Response to “A threshold without a map: Why isodose normalization must accompany D95/D2 TCP cutpoints” | Synapse