Tigre failed to reconstruct a large dataset, error in Atb mex

Hello, I am running the Tigre toolbox downloaded from your GitHub using MATLAB, but I am currently experiencing a problem. When I used the Tigre toolbox to develop algorithms such as FDK, SART, and PICCS for reconstructing CBCT projection data, the code ran perfectly and the reconstruction results were correct when reconstructing the projection image of the simulated 3D shepp-logan phantom. At this point, my projection data consists of 360 TIFF images with a size of 256 * 256 and 32-bit. However, when I was reconstructing the real experimental projection data, an error occurred, which seemed to be referring to exceeding the memory. My actual experimental projection data consists of 720 32 bit, 2304 * 2940 sized TIFF images. I am certain that this is not due to insufficient memory on my computer device, as I have successfully reconstructed using the ASRTA toolbox. Even by using the ASRTA toolbox to reconstruct 1200 32-bit projection images in TIFF format with a size of 2304 * 2940, I was able to successfully reconstruct them. Excuse me, is it due to the Tigre toolbox that does not support the reconstruction of large datasets? The error message is as follows: It seems to specifically point to an issue with Atb mex

## Code to reproduce the problem (If applicable)
Error pinning memory
Incorrect use of Atc_mex

part or all of the requested memory range is already mapped
Error Atb (line 61)

img=Atb_mex(projections,geo,angles,ptype, gpuids.devices);
Error FDK (line 85)

res=Atb((proj),geo,angles, 'gpuids', gpuids); % Weighting is inside

Another device reported an error:
Check mem error
Incorrect use of Atc_mex
out of memory

Error Atb (line 61)
img=Atb_mex(projections,geo,angles,ptype, gpuids.devices);

Error FDK (line 85)
res=Atb((proj),geo,angles, 'gpuids', gpuids); % Weighting is inside
```

```

## Specifications

- MATLAB/python version:R2018a and R2024a
- OS:windows10
- CUDA version:12.0