I have studied in Hatcher (Algebraic topology, p.169) an homological proof of the Jordan theorem. I would like to understand an upgrade of this theorem, the Schoenflies version.  
On my way to prove it, I have studied the idea of using the Caratheodory theorem. Even if I can understand this one, I am completely unable to conclude. 

The common idea (that is shared on internet), is to use the conformal mapping property, and I agree with that. But here is the problem, as everything related to the Jordan theorem, the fact that a Jordan domain is simply connected seems clear, but the demonstration is not.

So, I tried to prove this assertion. Here is my first idea:  
In the Hatcher's proof of Jordan theorem, we saw that 
$$H_i(\mathbb{S}^2 - h(\mathbb{S}^1))=0$$ 
Except for $i =2-1-1=0$ (here $h$ is any continuous function). So $H_1(G)=0$ where $G$ is the bounded component of $\mathbb{R}^2-\mathbb{S}^1$. However, this is not a characterization of simple connectedness, we would like to have the save result but with $\pi_1(G)$. So there is my question, does someone know (a link to) a proof that 
$$ \pi_1(G)=0 ?$$
It can be a rather conceptual explanation, I would like it.

Furthermore, I have heard about a definition of simple connectedness with Jordan curves, but I don't know how to prove it either.

Thank you