Datasets:

ArXiv:
File size: 31,249 Bytes
b4d7ac8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
# Copyright (c) MONAI Consortium
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#     http://www.apache.org/licenses/LICENSE-2.0
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""
A collection of "functional" transforms for spatial operations.
"""

from __future__ import annotations

import math
import warnings
from enum import Enum

import numpy as np
import torch

import monai
from monai.config import USE_COMPILED
from monai.config.type_definitions import NdarrayOrTensor
from monai.data.meta_obj import get_track_meta
from monai.data.meta_tensor import MetaTensor
from monai.data.utils import AFFINE_TOL, compute_shape_offset, to_affine_nd
from monai.networks.layers import AffineTransform
from monai.transforms.croppad.array import ResizeWithPadOrCrop
from monai.transforms.intensity.array import GaussianSmooth
from monai.transforms.inverse import TraceableTransform
from monai.transforms.utils import create_rotate, create_translate, resolves_modes, scale_affine
from monai.transforms.utils_pytorch_numpy_unification import allclose
from monai.utils import (
    LazyAttr,
    TraceKeys,
    convert_to_dst_type,
    convert_to_numpy,
    convert_to_tensor,
    ensure_tuple,
    ensure_tuple_rep,
    fall_back_tuple,
    optional_import,
)

nib, has_nib = optional_import("nibabel")
cupy, _ = optional_import("cupy")
cupy_ndi, _ = optional_import("cupyx.scipy.ndimage")
np_ndi, _ = optional_import("scipy.ndimage")

__all__ = ["spatial_resample", "orientation", "flip", "resize", "rotate", "zoom", "rotate90", "affine_func"]


def _maybe_new_metatensor(img, dtype=None, device=None):
    """create a metatensor with fresh metadata if track_meta is True otherwise convert img into a torch tensor"""
    return convert_to_tensor(
        img.as_tensor() if isinstance(img, MetaTensor) else img,
        dtype=dtype,
        device=device,
        track_meta=get_track_meta(),
        wrap_sequence=True,
    )


def spatial_resample(
    img, dst_affine, spatial_size, mode, padding_mode, align_corners, dtype_pt, lazy, transform_info
) -> torch.Tensor:
    """
    Functional implementation of resampling the input image to the specified ``dst_affine`` matrix and ``spatial_size``.
    This function operates eagerly or lazily according to
    ``lazy`` (default ``False``).

    Args:
        img: data to be resampled, assuming `img` is channel-first.
        dst_affine: target affine matrix, if None, use the input affine matrix, effectively no resampling.
        spatial_size: output spatial size, if the component is ``-1``, use the corresponding input spatial size.
        mode: {``"bilinear"``, ``"nearest"``} or spline interpolation order 0-5 (integers).
            Interpolation mode to calculate output values.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
            When it's an integer, the numpy (cpu tensor)/cupy (cuda tensor) backends will be used
            and the value represents the order of the spline interpolation.
            See also: https://docs.scipy.org/doc/scipy/reference/generated/scipy.ndimage.map_coordinates.html
        padding_mode: {``"zeros"``, ``"border"``, ``"reflection"``}
            Padding mode for outside grid values.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
            When `mode` is an integer, using numpy/cupy backends, this argument accepts
            {'reflect', 'grid-mirror', 'constant', 'grid-constant', 'nearest', 'mirror', 'grid-wrap', 'wrap'}.
            See also: https://docs.scipy.org/doc/scipy/reference/generated/scipy.ndimage.map_coordinates.html
        align_corners: Geometrically, we consider the pixels of the input as squares rather than points.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
        dtype_pt: data `dtype` for resampling computation.
        lazy: a flag that indicates whether the operation should be performed lazily or not
        transform_info: a dictionary with the relevant information pertaining to an applied transform.
    """
    original_spatial_shape = img.peek_pending_shape() if isinstance(img, MetaTensor) else img.shape[1:]
    src_affine: torch.Tensor = img.peek_pending_affine() if isinstance(img, MetaTensor) else torch.eye(4)
    img = convert_to_tensor(data=img, track_meta=get_track_meta())
    # ensure spatial rank is <= 3
    spatial_rank = min(len(img.shape) - 1, src_affine.shape[0] - 1, 3)
    if (not isinstance(spatial_size, int) or spatial_size != -1) and spatial_size is not None:
        spatial_rank = min(len(ensure_tuple(spatial_size)), 3)  # infer spatial rank based on spatial_size
    src_affine = to_affine_nd(spatial_rank, src_affine).to(torch.float64)
    dst_affine = to_affine_nd(spatial_rank, dst_affine) if dst_affine is not None else src_affine
    dst_affine = convert_to_dst_type(dst_affine, src_affine)[0]
    if not isinstance(dst_affine, torch.Tensor):
        raise ValueError(f"dst_affine should be a torch.Tensor, got {type(dst_affine)}")

    in_spatial_size = torch.tensor(original_spatial_shape[:spatial_rank])
    if isinstance(spatial_size, int) and (spatial_size == -1):  # using the input spatial size
        spatial_size = in_spatial_size
    elif spatial_size is None and spatial_rank > 1:  # auto spatial size
        spatial_size, _ = compute_shape_offset(in_spatial_size, src_affine, dst_affine)  # type: ignore
    spatial_size = torch.tensor(
        fall_back_tuple(ensure_tuple(spatial_size)[:spatial_rank], in_spatial_size, lambda x: x >= 0)
    )
    extra_info = {
        "dtype": str(dtype_pt)[6:],  # remove "torch": torch.float32 -> float32
        "mode": mode.value if isinstance(mode, Enum) else mode,
        "padding_mode": padding_mode.value if isinstance(padding_mode, Enum) else padding_mode,
        "align_corners": align_corners if align_corners is not None else TraceKeys.NONE,
        "src_affine": src_affine,
    }
    try:
        _s = convert_to_numpy(src_affine)
        _d = convert_to_numpy(dst_affine)
        xform = np.eye(spatial_rank + 1) if spatial_rank < 2 else np.linalg.solve(_s, _d)
    except (np.linalg.LinAlgError, RuntimeError) as e:
        raise ValueError(f"src affine is not invertible {_s}, {_d}.") from e
    xform = convert_to_tensor(to_affine_nd(spatial_rank, xform)).to(device=img.device, dtype=torch.float64)
    affine_unchanged = (
        allclose(src_affine, dst_affine, atol=AFFINE_TOL) and allclose(spatial_size, in_spatial_size)
    ) or (allclose(xform, np.eye(len(xform)), atol=AFFINE_TOL) and allclose(spatial_size, in_spatial_size))
    meta_info = TraceableTransform.track_transform_meta(
        img,
        sp_size=spatial_size,
        affine=None if affine_unchanged and not lazy else xform,
        extra_info=extra_info,
        orig_size=original_spatial_shape,
        transform_info=transform_info,
        lazy=lazy,
    )
    if lazy:
        out = _maybe_new_metatensor(img)
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else meta_info  # type: ignore
    if affine_unchanged:
        # no significant change or lazy change, return original image
        out = _maybe_new_metatensor(img, dtype=torch.float32)
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out  # type: ignore
    # drop current meta first
    img = img.as_tensor() if isinstance(img, MetaTensor) else img
    im_size = list(img.shape)
    chns, in_sp_size, additional_dims = im_size[0], im_size[1 : spatial_rank + 1], im_size[spatial_rank + 1 :]

    if additional_dims:
        xform_shape = [-1] + in_sp_size
        img = img.reshape(xform_shape)
    img = img.to(dtype_pt)
    if isinstance(mode, int) or USE_COMPILED:
        dst_xform = create_translate(spatial_rank, [float(d - 1) / 2 for d in spatial_size])
        xform = xform @ convert_to_dst_type(dst_xform, xform)[0]
        affine_xform = monai.transforms.Affine(
            affine=xform,
            spatial_size=spatial_size,
            normalized=True,
            image_only=True,
            dtype=dtype_pt,
            align_corners=align_corners,
        )
        with affine_xform.trace_transform(False):
            img = affine_xform(img, mode=mode, padding_mode=padding_mode)
    else:
        _, _m, _p, _ = resolves_modes(mode, padding_mode)
        affine_xform = AffineTransform(  # type: ignore
            normalized=False, mode=_m, padding_mode=_p, align_corners=align_corners, reverse_indexing=True
        )
        img = affine_xform(img.unsqueeze(0), theta=xform.to(img), spatial_size=spatial_size).squeeze(0)  # type: ignore
    if additional_dims:
        full_shape = (chns, *spatial_size, *additional_dims)
        img = img.reshape(full_shape)
    out = _maybe_new_metatensor(img, dtype=torch.float32)
    return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out  # type: ignore


def orientation(img, original_affine, spatial_ornt, lazy, transform_info) -> torch.Tensor:
    """
    Functional implementation of changing the input image's orientation into the specified based on `spatial_ornt`.
    This function operates eagerly or lazily according to
    ``lazy`` (default ``False``).

    Args:
        img: data to be changed, assuming `img` is channel-first.
        original_affine: original affine of the input image.
        spatial_ornt: orientations of the spatial axes,
            see also https://nipy.org/nibabel/reference/nibabel.orientations.html
        lazy: a flag that indicates whether the operation should be performed lazily or not
        transform_info: a dictionary with the relevant information pertaining to an applied transform.
    """
    spatial_shape = img.peek_pending_shape() if isinstance(img, MetaTensor) else img.shape[1:]
    xform = nib.orientations.inv_ornt_aff(spatial_ornt, spatial_shape)
    img = convert_to_tensor(img, track_meta=get_track_meta())

    spatial_ornt[:, 0] += 1  # skip channel dim
    spatial_ornt = np.concatenate([np.array([[0, 1]]), spatial_ornt])
    axes = [ax for ax, flip in enumerate(spatial_ornt[:, 1]) if flip == -1]
    full_transpose = np.arange(len(spatial_shape) + 1)  # channel-first array
    full_transpose[: len(spatial_ornt)] = np.argsort(spatial_ornt[:, 0])
    extra_info = {"original_affine": original_affine}

    shape_np = convert_to_numpy(spatial_shape, wrap_sequence=True)
    shape_np = shape_np[[i - 1 for i in full_transpose if i > 0]]
    meta_info = TraceableTransform.track_transform_meta(
        img,
        sp_size=shape_np,
        affine=xform,
        extra_info=extra_info,
        orig_size=spatial_shape,
        transform_info=transform_info,
        lazy=lazy,
    )
    out = _maybe_new_metatensor(img)
    if lazy:
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else meta_info  # type: ignore
    if axes:
        out = torch.flip(out, dims=axes)
    if not np.all(full_transpose == np.arange(len(out.shape))):
        out = out.permute(full_transpose.tolist())
    return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out  # type: ignore


def flip(img, sp_axes, lazy, transform_info):
    """
    Functional implementation of flip.
    This function operates eagerly or lazily according to
    ``lazy`` (default ``False``).

    Args:
        img: data to be changed, assuming `img` is channel-first.
        sp_axes: spatial axes along which to flip over.
            If None, will flip over all of the axes of the input array.
            If axis is negative it counts from the last to the first axis.
            If axis is a tuple of ints, flipping is performed on all of the axes
            specified in the tuple.
        lazy: a flag that indicates whether the operation should be performed lazily or not
        transform_info: a dictionary with the relevant information pertaining to an applied transform.
    """
    sp_size = img.peek_pending_shape() if isinstance(img, MetaTensor) else img.shape[1:]
    sp_size = convert_to_numpy(sp_size, wrap_sequence=True).tolist()
    extra_info = {"axes": sp_axes}  # track the spatial axes
    axes = monai.transforms.utils.map_spatial_axes(img.ndim, sp_axes)  # use the axes with channel dim
    rank = img.peek_pending_rank() if isinstance(img, MetaTensor) else torch.tensor(3.0, dtype=torch.double)
    # axes include the channel dim
    xform = torch.eye(int(rank) + 1, dtype=torch.double)
    for axis in axes:
        sp = axis - 1
        xform[sp, sp], xform[sp, -1] = xform[sp, sp] * -1, sp_size[sp] - 1
    meta_info = TraceableTransform.track_transform_meta(
        img, sp_size=sp_size, affine=xform, extra_info=extra_info, transform_info=transform_info, lazy=lazy
    )
    out = _maybe_new_metatensor(img)
    if lazy:
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else meta_info
    out = torch.flip(out, axes)
    return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out


def resize(
    img, out_size, mode, align_corners, dtype, input_ndim, anti_aliasing, anti_aliasing_sigma, lazy, transform_info
):
    """
    Functional implementation of resize.
    This function operates eagerly or lazily according to
    ``lazy`` (default ``False``).

    Args:
        img: data to be changed, assuming `img` is channel-first.
        out_size: expected shape of spatial dimensions after resize operation.
        mode: {``"nearest"``, ``"nearest-exact"``, ``"linear"``,
            ``"bilinear"``, ``"bicubic"``, ``"trilinear"``, ``"area"``}
            The interpolation mode.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.interpolate.html
        align_corners: This only has an effect when mode is
            'linear', 'bilinear', 'bicubic' or 'trilinear'.
        dtype: data type for resampling computation. If None, use the data type of input data.
        input_ndim: number of spatial dimensions.
        anti_aliasing: whether to apply a Gaussian filter to smooth the image prior
            to downsampling. It is crucial to filter when downsampling
            the image to avoid aliasing artifacts. See also ``skimage.transform.resize``
        anti_aliasing_sigma: {float, tuple of floats}, optional
            Standard deviation for Gaussian filtering used when anti-aliasing.
        lazy: a flag that indicates whether the operation should be performed lazily or not
        transform_info: a dictionary with the relevant information pertaining to an applied transform.
    """
    img = convert_to_tensor(img, track_meta=get_track_meta())
    orig_size = img.peek_pending_shape() if isinstance(img, MetaTensor) else img.shape[1:]
    extra_info = {
        "mode": mode,
        "align_corners": align_corners if align_corners is not None else TraceKeys.NONE,
        "dtype": str(dtype)[6:],  # dtype as string; remove "torch": torch.float32 -> float32
        "new_dim": len(orig_size) - input_ndim,
    }
    meta_info = TraceableTransform.track_transform_meta(
        img,
        sp_size=out_size,
        affine=scale_affine(orig_size, out_size),
        extra_info=extra_info,
        orig_size=orig_size,
        transform_info=transform_info,
        lazy=lazy,
    )
    if lazy:
        if anti_aliasing and lazy:
            warnings.warn("anti-aliasing is not compatible with lazy evaluation.")
        out = _maybe_new_metatensor(img)
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else meta_info
    if tuple(convert_to_numpy(orig_size)) == out_size:
        out = _maybe_new_metatensor(img, dtype=torch.float32)
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out
    out = _maybe_new_metatensor(img)
    img_ = convert_to_tensor(out, dtype=dtype, track_meta=False)  # convert to a regular tensor
    if anti_aliasing and any(x < y for x, y in zip(out_size, img_.shape[1:])):
        factors = torch.div(torch.Tensor(list(img_.shape[1:])), torch.Tensor(out_size))
        if anti_aliasing_sigma is None:
            # if sigma is not given, use the default sigma in skimage.transform.resize
            anti_aliasing_sigma = torch.maximum(torch.zeros(factors.shape), (factors - 1) / 2).tolist()
        else:
            # if sigma is given, use the given value for downsampling axis
            anti_aliasing_sigma = list(ensure_tuple_rep(anti_aliasing_sigma, len(out_size)))
            for axis in range(len(out_size)):
                anti_aliasing_sigma[axis] = anti_aliasing_sigma[axis] * int(factors[axis] > 1)
        anti_aliasing_filter = GaussianSmooth(sigma=anti_aliasing_sigma)
        img_ = convert_to_tensor(anti_aliasing_filter(img_), track_meta=False)
    _, _m, _, _ = resolves_modes(mode, torch_interpolate_spatial_nd=len(img_.shape) - 1)
    resized = torch.nn.functional.interpolate(
        input=img_.unsqueeze(0), size=out_size, mode=_m, align_corners=align_corners
    )
    out, *_ = convert_to_dst_type(resized.squeeze(0), out, dtype=torch.float32)
    return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out


def rotate(img, angle, output_shape, mode, padding_mode, align_corners, dtype, lazy, transform_info):
    """
    Functional implementation of rotate.
    This function operates eagerly or lazily according to
    ``lazy`` (default ``False``).

    Args:
        img: data to be changed, assuming `img` is channel-first.
        angle: Rotation angle(s) in radians. should a float for 2D, three floats for 3D.
        output_shape: output shape of the rotated data.
        mode: {``"bilinear"``, ``"nearest"``}
            Interpolation mode to calculate output values.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
        padding_mode: {``"zeros"``, ``"border"``, ``"reflection"``}
            Padding mode for outside grid values.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
        align_corners: See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
        dtype: data type for resampling computation.
            If None, use the data type of input data. To be compatible with other modules,
            the output data type is always ``float32``.
        lazy: a flag that indicates whether the operation should be performed lazily or not
        transform_info: a dictionary with the relevant information pertaining to an applied transform.

    """

    im_shape = img.peek_pending_shape() if isinstance(img, MetaTensor) else img.shape[1:]
    input_ndim = len(im_shape)
    if input_ndim not in (2, 3):
        raise ValueError(f"Unsupported image dimension: {input_ndim}, available options are [2, 3].")
    _angle = ensure_tuple_rep(angle, 1 if input_ndim == 2 else 3)
    transform = create_rotate(input_ndim, _angle)
    if output_shape is None:
        corners = np.asarray(np.meshgrid(*[(0, dim) for dim in im_shape], indexing="ij")).reshape((len(im_shape), -1))
        corners = transform[:-1, :-1] @ corners  # type: ignore
        output_shape = np.asarray(corners.ptp(axis=1) + 0.5, dtype=int)
    else:
        output_shape = np.asarray(output_shape, dtype=int)
    shift = create_translate(input_ndim, ((np.array(im_shape) - 1) / 2).tolist())
    shift_1 = create_translate(input_ndim, (-(np.asarray(output_shape, dtype=int) - 1) / 2).tolist())
    transform = shift @ transform @ shift_1
    extra_info = {
        "rot_mat": transform,
        "mode": mode,
        "padding_mode": padding_mode,
        "align_corners": align_corners if align_corners is not None else TraceKeys.NONE,
        "dtype": str(dtype)[6:],  # dtype as string; remove "torch": torch.float32 -> float32
    }
    meta_info = TraceableTransform.track_transform_meta(
        img,
        sp_size=output_shape,
        affine=transform,
        extra_info=extra_info,
        orig_size=im_shape,
        transform_info=transform_info,
        lazy=lazy,
    )
    out = _maybe_new_metatensor(img)
    if lazy:
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else meta_info
    _, _m, _p, _ = resolves_modes(mode, padding_mode)
    xform = AffineTransform(
        normalized=False, mode=_m, padding_mode=_p, align_corners=align_corners, reverse_indexing=True
    )
    img_t = out.to(dtype)
    transform_t, *_ = convert_to_dst_type(transform, img_t)
    output: torch.Tensor = xform(img_t.unsqueeze(0), transform_t, spatial_size=tuple(int(i) for i in output_shape))
    output = output.float().squeeze(0)
    out, *_ = convert_to_dst_type(output, dst=out, dtype=torch.float32)
    return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out


def zoom(img, scale_factor, keep_size, mode, padding_mode, align_corners, dtype, lazy, transform_info):
    """
    Functional implementation of zoom.
    This function operates eagerly or lazily according to
    ``lazy`` (default ``False``).

    Args:
        img: data to be changed, assuming `img` is channel-first.
        scale_factor: The zoom factor along the spatial axes.
            If a float, zoom is the same for each spatial axis.
            If a sequence, zoom should contain one value for each spatial axis.
        keep_size: Whether keep original size (padding/slicing if needed).
        mode: {``"bilinear"``, ``"nearest"``}
            Interpolation mode to calculate output values.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
        padding_mode: {``"zeros"``, ``"border"``, ``"reflection"``}
            Padding mode for outside grid values.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
        align_corners: See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
        dtype: data type for resampling computation.
            If None, use the data type of input data. To be compatible with other modules,
            the output data type is always ``float32``.
        lazy: a flag that indicates whether the operation should be performed lazily or not
        transform_info: a dictionary with the relevant information pertaining to an applied transform.

    """
    im_shape = img.peek_pending_shape() if isinstance(img, MetaTensor) else img.shape[1:]
    output_size = [int(math.floor(float(i) * z)) for i, z in zip(im_shape, scale_factor)]
    xform = scale_affine(im_shape, output_size)
    extra_info = {
        "mode": mode,
        "align_corners": align_corners if align_corners is not None else TraceKeys.NONE,
        "dtype": str(dtype)[6:],  # dtype as string; remove "torch": torch.float32 -> float32
        "do_padcrop": False,
        "padcrop": {},
    }
    if keep_size:
        do_pad_crop = not np.allclose(output_size, im_shape)
        if do_pad_crop and lazy:  # update for lazy evaluation
            _pad_crop = ResizeWithPadOrCrop(spatial_size=im_shape, mode=padding_mode)
            _pad_crop.lazy = True
            _tmp_img = MetaTensor([], affine=torch.eye(len(output_size) + 1))
            _tmp_img.push_pending_operation({LazyAttr.SHAPE: list(output_size), LazyAttr.AFFINE: xform})
            lazy_cropped = _pad_crop(_tmp_img)
            if isinstance(lazy_cropped, MetaTensor):
                xform = lazy_cropped.peek_pending_affine()
                extra_info["padcrop"] = lazy_cropped.pending_operations[-1]
            extra_info["do_padcrop"] = do_pad_crop
        output_size = [int(i) for i in im_shape]
    meta_info = TraceableTransform.track_transform_meta(
        img,
        sp_size=output_size,
        affine=xform,
        extra_info=extra_info,
        orig_size=im_shape,
        transform_info=transform_info,
        lazy=lazy,
    )
    out = _maybe_new_metatensor(img)
    if lazy:
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else meta_info
    img_t = out.to(dtype)
    _, _m, _, _ = resolves_modes(mode, torch_interpolate_spatial_nd=len(img_t.shape) - 1)
    zoomed: NdarrayOrTensor = torch.nn.functional.interpolate(
        recompute_scale_factor=True,
        input=img_t.unsqueeze(0),
        scale_factor=list(scale_factor),
        mode=_m,
        align_corners=align_corners,
    ).squeeze(0)
    out, *_ = convert_to_dst_type(zoomed, dst=out, dtype=torch.float32)
    if isinstance(out, MetaTensor):
        out = out.copy_meta_from(meta_info)
    do_pad_crop = not np.allclose(output_size, zoomed.shape[1:])
    if do_pad_crop:
        _pad_crop = ResizeWithPadOrCrop(spatial_size=img_t.shape[1:], mode=padding_mode)
        out = _pad_crop(out)
    if get_track_meta() and do_pad_crop:
        padcrop_xform = out.applied_operations.pop()
        out.applied_operations[-1]["extra_info"]["do_padcrop"] = True
        out.applied_operations[-1]["extra_info"]["padcrop"] = padcrop_xform
    return out


def rotate90(img, axes, k, lazy, transform_info):
    """
    Functional implementation of rotate90.
    This function operates eagerly or lazily according to
    ``lazy`` (default ``False``).

    Args:
        img: data to be changed, assuming `img` is channel-first.
        axes: 2 int numbers, defines the plane to rotate with 2 spatial axes.
            If axis is negative it counts from the last to the first axis.
        k: number of times to rotate by 90 degrees.
        lazy: a flag that indicates whether the operation should be performed lazily or not
        transform_info: a dictionary with the relevant information pertaining to an applied transform.
    """
    extra_info = {"axes": [d - 1 for d in axes], "k": k}
    ori_shape = img.peek_pending_shape() if isinstance(img, MetaTensor) else img.shape[1:]
    sp_shape = list(ori_shape)
    if k in (1, 3):
        a_0, a_1 = axes[0] - 1, axes[1] - 1
        sp_shape[a_0], sp_shape[a_1] = ori_shape[a_1], ori_shape[a_0]
    rank = img.peek_pending_rank() if isinstance(img, MetaTensor) else torch.tensor(3.0, dtype=torch.double)
    r, sp_r = int(rank), len(ori_shape)
    xform = to_affine_nd(r, create_translate(sp_r, [-float(d - 1) / 2 for d in sp_shape]))
    s = -1.0 if int(axes[0]) - int(axes[1]) in (-1, 2) else 1.0
    if sp_r == 2:
        rot90 = to_affine_nd(r, create_rotate(sp_r, [s * np.pi / 2]))
    else:
        idx = {1, 2, 3} - set(axes)
        angle: list[float] = [0, 0, 0]
        angle[idx.pop() - 1] = s * np.pi / 2
        rot90 = to_affine_nd(r, create_rotate(sp_r, angle))
    for _ in range(k):
        xform = rot90 @ xform
    xform = to_affine_nd(r, create_translate(sp_r, [float(d - 1) / 2 for d in ori_shape])) @ xform
    meta_info = TraceableTransform.track_transform_meta(
        img,
        sp_size=sp_shape,
        affine=xform,
        extra_info=extra_info,
        orig_size=ori_shape,
        transform_info=transform_info,
        lazy=lazy,
    )
    out = _maybe_new_metatensor(img)
    if lazy:
        return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else meta_info
    out = torch.rot90(out, k, axes)
    return out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out


def affine_func(
    img, affine, grid, resampler, sp_size, mode, padding_mode, do_resampling, image_only, lazy, transform_info
):
    """
    Functional implementation of affine.
    This function operates eagerly or lazily according to
    ``lazy`` (default ``False``).

    Args:
        img: data to be changed, assuming `img` is channel-first.
        affine: the affine transformation to be applied, it can be a 3x3 or 4x4 matrix. This should be defined
            for the voxel space spatial centers (``float(size - 1)/2``).
        grid: used in non-lazy mode to pre-compute the grid to do the resampling.
        resampler: the resampler function, see also: :py:class:`monai.transforms.Resample`.
        sp_size: output image spatial size.
        mode: {``"bilinear"``, ``"nearest"``} or spline interpolation order 0-5 (integers).
            Interpolation mode to calculate output values.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
            When it's an integer, the numpy (cpu tensor)/cupy (cuda tensor) backends will be used
            and the value represents the order of the spline interpolation.
            See also: https://docs.scipy.org/doc/scipy/reference/generated/scipy.ndimage.map_coordinates.html
        padding_mode: {``"zeros"``, ``"border"``, ``"reflection"``}
            Padding mode for outside grid values.
            See also: https://pytorch.org/docs/stable/generated/torch.nn.functional.grid_sample.html
            When `mode` is an integer, using numpy/cupy backends, this argument accepts
            {'reflect', 'grid-mirror', 'constant', 'grid-constant', 'nearest', 'mirror', 'grid-wrap', 'wrap'}.
            See also: https://docs.scipy.org/doc/scipy/reference/generated/scipy.ndimage.map_coordinates.html
        do_resampling: whether to do the resampling, this is a flag for the use case of updating metadata but
            skipping the actual (potentially heavy) resampling operation.
        image_only: if True return only the image volume, otherwise return (image, affine).
        lazy: a flag that indicates whether the operation should be performed lazily or not
        transform_info: a dictionary with the relevant information pertaining to an applied transform.

    """

    # resampler should carry the align_corners and type info
    img_size = img.peek_pending_shape() if isinstance(img, MetaTensor) else img.shape[1:]
    rank = img.peek_pending_rank() if isinstance(img, MetaTensor) else torch.tensor(3.0, dtype=torch.double)
    extra_info = {
        "affine": affine,
        "mode": mode,
        "padding_mode": padding_mode,
        "do_resampling": do_resampling,
        "align_corners": resampler.align_corners,
    }
    affine = monai.transforms.Affine.compute_w_affine(rank, affine, img_size, sp_size)
    meta_info = TraceableTransform.track_transform_meta(
        img,
        sp_size=sp_size,
        affine=affine,
        extra_info=extra_info,
        orig_size=img_size,
        transform_info=transform_info,
        lazy=lazy,
    )
    if lazy:
        out = _maybe_new_metatensor(img)
        out = out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else meta_info
        return out if image_only else (out, affine)
    if do_resampling:
        out = resampler(img=img, grid=grid, mode=mode, padding_mode=padding_mode)
        out = _maybe_new_metatensor(out)
    else:
        out = _maybe_new_metatensor(img, dtype=torch.float32, device=resampler.device)
    out = out.copy_meta_from(meta_info) if isinstance(out, MetaTensor) else out
    return out if image_only else (out, affine)