Wenqi Li

Currently `shuffle` and `epoch` are tightly coupled in the `ResizeSampler` -- the number of epochs is infinite if `shuffle` is on, and is 1 if `shuffle` is off. These two...

It seems that in many window-based segmentation/regression use cases, either setting`is_training=True` or setting `use_local_stats=True` in the [batch normalisation layer](https://github.com/NifTK/NiftyNet/blob/v0.4.0/niftynet/layer/bn.py#L79) at inference time leads to better performance than those with NiftyNet's...

**Is your feature request related to a problem? Please describe.** looking into optional spatial resampling backends using cupy-based implementations, mainly for the cuda-based high order interpolation mode. the existing solution...

fusing spatial transforms

2

# Lazy resampling (fusing spatial transforms) **Is your feature request related to a problem? Please describe.** Follow up of https://github.com/Project-MONAI/MONAI/discussions/4198, https://github.com/Project-MONAI/MONAI/issues/112, when there are various spatial transforms such as `Spacing`...

`monai.visualize.matshow3d` for metatensor

1

**Is your feature request related to a problem? Please describe.** when it's input is a metatensor, `monai.visualize.matshow3d` should have an option to display the image content with the actual aspect...

### Discussed in https://github.com/Project-MONAI/MONAI/discussions/4807 Originally posted by **ChenglongWang** August 1, 2022 Recently, I noticed that the `pixdim` argument from the `Spacing` transform is inconsistent with other transforms. For the `pixdim`...

**Is your feature request related to a problem? Please describe.** for backward compatibility, it might be useful to consider an option of `as_dict=True` to enable `FromMetaTensor` as part of the...

profiling the image reader/writer

1

**Is your feature request related to a problem? Please describe.** would be great to benchmark (and improve) the reader/writer performance, or recommend optimal options for those components. cc @AHarouni

remove `torch.testing.assert_allclose` calls

1

**Is your feature request related to a problem? Please describe.** ``` `torch.testing.assert_allclose()` is deprecated since 1.12 and will be removed in 1.14. Use torch.testing.assert_close() instead. For detailed upgrade instructions see...

**Is your feature request related to a problem? Please describe.** the previous implementations of `get_*_layer` https://github.com/Project-MONAI/MONAI/blob/ff9ff55dd9198dabcc31eebe3109332ea708b356/monai/networks/blocks/dynunet_block.py#L133-L137 could now be replaced by `Convolution ` https://github.com/Project-MONAI/MONAI/blob/ff9ff55dd9198dabcc31eebe3109332ea708b356/monai/networks/blocks/convolutions.py#L24 this is however a breaking change,...