My favorite figures (to date)

In my previous post, I discussed my pipeline for making publication-ready figures. Below are some of my favorite figures that I’ve published over the years using this exact pipeline.

Event-related potentials of verbal analogical reasoning

Reprinted from Brain and Cognition, 129, Kmiecik, M. J., Brisson, R. J., & Morrison, R. G., The time course of semantic and relational processing during verbal analogical reasoning, 25-34, Copyright (2019), with permission from Elsevier

The above figure¹ was one of the first figures I published after reading Edward Tufte. Notice how the figure is colorless? This project was no longer funded and I didn’t have the means to pay the journal to publish color figures; therefore, I created the figure in black/white. There are two conditions each with two levels. Because the conditions were separated by the electrode topographies in the middle column, I used titles and two linetypes to keep these conditions distinct, but allowed them to share a linetypes for a clean look.

Color can often make figures pop and look nice, but a classic grayscale figure can get the job done.

Learning rates of relationally complex problems

I love figures that maximize the amount of information in a small space. This multi-paneled figure above² uses a colorblind friendly color palette, shows individual data points, and attempts to reduce chart junk—unnecessary plot elements that don’t add information—most evident in the correlation plot on the bottom right. The line plots with the dots in the background feature a 2x2 experimental design with the first factor being represented by linetype and the other by color. However, I also wanted to show the individual data points. Because circles cannot have a linetype (well, at least not easily deciphered in this case) I decided to use opacity instead. 

Bladder Pain and Visual Evoked Potentials

Again, lots of information in a small space (see the figure above³). Note the color palettes used for the electrode topographies. They are two-color diverging color palettes that have a central/zero white color—a perfect palette to use when the estimate you are plotting has a natural zero point (in this case, slopes). One of my biggest pet peeve in the neuroimaging literature is that researchers are still plotting estimates on structural MRIs using rainbow colors that use green as the zero/center point. Our eyes have a difficult time deciphering the rainbow color palette and cannot fully appreciate the differences that are intended to be shown. Instead, opt for color palettes seen in nature, like the sunset.

Multimodal hypersensitivity predicts future pain

At the time I was creating this figure above⁴, I had recently discovered the amazing Studio Ghibli and was watching Spirited Away, Howl’s Moving Castle, Princess Mononoke, and Kiki’s Delivery Service. I came to discover the R package ghibli that contains the color palettes from several of the films. I was obsessed with these color palettes and knew I had to use them in a publication. Also, I found it visually appealing to use transparent circles with no borders when plotting the distribution of individual data points (see next to the box plots). 

LRRK2 G2019S Parkinson’s Disease

My latest work combines both spatial information and quantitative data to create a powerful message of neurodegeneration in Parkinson’s Disease (see figure above⁵). When examining the brain, the eye is drawn to the dark red area of the midbrain called the substantia nigra, emphasizing the devastating impact of Parkinson’s disease on this region. The lighter pink area just below in the pons and brainstem is thought responsible for REM sleep behavior disorder (RBD). RBD is thought to be a prodromal symptom of Parkinson’s with afflicted individuals later developing Parkinson’s at greater rates than those that do not exhibit symptoms of RBD. The lighter color signals that RBD is less prevalent in carriers of the LRRK2 G2019S genetic variant. More detailed information on the prevalence of symptoms surrounds this brain image, allowing readers a comprehensive picture of how symptoms relate to neurodegeneration in Parkinson’s. 

1. Kmiecik, M. J., Brisson, R. J., & Morrison, R. G. (2019). The time course of semantic and relational processing during verbal analogical reasoning. Brain and Cognition, 129, 25-34. https://doi.org/10.1016/j.bandc.2018.11.012

2. Kmiecik, M. J., Perez, R., Krawczyk, D.C. (2021). Navigating Increasing Levels of Relational Complexity: Perceptual, Analogical, and System Mappings. Journal of Cognitive Neuroscience, 33(3), 357-376. https://doi.org/10.1162/jocna01618

3. Kmiecik, M. J., Tu, F. F., Silton, R. L., Dillane, K. E., Roth, G. E., Harte, S. E., Hellman, K. M. (2022). Cortical Mechanisms of Visual Hypersensitivity in Women at Risk for Chronic Pelvic Pain. PAIN, 163(6), 1035-1048. doi: 10.1097/j.pain.0000000000002469

4. Kmiecik, M. J., Tu, F. F., Clauw, D. J., & Hellman, K. M. (2023). Multimodal Hypersensitivity Derived from Quantitative Sensory Testing Predicts Pelvic Pain Outcome: an Observational Cohort Study. PAIN, 164(9), 270-283. http://dx.doi.org/10.1097/j.pain.0000000000002909

5. Kmiecik, M. J., Micheletti, S., Coker, D., Heilbron, K., Shi, J., Stagaman, K., Filshtein Sonmez, T., Fontanillas, P., Shringarpure, S., Wetzel, M., Rowbotham, H. M., Cannon, P., Shelton, J., Hinds, D. A., Tung, J. Y., 23andMe Research Team, Holmes, M. V., Aslibekyan, S., Norcliffe-Kaufmann, L. (2024). Genetic analysis and natural history of Parkinson’s disease due to the LRRK2 G2019S variant. Brain, 147(6), 1996–2008. https://doi.org/10.1093/brain/awae073