motion system in the sensory booth. A few of the kids get to stick dot-sensors to themselves and go dance around in the sensory booth while we see their “form” (it’s a collection of dots) appear on the screen.

Two people dancing in the motion cap system. If you look at the dots, you can see a human outline.

A Scary Thing

Here’s something really scary to do: randomly email people and ask things of them. Even more scary? Randomly find people on LinkedIn and ask if they will meet with you, a total stranger. Now why am I stalking people you ask? One of my professors at school recommend that we meet with people in industry over the summer and just ask them about their lives (after all people love talking about themselves 🙂 ).

Was the project a success you ask? After messaging multiple people on LinkedIn who were working as an engineer or researcher in the biomedical industry here in Minneapolis and were conveniently Notre Dame alumni, I had three people get back to me to either meet for coffee or talk on the phone. I would call that a success!

Another month in my jump into academia: June 2019

Week 2: Week of June 10

Well this first week in the lab went as one could expect. I got introduced to the rhythms of lab and went through the patterns of short term memory loss expected with learning a lot of information in a very short time. I learned what my project was going to be and started designing the stimuli that would be used for the test. Expect a post soon where I talk more about what I am doing and why. Short story short I am working to see which regions of the brain are activated by pitch.

A frequency spectra showing what each of the stimuli look like. The top row are low-frequency (like a snare) sounds and the bottom row are high-frequency sounds (like a flute).

Week 3: Week of June 17

This week was exciting because I started building my behavioral test for the experiment! If only I had remembered that it would take longer than my simple brain thought. 😉 This week also brought a lot of paper reading that I had put off the first week. I think I read 4 or 5 papers over the course of 2 days plus parts of a few different textbooks. My brain was definitely on information overload.

With one of my postdocs gone, I also found a few other things to keep me occupied. I had to give a presentation to my seminar group explaining what I was doing and how I was planning on doing it. I had a week to prepare this presentation and thankfully I had started building it at 3 in the morning on the day it was assigned, so I actually had it done during Week 2 to show to my PI and postdocs for critique. My postdocs said it was really good and I also got really good feedback from my seminar coordinate on it. Glad that design class I took finally came in handy!

I also had the privilege to take part in a friend’s vision study. They were testing to see if this crazy version of a visual phenomena worked the same as the original version. I got told that my data looked really good, which made the motion sickness worth it! Definitely a good experience before running my own subjects.

Week 4: Week of June 24

More paper reading and coding…. the days of academia as a research assistant. Four papers in one day, this must be a new record. I am finally at that magical point where a paper is referencing another paper and I can say “Ha! I’ve already read that one and understand what is happening!”; it’s a pretty great feeling. After meeting with my team, I moved on to version 2 of the code. This resulted in me finding out I am a very fast coder… and the only RA they have had that can really code. So I spent Thursday playing around with pictures of my brain and looking how to make it look cool in a bunch of different ways. Man I love when people make good documentation! Other exciting thing that happened: I got the go ahead to run subjects the next week!

Here’s what my brain looks like in case anyone was wondering.

Week 5: Week of July 1

Yay for holiday weeks! Can I manage to get all five of my subjects run before the 4th? Monday consisted of setting up the test and making sure everything will work correctly and all the information is correctly saved. Everything runs smoothly by the end of the day! Tuesday: time to meet with the team. I’m very excited to tell them I have all my subjects lined up for the next day ( 5 subjects in 8 hours!!) and see if we’re gonna start calibrating and if my dreams go right, run subjects at the scanner! The meeting results in the subjects being delayed a week and version 3 of the code being built with a new task for the subject. Instead of the original oddball detection the test now uses a one-back methodology. It’s harder to hear and we don’t have to throw out any data. Back to the computer and back to the code. I spent 9 hours on MATLAB that day… but the test was all done and ready to go by the end of the day. Wednesday resulted in me making the code fancier to do some different calculations and then writing up a 6-page document on how the code works and what exactly the test does. Maybe it’ll come into use at some point (like a paper). Then it was time for a great holiday weekend with family! (More to come on all my social happenings at some other point)