RuntimeError: quantile() input tensor is too large

[TerminalVelocityDPro]

Hello! I was wanting to use your NeSVoR tookit to reconstruct some misaligned 2D slices of liver MRI I was working on, but I'm running into some issues regarding usage. Firstly, I'm using three stacks of slices that I believe to be orthogonal to one another of the liver MRI that I converted from dicom to nifti. However, when I try to run the default reconstruction algorithm or stack to stack registration, I keep on receiving the same error:

========== == CUDA ==

CUDA Version 11.7.1

Container image Copyright (c) 2016-2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved.

This container image and its contents are governed by the NVIDIA Deep Learning Container License. By pulling and using the container, you accept the terms and conditions of this license: https://developer.nvidia.com/ngc/nvidia-deep-learning-container-license

A copy of this license is made available in this container at /NGC-DL-CONTAINER-LICENSE for your convenience.

/usr/local/lib/python3.10/dist-packages/tinycudann/modules.py:52: UserWarning: tinycudann was built for lower compute capability (70) than the system's (86). Performance may be suboptimal. warnings.warn(f"tinycudann was built for lower compute capability ({cc}) than the system's ({system_compute_capability}). Performance may be suboptimal.") 2024-09-15 19:53:47 [INFO] Data loading starts ... 2024-09-15 19:53:47 [INFO] loading stacks 2024-09-15 19:55:20 [INFO] background thresholding 2024-09-15 19:55:20 [INFO] Data loading finished in 92.5 s 2024-09-15 19:55:20 [INFO] Registration starts ... 2024-09-15 19:55:22 [WARNING] ROI in input stack "/incoming/orthogonal1.nii.gz" is too large for SVoRT 2024-09-15 19:55:22 [WARNING] ROI in input stack "/incoming/orthogonal2.nii.gz" is too large for SVoRT 2024-09-15 19:55:22 [WARNING] ROI in input stack "/incoming/orthogonal3.nii.gz" is too large for SVoRT 2024-09-15 19:56:09 [INFO] similarity score for SVoRT = 0.177652 2024-09-15 19:56:09 [INFO] similarity score for stack registration = 0.412441 2024-09-15 19:56:09 [INFO] use stack transformation 2024-09-15 19:56:09 [ERROR] Unhandled exception: Traceback (most recent call last): File "/usr/local/bin/nesvor", line 33, in sys.exit(load_entry_point('nesvor', 'console_scripts', 'nesvor')()) File "/usr/local/NeSVoR/nesvor/cli/main.py", line 23, in main run(args) File "/usr/local/NeSVoR/nesvor/cli/main.py", line 48, in run getattr(commands, command_class)(args).main() File "/usr/local/NeSVoR/nesvor/cli/commands.py", line 75, in main self.exec() File "/usr/local/NeSVoR/nesvor/cli/commands.py", line 160, in exec input_dict = self.preprocess() File "/usr/local/NeSVoR/nesvor/cli/commands.py", line 150, in preprocess input_dict["input_slices"] = _register( File "/usr/local/NeSVoR/nesvor/cli/commands.py", line 393, in _register slices = svort_predict( File "/usr/local/NeSVoR/nesvor/svort/inference.py", line 561, in svort_predict return run_svort(dataset, model, svort, vvr, force_vvr, force_scanner) File "/usr/local/NeSVoR/nesvor/svort/inference.py", line 525, in run_svort stack.slices /= torch.quantile(stack.slices[stack.mask], 0.99) # normalize RuntimeError: quantile() input tensor is too large

That runtime error keeps appearing, and I was wondering if you know what might be happening to cause it to occur.

I know that the toolkit was originally intended for fetal brain/body reconstruction, but I thought that it could be extended to usage for other body parts as well.

Thanks for your help, I'd appreciate any guidance you can provide.