monailabel.deepedit.interaction module

class monailabel.deepedit.interaction.Interaction(deepgrow_probability, transforms, train, click_probability_key='probability', max_interactions=1)[source]

Bases: object

Ignite process_function used to introduce interactions (simulation of clicks) for DeepEdit Training/Evaluation.

More details about this can be found at:

Diaz-Pinto et al., MONAI Label: A framework for AI-assisted Interactive Labeling of 3D Medical Images. (2022) https://arxiv.org/abs/2203.12362

Parameters

  • deepgrow_probability (float) – probability of simulating clicks in an iteration

  • transforms (Union[Sequence[Callable], Callable]) – execute additional transformation during every iteration (before train). Typically, several Tensor based transforms composed by Compose.

  • train (bool) – True for training mode or False for evaluation mode

  • click_probability_key (str) – key to click/interaction probability

  • max_interactions (int) – maximum number of interactions per iteration

previous

monailabel.deepedit.handlers module

next

monailabel.deepedit.transforms module