# Version 3.4.0 (December 2025)

## What's new in 3.4.0

- Tested with the latest MONAI v1.5.1 and Holoscan v3.9 CUDA 12 version, [holoscan-cu12](https://pypi.org/project/holoscan-cu12).
- Added a new compressed DICOM image decoder for PyDicom, which is based on nvImageCodec for GPU-accelerated image frame decoding. This decoder is dynamically registered at runtime with PyDicom as the first choice for the supported transfer syntaxes.
  - It is worth noting that the invocation of the decoder is fully controlled by the PyDicom decode runner. A single frame is decoded at each invocation, even for multi-frame images. For this reason, batch decoding is not feasible, and the GPU acceleration benefit is offset by the overhead of moving pixels from system to GPU memory.
  - A custom decode runner can be used instead, in combination with PyDicom metadata parsing, to utilize this new decoder with batch decoding to improve performance. 


Please also see the <a href="https://github.com/Project-MONAI/monai-deploy-app-sdk/issues?q=is%3Aissue+is%3Aclosed">closed issues on Github</a> and the <a href="https://github.com/Project-MONAI/monai-deploy-app-sdk/pulls?q=is%3Apr+is%3Aclosed">closed pull requests on Github</a>.

## Additional information
Please visit [GETTING STARTED](/getting_started/index) guide and follow the tutorials.

You can learn more about SDK usage through [DEVELOPING WITH SDK](/developing_with_sdk/index).

Please let us know how you like it and what could be improved by [submitting an issue](https://github.com/Project-MONAI/monai-deploy-app-sdk/issues/new/choose) or [asking questions](https://github.com/Project-MONAI/monai-deploy-app-sdk/discussions).