Law of Ito Integral

Question

Let $\sigma$ be an $\mathcal{F}_t$-adapted caglad stochastic process. Let $W$ be a Brownian motion independent of $\sigma$. Let $r>0$ be a strictly positive real number. How can I prove that $$ \left|\int_{0}^1\sigma_s\,dW_s\right|^r \stackrel{d}{=} \left|U\right|^r\,\left(\int_0^1\sigma_s^2\,ds\right)^{r/2}, $$ where $U\stackrel{d}{=}$ is a $\text{N}(0,1)$ gaussian variable?