Search

There exists a large number of datasets for organ segmentation, which are partially annotated and sequentially constructed. A typical dataset is constructed at a certain time by curating medical images and annotating the organs of interest. In other words, new datasets with annotations of new organ categories are built over time. To unleash the potential behind these partially labeled, sequentially-constructed datasets, we propose to incrementally learn a multi-organ segmentation model. In each incremental learning (IL) stage, we lose the access to previous data and annotations, whose knowledge is assumingly captured by the current model, and gain the access to a new dataset with annotations of new organ categories, from which we learn to update the organ segmentation model to include the new organs. While IL is notorious for its `catastrophic forgetting' weakness in the context of natural image analysis, we experimentally discover that such a weakness mostly disappears for CT multi-organ segmentation. To further stabilize the model performance across the IL stages, we introduce a light memory module and some loss functions to restrain the representation of different categories in feature space, aggregating feature representation of the same class and separating feature representation of different classes. Extensive experiments on five open-sourced datasets are conducted to illustrate the effectiveness of our method., Comment: arXiv admin note: text overlap with arXiv:2103.04526

Chemical pressure is an effective method to tune physical properties, particularly for diluted magnetic semiconductors (DMS) of which ferromagnetic ordering is mediated by charge carriers. Via substitution of smaller Ca for larger Sr, we introduce chemical pressure on (Sr,Na)(Cd,Mn)2As2 to fabricate a new DMS material (Ca,Na)(Cd,Mn)2As2. Carriers and spins are introduced by substitutions of (Ca,Na) and (Cd,Mn) respectively. The unit cell volume reduces by 6.2% after complete substitution of Ca for Sr, suggesting a subsistent chemical pressure. Importantly the local geometry of [Cd/MnAs4] tetrahedron is optimized via chemical compression that increases the Mn-As hybridization leading to enhanced ferromagnetic interactions. As a result, the maximum Curie temperature (TC) is increased by about 50% while the the maximum saturation moment increases by over 100% from (Sr,Na)(Cd,Mn)2As2 to (Ca,Na)(Cd,Mn)2As2. The chemical pressure estimated from the equation of state is equal to an external physical pressure of 3.6 GPa., Comment: 19 pages, 4 figures