Knee Segmentation Tool is a software application used to segment structures in 3D medical images. Knee Segmentation Tool provides fully automatic segmentation using artificial intelligence, as well as manual delineation and image navigation. In addition to these core functions, Knee Segmentation Tool offers many supporting utilities.
Knee Segmentation Tool offers many supporting utilities. Some of the core advantages of Knee Segmentation Tool include:
Knee Segmentation Tool will automatically detect different DICOM series and convert them to the correct NIFTI files.
Knee Segmentation Tool supports visualisation and navigation in both 2D and 3D. Including that, does Knee Segmentation Tool volume sampling and opacity control.
Knee Segmentation Tool uses AI to do sematic segmentation with the accuracy of a professional trained surgeon, but uses just a fraction of the time.
Knee Segmentation Tool supports manual annotation to add/subtract structures, or create a additional structures to indicate abnormalities.
Knee Segmentation Tool can collect statistics as voxel counts and cubic millimeters in the different structures of the 3D volume.
Knee Segmentation Tool supports 3D model exports to industri standard foramts. Making it possible to visualize the models in VR/AR or 3D print them.
Knee Segmentation Tool makes it easy from start to end. With only a few steps along the way, can you visualize your MRI images in 3D in a few minutes.
For best results, use the suggested protocols provided under resources. Choose a protocol consisting of T1, PD, FS and an optional ANGIO sequence. All volumes scanned must share the same geometry.
Convert your MRI DICOM images into the Nifti format using this feature. Watch and learn from our tutorials.
Select the input nifti file directory and choose the desired window resolution. Keep in mind that higher resolutions demand more available computer resources. Please be patient as the software is inferring.
You may now view, annotate and export your results.
This study explores deep learning techniques for segmenting knee anatomy into 13 classes using a specialized MR protocol. By analyzing 40 healthy knee volumes and a knee with an ACL tear, the research utilized DenseVNet neural networks across multiple sequence combinations. The comprehensive analysis revealed high accuracy in anatomical labeling, with superior performance observed when combining all MR sequences. Notably, the deep learning model successfully identified an ACL tear, showcasing potential as a diagnostic and preoperative planning tool in orthopedics.
Read MoreCelebrating the recognition and awards received for our innovative solutions in medical imaging.
EuroMinnies award "Best New Radiology Software 2021"