Get learning rates and their corresponding losses Args: skip_start: number of batches to trim from the start. skip_en

Return correct shortcut_type and bias_downsample for pretrained MedicalNet weights according to resnet depth.

Returns the total number of parameters in a Module.

Get number of classes.

Return the number of frames in a video file. Raises: RuntimeError: no frames found.

Get the number of times that the data need to be converted (e.g., numpy to torch). Conversions between different devices are also counted (e.

Get a numpy dtype (e.g., `np.float32`) from its string (e.g., `"float32"`).

Given a sample input shape, calculate min/opt/max shapes according to dynamic_batchsize.

Get estimated output tensor size to approximate RAM consumption. Args: in_size: input image shape (4D/5D, ``[BCHW[D]]``)

Returns a tuple of slices to define a random patch in an array of shape `dims` with size `patch_size` or the as close to it as possible withi

This is equivalent to self.layer_blocks[idx](x), but torchscript doesn't support this yet

Returns the size (height, width) of the whole slide image at a given level. Args: wsi: a whole slide image object loaded

Get either a sequence or single instance of np.ndarray/torch.Tensor. Return single torch.Tensor.

Get learning rate which has steepest gradient and its corresponding loss Args: skip_start: number of batches to trim from the sta

Calculate the target spacing according to all spacings. If the target spacing is very anisotropic, decrease the spacing value

Achieve the referenced UID from the referenced Series Sequence for the input pydicom dataset object. The referenced UID could be Series Insta

Returns a dictionary mapping results entries to tuples containing the number of items, time sum, time average, time std dev, time min

Get a torch dtype (e.g., `torch.float32`) from its string (e.g., `"float32"`).

Returns: tuple of ints represents the pytorch major/minor version.

Get the backends of all MONAI transforms. Returns: Dictionary, where each key is a transform, and its corresponding values are a

Get version information or return default if unable to do so.

Gets a zero tensor. Args: input: tensor which shape you want the zeros tensor to correspond to. Returns:

Checks whether a given tensor is has a particular status key message on any of its applied operations. If it doesn't, it returns the tuple `(

Runs TRT engine. Args: stream: CUDA stream to run on use_cuda_graph: use CUDA graph. Note: requires all input

Initialize the image writer modules according to the filename extension.

Instantiate the target component and return the instance.

Instantiate component based on ``self.config`` content. The target component must be a `class` or a `function`, otherwise, return `No

Inverse of the crop transform, restoring the original spatial dimensions via padding. For the string-key path, the cropper used in `

De-pad network output to match its original shape Args: x: input of shape (B,C,H,W) for 2D data or (B,C,H,W,D) for 3D data p

Check if the data is a DDF.

Calculate if the input shape is divisible by the minimum factors for the current network configuration

Warping with python itk Args: image: itk image of array shape 2D: (H, W) or 3D: (D, H, W) ddf: numpy array of

Keep connected regions of img_pos and img_neg that include the positive points and negative points separately. The function is used for mergi

Extract kwargs from a pending transform item. When ``pending_item`` is a dict, ``align_corners`` is also extracted from its ``extra_info`` entry

Restore state from a dictionary. Args: state: A dictionary containing the state to restore.

Create random affine transformation (with values == -1, 0 or 1).

Match regular expression for the input string to find the references. The reference string starts with ``"@"``, like: ``"@XXX::YYY::Z

Based on https://github.com/Tencent/MedicalNet/blob/18c8bb6cd564eb1b964bffef1f4c2283f1ae6e7b/datasets/brains18.py#L133

Update the `args` dict-like object with the key/value pair `k` and `v`.

Converts a MetaTensor object to an ITK image. Expects the MetaTensor to be in ChannelFirst format. Args: meta_tensor: The MetaTensor

Converts a MONAI affine matrix to an ITK affine matrix (2x2 for 2D or 3x3 for 3D matrix and translation vector). See also 'itk_to_monai_affin

Warping with MONAI Args: image_tensor: torch tensor of shape 2D: (1, 1, H, W) and 3D: (1, 1, D, H, W) ddf_ten

Wrapper handles inference for 2D models over 3D volume inputs.

Update deprecated identifiers in `config` using `DEPRECATED_ID_MAPPING`. This will replace names that are marked as deprecated with t

Use OpenCV to open a video source from either file or capture device. Args: video_source: filename or index referring to

Parse the version strings.

Parzen windowing with b-spline kernel (adapted from ITK) Args: img: the shape should be B[NDHW]. order: int.

Parzen windowing with gaussian kernel (adapted from DeepReg implementation) Note: the input is expected to range between 0 and 1

Apply pixel shuffle to the tensor `x` with spatial dimensions `spatial_dims` and scaling factor `scale_factor`. See: Shi et al., 2016, "Real

Run prediction, ensemble, and/or postprocessing optionally. Args: folds: which folds to use run_ensemble: wh

Prepares onehot encoded input for metric call.

Print the package versions to `file`. Args: file: `print()` text stream file. Defaults to `sys.stdout`.

Print config (installed dependencies, etc.) and system info for debugging. Args: file: `print()` text stream file. Defaults to `sys.

Prints a list of backends of all MONAI transforms.

Compute the lab from the probability of predicted feature maps Args: sigmoid: If the sigmoid function should be used. thresh

Given an input it projects it to the quantized space and returns additional tensors needed for EMA loss. Args: inputs: E

If multiple values are provided for center_fractions and accelerations, this function selects one value uniformly for each tr

Return the item at a randomized location `self._idx` in `buffer`.

Read image data from specified file or files, it can read a list of images and stack them together as multi-channel data in `get_data

This function is used to rename all instance id of `pred`, so that the id is contiguous. For example: all ids of the input can be [0, 1,

Remove the resolved ``ConfigItem`` by id. Args: id: id name of the expected item.

`np.repeat` with equivalent implementation for torch (`repeat_interleave`). Args: a: input data to repeat. repeats: number o

Replace sub-module(s) in a parent module. The name of the module to be replace can be nested e.g., `features.denseblock1.denselayer1.lay

Convert the ``data_array`` into the coordinate system specified by ``target_affine``, from the current coordinate definition of ``aff

Reset the pandas `TextFileReader` iterable object to read data. For more details, please check: https://pandas.pydata.org/pandas-docs

Restores the model and optimizer to their initial states.

Detaches batch and channel dimensions. Args: x: input of shape (B*C,1,H,W,2) for 2D data or (B*C,1,H,W,D,2) for 3D data batc

Swaps the complex dimension with the channel dimension so that the network output has 2 as its last dimension Args: x: input of shap

Combines batch and channel dimensions. Args: x: input of shape (B,C,H,W,2) for 2D data or (B,C,H,W,D,2) for 3D data Returns: