In the book Concrete Mathematics 2ed by Knuth Graham and Patashnik there is an exercise 5.45. The exercise is the following
$$ \text { 45 Find a closed form for } \sum_{k \leqslant n}\left(\begin{array}{l}{2 k} \\ {k}\end{array}\right) 4^{-k} $$
From what they hint in the chapter this is a known identity that has a closed form of $\left(\begin{array}{c}{n+1 / 2} \\ {n}\end{array}\right)$. However they mention that one can use the parallel summation identity from the book to reach this answer. The parallel summation identity from the book is the following, $$ \sum_{k \leqslant n}\left(\begin{array}{c}{r+k} \\ {k}\end{array}\right)=\left(\begin{array}{c}{r+n+1} \\ {n}\end{array}\right), \quad \text { integer } n . \quad \text { parallel summation } $$
However this implies that the following should be true $$ \sum_{k \leqslant n}\binom{k-1/2}{k} =\sum_{k \leqslant n}\left(\begin{array}{l}{2 k} \\ {k}\end{array}\right) 4^{-k} $$
And I cannot see how this is true?
That formula is indeed correct. We can calculate that $\binom{k-1/2}{k}=\binom{2k}{k}4^{-k}$.
Indeed, by manipulating our expression, you can calculate \begin{align*} \binom{k-1/2}{k}&=\frac{\left(k-\frac{1}{2}\right)\left(k-1-\frac{1}{2}\right)\dots\left(\frac{1}{2}\right)}{k!}\\ &=\frac{\left(\frac{2k-1}{2}\right)\left(\frac{2k-3}{2}\right)\dots\left(\frac{1}{2}\right)}{k!}\\ &=\frac{(2k-1)(2k-3)\dots(1)2^{-k}}{k!}\cdot \frac{(2k)(2k-2)\dots2}{(2k)(2k-2)\dots2}\\ &=\frac{(2k)!}{k!}\cdot \frac{1}{(2k)(2k-2)\dots2}2^{-k}\\ &=\frac{(2k)!}{k!}\cdot \frac{1}{k(k-1)\dots1 \cdot 2^k}2^{-k}\\ &=\frac{(2k)!}{k!k!}2^{-2k}\\ &=\binom{2k}{k}4^{-k} \end{align*}