Roadmap - innat/medic-ai GitHub Wiki
Currently
- Tutorials
- Covid-19 CT Scan Segmentation
- Covid-19 CT Scan Classification
- BraTS (Brain Tumor Segmentation) 3D Segmentation
- BTCV Abdomen 3D Segmentation
- Spleen Segmentation
- Models
- UNETR
- 3D-TransUnet
- MediCLIP
- VISTA
- MedSAM | MedSAM-V2 | SAM-Med-3D
- 3D ImageNet (Classification + Segmentation-UNet)
- DenseNet
- EfficientNet
- ConvNeXt
- SegResNet
- ResNet-V2
- Utility
- MASI
- Gradient Checkpointing
- Sliding Window Inference for ONNX runtime
Backend Compatibility
- TensorFlow backend support.
- Torch backend support.
- Jax backend support.
Device Compatibility
- Multi-GPU support.
- TPU-V3 support.
2025.04.15
- Callback
-
SlidingWindowInferenceCallback
-
- Inference
-
SlidingWindowInference
-
2025.03.20
- Model
-
Swin Transformer
: 3D Classification Model.
-
- Transformation
-
RandRotate90
-
RandSpatialCrop
-
Resize
-
RandShiftIntensity
-
NormalizeIntensity
-
TensorBundle
-
- Metrics
-
SparseDiceMetric
-
- Losses
-
BinaryDiceCELoss
-
CategoricalDiceCELoss
-
SparseDiceCELoss
-
SparseDiceLoss
-
2025.02.25
- Model
-
SwinUNETR
: 3D Segmentation Model.
-
- Transformation
-
RandCropByPosNegLabel
-
CropForeground
-
Spacing
-
SignalFillEmpty
-
Orientation
-
RandFlip
-
- Metrics
-
BinaryDiceMetric
-
CategoricalDiceMetric
-
- Losses
-
BinaryDiceLoss
-
CategoricalDiceLoss
-