1. It will not be. Neuro imaging, computational neuroscience, neural computation HEAVILY recruits from Comp sci, mathematics, and engineering backgrounds. You will just have a lot to learn on your own, or if you decide to do a graduate/phd, you will have a lot of classes that will be new to you.

2. Not too late. It's a bad way of thinking anyway. If someone were to tell me "it's too late", I'd still try it anyway. Why listen to someone on the offchance they are right? Neuroscience is the study of the nervous system, so a good place to start would be a textbook that covers all aspects. My school used a book called Neuroscience: Exploring the Brain by Mark Bear. It seems to be a standard book because I've seen it recommended here and there. I dont think it's anything amazing but it covers pretty much all the topics you will need

3. Well there is cellular, molecular, pharmacological research, and then there is more computational oriented research that's done in computer 'labs' which are essentially offices with computers. At least that's the environment I work in. I'm only an assistant but a lot of the data analysis I do is on data taken from the hospital (UCLA). I love it so far because it's always exciting talking about new projects and what not. And then if you have your phd you can join your peers and fly around to attend conferences and what not

4. Countries...I'm no expert on this but I would look at SCHOOLS instead of countries. Off the top of my head, I THINK UCLA, UCSD, Caltech, MIT, all do a good amount of neuro research. NTNU in Norway I've heard has a strong program. It's one of the schools I'm looking into for masters. I'm sure most asian countries have good research programs too because I noticed a lot of our collaborations have been with people in China. Like I said though, I would look into schools. Country matters more if you're talking about tuition, funding, language barrier, living environment

5. I want to say biometrics and artificial NN are well within the realm of neuroscience, but I'm not familiar with either so someone else should give their opinion

I'm a neuroscience PhD student.

There are plenty of neuroscientists who had no exposure to biology until graduate school. You'll be fine. In fact, around a quarter of neuroscientists have PhDs in physics.

It's going to take some work. There's a lot of biology that's essential to doing good neuroscience, especially on the level of the synapse. Too many theoretical neuroscientists nowadays have no understanding of the biology, and, as such, their models are off. It's certainly not insurmountable, but you will have to dedicate yourself to understanding molecular biology, biochemistry, etc. As someone mentioned, "Principles of Neural Science" is fantastic.

Most neuroscientists are in a lab of some sort, but it's generally pretty social.

Neural networks rely on a lot of engineering and math concepts so you'll find that that background will become incredibly useful.

As for countries/schools, the US has the strongest neuroscience currently because of funding, but there is fantastic neuroscience all around the world. In Japan, the Riken institute has some really wonderful people. In Sweden, the Karolinska Institute does some fantastic work. Throughout Germany, the Max Plancks are great. In England, UCL, Oxford, and Cambridge all have great scientists. In the US, the hands down best places for computational neuroscience are Columbia and UCSD. Columbia has Larry Abbot and Ken Miller, who are leaders in modeling. UCSD has Terry Sejnowski, who essentially founded computational neuroscience. Caltech and MIT are also strong. For more hands-on, systems neuroscience (what I do), Harvard and UCSF have the strongest programs, though Columbia, Johns Hopkins, UCSD, and Rockefeller are also strong.

Hope this helps!

I'm currently a 2nd-year computational neuroscience PhD student. My undergraduate background was in electrical and computer engineering, and it's been a fine preparation for dealing with the issues I face at both a research and coursework level. Just make sure you apply to a neuroscience program that has a strong focus on computational techniques and/or theoretical modeling.

Most comp neuro graduate programs will assume you know nothing about neuroscience coming in besides a few things you may have picked up from a high school biology course. As the other commenter said, don't think of things in terms of it being too late. I'll also recommend a textbook: Kandel, Schwartz and Jessell's "Principles of Neural Science" is more or less the best starting point for graduate-level neuroscience study.

My current grad program is a great environment. We spend most of our time in front of a computer writing code or drawing systems on white boards. With many (but not all) programs in computational neuroscience you don't need to get your hands dirty with real brains if you don't want to - although there's usually plenty of opportunity for it if you decide you do.

There are great neuroscience programs in every country. As far as American universities, I would also add Carnegie Mellon and Boston University for computational neuroscience to the list by the previous commenter.

Biometrics is a huge topic in comp. neuro these days, if you mean what I think you do by the word, that is, the development of diagnostic tools for biological systems. Also , one of the great things about neuroscience is you can find an application for most engineering topics in the field.

Hope this helps answer your questions. Good luck!!