I'm from Algeria. I now live in Tianjin, China, doing my PhD in Information and Communication Engineering at Tianjin University. If you told me at the start of my PhD that's where I'd end up, I probably wouldn't have believed you.
This is the story of how I got here. Not the polished version. The real one.
Why Bioengineering
I chose bioengineering because I couldn't pick between biology and engineering, so I picked both. Honestly, that's it. I was fascinated by the idea that the human body behaves like a system. That you can measure it, model it, and sometimes predict what it's going to do. Every signal the body produces carries information. I found that endlessly interesting.
I was especially drawn to neurological signals. EEG, ECG, the way electrical activity encodes what's happening inside us. There's something almost strange about it. A waveform on a screen that represents someone's thoughts, or their heartbeat, or the tremor in their hand.
The Moment That Actually Decided Everything
It was early 2023. I had arrived in Tianjin just a few months before, in December 2022, and I was still in the phase of figuring out what my research topic would be. I had a bad migraine one day and went to the hospital.
While I was sitting in the waiting room, I noticed an old man across from me. He was trying to button his shirt. His hands were shaking, and he was trying quietly, on his own, not asking anyone for help. It took him a long time.
I couldn't stop watching. Something about that moment just got to me.
I went home and started reading about Parkinson's Disease. I learned that by the time the tremors show up, a lot of the neurological damage has already happened. There's no simple early test. Diagnosis relies on visible motor symptoms that appear late. I kept thinking: the body must be giving off earlier signals. In the voice. In the way someone walks. In their handwriting. We just haven't been listening carefully enough.
That became my research topic. No spreadsheet, no framework. Just that image of a man trying to button his shirt.
Learning to Code from Scratch
I want to be honest here because I think it matters. I had no real coding background when I started. I had done some basic programming during my engineering courses but nothing close to machine learning. The first few months were rough. I broke things constantly and spent more time reading error messages than actually writing useful code.
What saved me was working with real data immediately. The moment I loaded my first voice dataset and plotted a waveform, everything clicked. I had a real problem I cared about and suddenly the code had a purpose. Abstract tutorials alone never did it for me. The problem had to come first.
Where Things Stand Now
What I've Actually Learned
The technical stuff matters, obviously. But the thing nobody tells you is how much the clinical understanding matters just as much. If you don't know why a symptom appears, you can't build a meaningful feature. If you don't understand the disease, you can't really interpret what your model is doing. My bioengineering background wasn't a detour from AI. It became the whole foundation.
I also learned to be careful about accuracy numbers. Medical AI is full of optimistic results that don't survive contact with the real world. Small datasets, noisy conditions, the gap between a research paper and something a doctor can actually use in a clinic. I try to stay honest about that in my own work.
Why I'm Writing This
Because I couldn't find many stories like mine when I was starting out. Someone from Algeria, figuring out machine learning in China, starting from a moment in a hospital rather than from a computer science degree. If you're early in your path and things feel uncertain, that's okay. Most paths look clearer from the end than they do from the middle.