I disagree with you: your definition of killing seems to be: any action which can be known to have a high probability of resulting in the death of a human.
This differs from our definition, which is to take an action which directly targets the human for death.
To us, there is ac difference between an abortion, an act whose sole purpose and direction is to attack the unborn baby and cause its death, and the removal of a diseased organ which, unfortunately, contains an unborn child at the time.
In the first case, the result desired is the death of the baby. If the mother were not pregnant, she would not be having the abortion at all. In the second, the result desired is the restoration of the mother to health; she would need the operation whether or not she was pregnant. The death of the baby is an undesired result, a side-effect.
You seem to be arguing that because the baby dies in each case, both acts carry the same morality. And yet you know that this reasoning is false, because when applied elsewhere, it is clearly erroneous.
Imagine that Bob and Joe each need money for an operation to save their mothers’ lives. Bob gets an evening job as a security guard and a weekend job mowing lawns to obtain the money.
Joe starts robbing people at gunpoint to get the money.
Obviously Bob is not doing something wrong and Joe is doing something wrong, and yet the outcome in each instance is the same: they get the money for their mothers’ operations.