July 23, 2025
Day 42 - Wrapping Up Methodology Draft
What I Learned
Today marked a major milestone in my research internship — I officially completed the full draft of the Methodology section for our ECG-based deep learning paper. Over the past few days, I’ve been carefully refining how we present our preprocessing pipeline, including the wavelet denoising, bandpass filtering, and the signal reconstruction from 2D grayscale ECG images to usable 1D time-series vectors. I also finalized the documentation of our feature extraction techniques across time-domain and frequency-domain representations, and how these features power our 1D-CNN, LSTM, and 2D-CNN model architectures. It felt good to step back and see the structure come together — clear, reproducible, and aligned with the scientific rigor expected in a research paper. Now that the methodology is locked in, I can shift focus to refining the Results and Discussion sections once all model evaluations (including noisy data augmentation) are finalized. Every part of this project is pushing me to think like a real researcher, not just a coder. And today, I felt that growth.
Blockers
No blockers.
Reflection
Today I finally finished writing the methodology section of our research paper — and honestly, it feels like a quiet but important win. For the past few days, I’ve been deep in the weeds of this section: explaining how we transformed raw ECG images into meaningful 1D signals, how we applied wavelet denoising and bandpass filtering, and how we extracted both time-domain and frequency-domain features for our models. It wasn’t easy — translating complex processes into clear, scientific writing takes more patience than I expected. But it also made me realize how much I’ve grown. A few weeks ago, I was just trying to understand the pipeline. Today, I was able to explain it. It feels good — not in a flashy way, but in that steady, internal kind of way. Like I’ve built something that matters, even if no one sees it yet. There’s still more to do — evaluation, results, and final revisions — but this felt like an anchor point. The kind of moment you look back on later and realize: yeah, that’s when it started coming together.