diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index 9621a1fe95..86c01b09dc 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -30,22 +30,14 @@ repos:
     rev: v0.7.0
     hooks:
     -   id: ruff
-        args:
-        - --fix
-
-  - repo: https://github.com/asottile/pyupgrade
-    rev: v3.19.0
-    hooks:
-      - id: pyupgrade
-        args: [--py39-plus, --keep-runtime-typing]
-        name: Upgrade code with exceptions
+        args: ["--fix"]
         exclude: |
           (?x)(
               ^versioneer.py|
               ^monai/_version.py|
-              ^monai/networks/| # avoid typing rewrites
-              ^monai/apps/detection/utils/anchor_utils.py| # avoid typing rewrites
-              ^tests/test_compute_panoptic_quality.py # avoid typing rewrites
+              ^monai/networks/| # todo: avoid typing rewrites
+              ^monai/apps/detection/utils/anchor_utils.py| # todo: avoid typing rewrites
+              ^tests/test_compute_panoptic_quality.py # todo: avoid typing rewrites
           )
 
   - repo: https://github.com/asottile/yesqa
diff --git a/monai/apps/deepgrow/dataset.py b/monai/apps/deepgrow/dataset.py
index 802d86e0c7..e597188e74 100644
--- a/monai/apps/deepgrow/dataset.py
+++ b/monai/apps/deepgrow/dataset.py
@@ -201,14 +201,9 @@ def _save_data_2d(vol_idx, vol_image, vol_label, dataset_dir, relative_path):
         logging.warning(f"Unique labels {unique_labels_count} exceeds 20. Please check if this is correct.")
 
     logging.info(
-        "{} => Image Shape: {} => {}; Label Shape: {} => {}; Unique Labels: {}".format(
-            vol_idx,
-            vol_image.shape,
-            image_count,
-            vol_label.shape if vol_label is not None else None,
-            label_count,
-            unique_labels_count,
-        )
+        f"{vol_idx} => Image Shape: {vol_image.shape} => {image_count};"
+        f" Label Shape: {vol_label.shape if vol_label is not None else None} => {label_count};"
+        f" Unique Labels: {unique_labels_count}"
     )
     return data_list
 
@@ -259,13 +254,8 @@ def _save_data_3d(vol_idx, vol_image, vol_label, dataset_dir, relative_path):
         logging.warning(f"Unique labels {unique_labels_count} exceeds 20. Please check if this is correct.")
 
     logging.info(
-        "{} => Image Shape: {} => {}; Label Shape: {} => {}; Unique Labels: {}".format(
-            vol_idx,
-            vol_image.shape,
-            image_count,
-            vol_label.shape if vol_label is not None else None,
-            label_count,
-            unique_labels_count,
-        )
+        f"{vol_idx} => Image Shape: {vol_image.shape} => {image_count};"
+        f" Label Shape: {vol_label.shape if vol_label is not None else None} => {label_count};"
+        f" Unique Labels: {unique_labels_count}"
     )
     return data_list
diff --git a/monai/apps/nnunet/nnunet_bundle.py b/monai/apps/nnunet/nnunet_bundle.py
index df8f09bf4b..47a0755ddf 100644
--- a/monai/apps/nnunet/nnunet_bundle.py
+++ b/monai/apps/nnunet/nnunet_bundle.py
@@ -13,7 +13,7 @@
 import os
 import shutil
 from pathlib import Path
-from typing import Any, Optional, Union
+from typing import Any
 
 import numpy as np
 import torch
@@ -36,9 +36,9 @@
 
 
 def get_nnunet_trainer(
-    dataset_name_or_id: Union[str, int],
+    dataset_name_or_id: str | int,
     configuration: str,
-    fold: Union[int, str],
+    fold: int | str,
     trainer_class_name: str = "nnUNetTrainer",
     plans_identifier: str = "nnUNetPlans",
     use_compressed_data: bool = False,
@@ -46,7 +46,7 @@ def get_nnunet_trainer(
     only_run_validation: bool = False,
     disable_checkpointing: bool = False,
     device: str = "cuda",
-    pretrained_model: Optional[str] = None,
+    pretrained_model: str | None = None,
 ) -> Any:  # type: ignore
     """
     Get the nnUNet trainer instance based on the provided configuration.
@@ -166,7 +166,7 @@ class ModelnnUNetWrapper(torch.nn.Module):
     restoring network architecture, and setting up the predictor for inference.
     """
 
-    def __init__(self, predictor: object, model_folder: Union[str, Path], model_name: str = "model.pt"):  # type: ignore
+    def __init__(self, predictor: object, model_folder: str | Path, model_name: str = "model.pt"):  # type: ignore
         super().__init__()
         self.predictor = predictor
 
@@ -294,7 +294,7 @@ def forward(self, x: MetaTensor) -> MetaTensor:
         return MetaTensor(out_tensor, meta=x.meta)
 
 
-def get_nnunet_monai_predictor(model_folder: Union[str, Path], model_name: str = "model.pt") -> ModelnnUNetWrapper:
+def get_nnunet_monai_predictor(model_folder: str | Path, model_name: str = "model.pt") -> ModelnnUNetWrapper:
     """
     Initializes and returns a `nnUNetMONAIModelWrapper` containing the corresponding `nnUNetPredictor`.
     The model folder should contain the following files, created during training:
@@ -426,9 +426,9 @@ def get_network_from_nnunet_plans(
     plans_file: str,
     dataset_file: str,
     configuration: str,
-    model_ckpt: Optional[str] = None,
+    model_ckpt: str | None = None,
     model_key_in_ckpt: str = "model",
-) -> Union[torch.nn.Module, Any]:
+) -> torch.nn.Module | Any:
     """
     Load and initialize a nnUNet network based on nnUNet plans and configuration.
 
@@ -518,7 +518,7 @@ def convert_monai_bundle_to_nnunet(nnunet_config: dict, bundle_root_folder: str,
     from nnunetv2.utilities.dataset_name_id_conversion import maybe_convert_to_dataset_name
 
     def subfiles(
-        folder: Union[str, Path], prefix: Optional[str] = None, suffix: Optional[str] = None, sort: bool = True
+        folder: str | Path, prefix: str | None = None, suffix: str | None = None, sort: bool = True
     ) -> list[str]:
         res = [
             i.name
diff --git a/monai/handlers/stats_handler.py b/monai/handlers/stats_handler.py
index 214872fef4..1dafde0f54 100644
--- a/monai/handlers/stats_handler.py
+++ b/monai/handlers/stats_handler.py
@@ -260,7 +260,7 @@ def _default_iteration_print(self, engine: Engine) -> None:
                         "ignoring non-scalar output in StatsHandler,"
                         " make sure `output_transform(engine.state.output)` returns"
                         " a scalar or dictionary of key and scalar pairs to avoid this warning."
-                        " {}:{}".format(name, type(value))
+                        f" {name}:{type(value)}"
                     )
                     continue  # not printing multi dimensional output
                 out_str += self.key_var_format.format(name, value.item() if isinstance(value, torch.Tensor) else value)
@@ -273,7 +273,7 @@ def _default_iteration_print(self, engine: Engine) -> None:
                 "ignoring non-scalar output in StatsHandler,"
                 " make sure `output_transform(engine.state.output)` returns"
                 " a scalar or a dictionary of key and scalar pairs to avoid this warning."
-                " {}".format(type(loss))
+                f" {type(loss)}"
             )
 
         if not out_str:
diff --git a/monai/handlers/tensorboard_handlers.py b/monai/handlers/tensorboard_handlers.py
index 44a03710de..20e2d74c8c 100644
--- a/monai/handlers/tensorboard_handlers.py
+++ b/monai/handlers/tensorboard_handlers.py
@@ -257,7 +257,7 @@ def _default_iteration_writer(self, engine: Engine, writer: SummaryWriter | Summ
                         "ignoring non-scalar output in TensorBoardStatsHandler,"
                         " make sure `output_transform(engine.state.output)` returns"
                         " a scalar or dictionary of key and scalar pairs to avoid this warning."
-                        " {}:{}".format(name, type(value))
+                        f" {name}:{type(value)}"
                     )
                     continue  # not plot multi dimensional output
                 self._write_scalar(
@@ -280,7 +280,7 @@ def _default_iteration_writer(self, engine: Engine, writer: SummaryWriter | Summ
                 "ignoring non-scalar output in TensorBoardStatsHandler,"
                 " make sure `output_transform(engine.state.output)` returns"
                 " a scalar or a dictionary of key and scalar pairs to avoid this warning."
-                " {}".format(type(loss))
+                f" {type(loss)}"
             )
         writer.flush()
 
diff --git a/monai/losses/adversarial_loss.py b/monai/losses/adversarial_loss.py
index c7be79243f..b2c27a41ee 100644
--- a/monai/losses/adversarial_loss.py
+++ b/monai/losses/adversarial_loss.py
@@ -57,8 +57,7 @@ def __init__(
 
         if criterion.lower() not in list(AdversarialCriterions):
             raise ValueError(
-                "Unrecognised criterion entered for Adversarial Loss. Must be one in: %s"
-                % ", ".join(AdversarialCriterions)
+                f"Unrecognised criterion entered for Adversarial Loss. Must be one in: {', '.join(AdversarialCriterions)}"
             )
 
         # Depending on the criterion, a different activation layer is used.
diff --git a/monai/losses/dice.py b/monai/losses/dice.py
index ed88100edd..948749606b 100644
--- a/monai/losses/dice.py
+++ b/monai/losses/dice.py
@@ -494,7 +494,7 @@ def __init__(
             raise ValueError(f"dist_matrix must be C x C, got {dist_matrix.shape[0]} x {dist_matrix.shape[1]}.")
 
         if weighting_mode not in ["default", "GDL"]:
-            raise ValueError("weighting_mode must be either 'default' or 'GDL, got %s." % weighting_mode)
+            raise ValueError(f"weighting_mode must be either 'default' or 'GDL', got {weighting_mode}.")
 
         self.m = dist_matrix
         if isinstance(self.m, np.ndarray):
diff --git a/monai/losses/ds_loss.py b/monai/losses/ds_loss.py
index 6a604aa22d..ef359bcfd0 100644
--- a/monai/losses/ds_loss.py
+++ b/monai/losses/ds_loss.py
@@ -11,7 +11,6 @@
 
 from __future__ import annotations
 
-from typing import Union
 
 import torch
 import torch.nn.functional as F
@@ -70,7 +69,7 @@ def get_loss(self, input: torch.Tensor, target: torch.Tensor) -> torch.Tensor:
             target = F.interpolate(target, size=input.shape[2:], mode=self.interp_mode)
         return self.loss(input, target)  # type: ignore[no-any-return]
 
-    def forward(self, input: Union[None, torch.Tensor, list[torch.Tensor]], target: torch.Tensor) -> torch.Tensor:
+    def forward(self, input: None | torch.Tensor | list[torch.Tensor], target: torch.Tensor) -> torch.Tensor:
         if isinstance(input, (list, tuple)):
             weights = self.get_weights(levels=len(input))
             loss = torch.tensor(0, dtype=torch.float, device=target.device)
diff --git a/monai/losses/focal_loss.py b/monai/losses/focal_loss.py
index 28d1c0cdc9..a1145c521e 100644
--- a/monai/losses/focal_loss.py
+++ b/monai/losses/focal_loss.py
@@ -13,7 +13,6 @@
 
 import warnings
 from collections.abc import Sequence
-from typing import Optional
 
 import torch
 import torch.nn.functional as F
@@ -153,7 +152,7 @@ def forward(self, input: torch.Tensor, target: torch.Tensor) -> torch.Tensor:
         if target.shape != input.shape:
             raise ValueError(f"ground truth has different shape ({target.shape}) from input ({input.shape})")
 
-        loss: Optional[torch.Tensor] = None
+        loss: torch.Tensor | None = None
         input = input.float()
         target = target.float()
         if self.use_softmax:
@@ -203,7 +202,7 @@ def forward(self, input: torch.Tensor, target: torch.Tensor) -> torch.Tensor:
 
 
 def softmax_focal_loss(
-    input: torch.Tensor, target: torch.Tensor, gamma: float = 2.0, alpha: Optional[float] = None
+    input: torch.Tensor, target: torch.Tensor, gamma: float = 2.0, alpha: float | None = None
 ) -> torch.Tensor:
     """
     FL(pt) = -alpha * (1 - pt)**gamma * log(pt)
@@ -225,7 +224,7 @@ def softmax_focal_loss(
 
 
 def sigmoid_focal_loss(
-    input: torch.Tensor, target: torch.Tensor, gamma: float = 2.0, alpha: Optional[float] = None
+    input: torch.Tensor, target: torch.Tensor, gamma: float = 2.0, alpha: float | None = None
 ) -> torch.Tensor:
     """
     FL(pt) = -alpha * (1 - pt)**gamma * log(pt)
diff --git a/monai/losses/perceptual.py b/monai/losses/perceptual.py
index 2ae03bc8dc..7774c39f10 100644
--- a/monai/losses/perceptual.py
+++ b/monai/losses/perceptual.py
@@ -95,8 +95,7 @@ def __init__(
 
         if network_type.lower() not in list(PercetualNetworkType):
             raise ValueError(
-                "Unrecognised criterion entered for Adversarial Loss. Must be one in: %s"
-                % ", ".join(PercetualNetworkType)
+                "Unrecognised criterion entered for Perceptual Loss. Must be one in: {}".format(", ".join(PercetualNetworkType))
             )
 
         if cache_dir:
diff --git a/monai/losses/spatial_mask.py b/monai/losses/spatial_mask.py
index a4c16236a2..0f823410dd 100644
--- a/monai/losses/spatial_mask.py
+++ b/monai/losses/spatial_mask.py
@@ -14,7 +14,7 @@
 import inspect
 import warnings
 from collections.abc import Callable
-from typing import Any, Optional
+from typing import Any
 
 import torch
 from torch.nn.modules.loss import _Loss
@@ -47,7 +47,7 @@ def __init__(
         if not callable(self.loss):
             raise ValueError("The loss function is not callable.")
 
-    def forward(self, input: torch.Tensor, target: torch.Tensor, mask: Optional[torch.Tensor] = None) -> torch.Tensor:
+    def forward(self, input: torch.Tensor, target: torch.Tensor, mask: torch.Tensor | None = None) -> torch.Tensor:
         """
         Args:
             input: the shape should be BNH[WD].
diff --git a/monai/losses/sure_loss.py b/monai/losses/sure_loss.py
index fa8820885d..30aed2a4dc 100644
--- a/monai/losses/sure_loss.py
+++ b/monai/losses/sure_loss.py
@@ -11,7 +11,7 @@
 
 from __future__ import annotations
 
-from typing import Callable, Optional
+from typing import Callable
 
 import torch
 import torch.nn as nn
@@ -42,10 +42,10 @@ def sure_loss_function(
     operator: Callable,
     x: torch.Tensor,
     y_pseudo_gt: torch.Tensor,
-    y_ref: Optional[torch.Tensor] = None,
-    eps: Optional[float] = -1.0,
-    perturb_noise: Optional[torch.Tensor] = None,
-    complex_input: Optional[bool] = False,
+    y_ref: torch.Tensor | None = None,
+    eps: float = -1.0,
+    perturb_noise: torch.Tensor | None = None,
+    complex_input: bool = False,
 ) -> torch.Tensor:
     """
     Args:
@@ -131,7 +131,7 @@ class SURELoss(_Loss):
     (https://arxiv.org/pdf/2310.01799.pdf)
     """
 
-    def __init__(self, perturb_noise: Optional[torch.Tensor] = None, eps: Optional[float] = None) -> None:
+    def __init__(self, perturb_noise: torch.Tensor | None = None, eps: float | None = None) -> None:
         """
         Args:
             perturb_noise (torch.Tensor, optional): The noise vector of shape
@@ -149,8 +149,8 @@ def forward(
         operator: Callable,
         x: torch.Tensor,
         y_pseudo_gt: torch.Tensor,
-        y_ref: Optional[torch.Tensor] = None,
-        complex_input: Optional[bool] = False,
+        y_ref: torch.Tensor | None = None,
+        complex_input: bool = False,
     ) -> torch.Tensor:
         """
         Args:
diff --git a/monai/metrics/utils.py b/monai/metrics/utils.py
index 972ec0061e..606a54669b 100644
--- a/monai/metrics/utils.py
+++ b/monai/metrics/utils.py
@@ -13,7 +13,7 @@
 
 import warnings
 from collections.abc import Iterable, Sequence
-from functools import lru_cache, partial
+from functools import partial, cache
 from types import ModuleType
 from typing import Any
 
@@ -465,7 +465,7 @@ def prepare_spacing(
 ENCODING_KERNEL = {2: [[8, 4], [2, 1]], 3: [[[128, 64], [32, 16]], [[8, 4], [2, 1]]]}
 
 
-@lru_cache(maxsize=None)
+@cache
 def _get_neighbour_code_to_normals_table(device=None):
     """
     returns a lookup table. For every binary neighbour code (2x2x2 neighbourhood = 8 neighbours = 8 bits = 256 codes)
diff --git a/monai/transforms/utility/array.py b/monai/transforms/utility/array.py
index 2ac37f2f81..3dc7897feb 100644
--- a/monai/transforms/utility/array.py
+++ b/monai/transforms/utility/array.py
@@ -21,7 +21,7 @@
 from collections.abc import Hashable, Mapping, Sequence
 from copy import deepcopy
 from functools import partial
-from typing import Any, Callable, Union
+from typing import Any, Callable
 
 import numpy as np
 import torch
@@ -1217,7 +1217,7 @@ def __init__(self, name: str, *args, **kwargs) -> None:
         transform, _ = optional_import("torchio.transforms", "0.18.0", min_version, name=name)
         self.trans = transform(*args, **kwargs)
 
-    def __call__(self, img: Union[NdarrayOrTensor, Mapping[Hashable, NdarrayOrTensor]]):
+    def __call__(self, img: NdarrayOrTensor | Mapping[Hashable, NdarrayOrTensor]):
         """
         Args:
             img: an instance of torchio.Subject, torchio.Image, numpy.ndarray, torch.Tensor, SimpleITK.Image,
@@ -1249,7 +1249,7 @@ def __init__(self, name: str, *args, **kwargs) -> None:
         transform, _ = optional_import("torchio.transforms", "0.18.0", min_version, name=name)
         self.trans = transform(*args, **kwargs)
 
-    def __call__(self, img: Union[NdarrayOrTensor, Mapping[Hashable, NdarrayOrTensor]]):
+    def __call__(self, img: NdarrayOrTensor | Mapping[Hashable, NdarrayOrTensor]):
         """
         Args:
             img: an instance of torchio.Subject, torchio.Image, numpy.ndarray, torch.Tensor, SimpleITK.Image,
diff --git a/pyproject.toml b/pyproject.toml
index add6642dba..ef85068ad7 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -46,6 +46,7 @@ target-version = "py39"
 select = [
   "E", "F", "W", # flake8
   "NPY",         # NumPy specific rules
+  "UP",          # pyupgrade
 ]
 extend-ignore = [
   "E741", # ambiguous variable name
diff --git a/tests/profile_subclass/profiling.py b/tests/profile_subclass/profiling.py
index ffa6a8b17d..18aecea2fb 100644
--- a/tests/profile_subclass/profiling.py
+++ b/tests/profile_subclass/profiling.py
@@ -63,9 +63,8 @@ def main():
 
         b_min, b_avg, b_med, b_std = bench(tensor_1, tensor_2)
         print(
-            "Type {} time (microseconds):  min: {}, avg: {}, median: {}, and std {}.".format(
-                t.__name__, (10**6 * b_min), (10**6) * b_avg, (10**6) * b_med, (10**6) * b_std
-            )
+            f"Type {t.__name__} time (microseconds):"
+            f"  min: {10**6 * b_min}, avg: {(10**6) * b_avg}, median: {(10**6) * b_med}, and std {(10**6) * b_std}."
         )