6 Weeks Until I Graduate Stanford's Biomedical Informatics Graduate Certificate

It feels like forever since I started my first class, BIOMEDIN260 - Data Science for Medicine, less than a year ago in September 2022. They say time flies when you're having fun, but that's also true when you're extremely busy. As I near the completion of Stanford University School of Medicine's Biomedical Informatics (BMI) Graduate Certificate in 6 weeks and begin my journey through Stanford University School of Engineering's Computer Science (CS) courses, I wanted to take some time to reflect on my experience thus far.

Perpetual oscillation

I've come to realize that I personally enjoy challenges. As I think on this, I realize that this didn't really start during pharmacy school either as I've always been extremely competitive when I used to play video games. My choice was always the solo/1v1 matches in real-time strategy games like Blizzard Entertainment's Warcraft III or Starcraft 2 as the sole predictor of success was myself and I'd remove the variable of loss due to a teammate. All that to say, challenges, in and of itself, typically provide a large dose of dopamine whenever something challenging presents itself.

Conversely, the amount of dopamine released in my body likely resembles more of a bolus rather than a continuous stream of reward as I'm usually hit with stress and frustration shortly thereafter. Aside from the relationship between these two neurotransmitters, the last few months have certainly been a back and forth between excitement and stress as I work through learning new things and struggle to solve challenging problems in the form of math, statistics, and programming.

I imagine this feeling can likely be extrapolated to learning and/or doing anything new as and, if I were to guess, likely somewhat of a U-shaped curve with dopamine on the y-axis and time on the x-axis. My confidence in my abilities has grown steadily over the last few months and the positive feelings have certainly surpassed the negative ones. For my clinical colleagues who are contemplating pursuing a similar path, let's just say it's a rollercoaster of emotions, but I think it's worth it.

A new lexicon

In lieu of this article, I actually wanted to write about knowledge engineering. However, figured an introspective article about my experience thus far would allow me to express some creativity and also reduce my cognitive load that'd likely come from deep diving into knowledge engineering. Ironically, this paragraph can also serve as a primer for that future article :).

Sometime between 2017-2019, I felt very compelled to do a PhD. I don't remember if I wrote about it at the time, but I created a YouTube video about it here: https://youtu.be/hvtxMSSrq-4. Taking a step back, my impetus to do so was because I was fascinated with the idea of interoperability and how to tackle the challenge of representing the complexity of healthcare in a way that would allow computers to consume it. Unfortunately, I never really knew what I would do my PhD in to solve a problem like that....until I started taking these courses.

To be fair, I believe my first encounter with the word, "knowledge engineering" was in the Health Sciences Informatics, Knowledge Engineering and Decision Support course that I took during my MPH. I actually used ~20 of my credit hours (out of 80) exploring biomedical informatics courses at Johns Hopkins Bloomberg School of Public Health and really began to familiarize with the words used to describe these interoperability challenges in healthcare. The deep dives into the technical world in the last few months have certainly helped cement my areas of interest and build my lexicon in describing what I really want to spend my life doing. For those who are interested in tackling the interoperability problem, especially clinicians, I think side quests into biomedical informatics or computer science can prove quite valuable in helping you articulate what you really want to work on at the intersection of healthcare and technology.

Math is scary

I fit the Asian stereotype quite well as I used to be good at math. The only reason I scored an 86 percentile on my PCAT or did average on the MCAT (got a 25) was solely due to my math scores bringing the abysmal biology, reading and writing scores up. I can't recall my math score on the MCAT, but my math on the PCAT was around a 99th percentile. I loved math growing up as it felt objective and natural to me. Though, having been more on the healthcare side of things for over a decade, my confidence and knowledge of difficult math concepts in calculus and linear algebra have dwindled to the point where I'm actually scared when quantitative skills are mentioned as a prerequisite for these courses. Typically, it's a mixture of math, statistics, biology, and programming.

I suspect I naturally desire instant gratification as a millennial as I opted to skip a lot of the recommended courses, especially the quantitative ones, to see how difficult some of these courses really were. Well, I've come to learn they're really difficult. To provide a more objective qualifier, let me put it this way. Stanford University has a course exploration tool called Carta. It's kind of like a rate my professors kind of tool that helps you get a sense of what you can expect from the course. One of the most helpful features though, is the intensity charts that display an aggregation of student responses about the number of hours they spend per week. For me, I almost always spend 2 to 3 times as long on these courses because of how challenging the problem sets and projects are to me. Especially when I compare my experiences with my classmates, many of whom come from computer science backgrounds, it's overtly obvious that I'm an outlier.

My inability to completely comprehend, and in some cases not even complete, some of the problem sets in previous classes made me realize this was a huge gap. That gap is became more clear in my current class, BIOMEDIN260 - Computational Methods for Biomedical Image Analysis and Interpretation, that goes into painstaking detail the mathematical approaches that form the basis of many deep learning approaches like convolutional neural networks (#CNNs). As I near the completion of my BMI courses and transition into foundational CS courses, I'm somewhat relieved to know that I'll reinforce my understanding of the basics, but I'm also now traumatized by math courses in general lol.

Journey to CS

As I wrap up the last 6 weeks of my BMI and embark on my new CS journey, I'm somewhat worried about my motivation will decrease given the more abstract nature of the courses and an absent focus on medical applications. Further, I'm starting my first course off with CS103 - Mathematical Foundations of Computer that comes with disclaimers like:

There's going to be some point in this course where you find yourself struggling. Chances are it'll be a problem set question that you can't make heads or tails of, but it could also be a proof in lecture that just doesn't make sense or a definition that just seems plain wrong.

I think I'll be in for a treat. On that note, I'll also be taking a short break after that class for a few months as I'm already quite burnt out from balancing work and school. I've also been contemplating whether or not I should continue given the arduous path ahead. As I review my planned schedule of courses I'll have to take, if I get admitted to the MSCS program, I'm worried about the sacrifices I'll have to make over the next few years. If you're curious, here's a link to my tentative schedule over the next 4 years. The main reasons I haven't completely scrapped the idea is because 1) the recent breakthroughs in #AI technology with OpenAI's large language model (#LLM), #chatGPT and 2) the opportunity to learn how I can apply the CS knowledge to solve my lifelong goal of connecting the world's healthcare data.

Concluding thoughts

Although this article wasn't as course-specific as I would have liked, hopefully it provides a bit of insight to others that are considering venturing down this path. I'll be sure to share my thoughts as I progress through the CS program and we'll see if the first course I take next quarter will be my last.