Three Benchmarks Running until March/April 2015
The VISCERAL project is organising three 3D medical imaging Benchmarks over the next 6 months:
- VISCERAL Anatomy3 Segmentation and Landmark Detection Benchmark
- VISCERALdetection Lesion Detection Benchmark
- VISCERALretrieval Case Retrieval Benchmark
Participants in these benchmarks will be able to present and discuss their results at a workshop to be held at the ISBI 2015 (Anatomy3 and VISCERALdetection) and a workshop to be held at the ECIR 2015 (VISCERALretrieval) (more details on the web pages).
More details on the individual benchmarks are below:
VISCERAL Anatomy3 Segmentation and Landmark Detection Benchmark: A set of annotated medical imaging data is provided to the participants, along with a powerful complimentary cloud-computing instance (8-core CPU with 16GB RAM) where participant algorithms can be developed and evaluated. The available data contains segmentations and landmarks of several different anatomical structures in different image modalities, e.g. CT and MRI. The participants, however, do NOT have to address all the tasks involved in such data, but rather they can attempt any sub-problem thereof. The following annotated structures are found in the training corpus. Segmentations: left/right kidney, spleen, liver, left/right lung, urinary bladder, rectus abdominis muscle, 1st lumbar vertebra, pancreas, left/right psoas major muscle, gallbladder, sternum, aorta, trachea, left/right adrenal gland. Landmarks: Lateral end of clavicula, crista iliaca, symphysis below, trochanter major, trochanter minor, tip of aortic arch, trachea bifurcation, aortic bifurcation, crista iliaca.
VISCERALdetection Lesion Detection Benchmark: We make available medical imaging data (CT, MRI, w & w/o contrast enhancement) that contains various lesions in anatomical regions such as the bones, liver, brain, lung, or lymph nodes. There are overall 308 annotated lesions in the data-set. During the training phase, participants will get imaging data and annotations in the form of lesion center position, and for large lesions, annotations that indicate the radius. During the benchmark phase, we will run participant algorithms on new data, and evaluate detection performance against expert annotations of lesions.
VISCERALretrieval Case Retrieval Benchmark: The VISCERALretrieval data set consists 2311 volumes originated from three different modalities (CT, MRT1, MRT2). For a subset of these volumes we provide anatomy-pathology terms extracted from the volume’s radiology report. The following scenario is modelled: a radiologist is assessing a query case in a clinical setting, e.g., a CT volume, and is searching for cases that are relevant in this assessment. The algorithm has to find cases that are relevant in a large database of cases. The participants have to develop an algorithm that finds clinically-relevant (related or useful for differential diagnosis) cases given a query case (imaging data only or imaging and text data), but not necessarily the final diagnosis.