arXiv subjects

Satoshi Kasai

Publications and source records attributed to Satoshi Kasai.

22 records · Page 2Linked to original sources

arXiv2022

Multi-Modality Abdominal Multi-Organ Segmentation with Deep Supervised 3D Segmentation Model

To promote the development of medical image segmentation technology, AMOS, a large-scale abdominal multi-organ dataset for versatile medical image segmentation, is provided and AMOS 2022 challenge is held by using the dataset. In this report, we present our solution for the AMOS 2022 challenge. We employ residual U-Net with deep super vision as our base model. The experimental results show that the mean scores of Dice similarity coefficient and normalized surface dice are 0.8504 and 0.8476 for CT only task and CT/MRI task, respectively.

eess.IV↗

arXiv2022

Computer Aided Diagnosis and Out-of-Distribution Detection in Glaucoma Screening Using Color Fundus Photography

Artificial Intelligence for RObust Glaucoma Screening (AIROGS) Challenge is held for developing solutions for glaucoma screening from color fundus photography that are robust to real-world scenarios. This report describes our method submitted to the AIROGS challenge. Our method employs convolutional neural networks to classify input images to "referable glaucoma" or "no referable glaucoma". In addition, we introduce an inference-time out-of-distribution (OOD) detection method to identify ungradable images. Our OOD detection is based on an energy-based method combined with activation rectification.

cs.CV↗

arXiv2022

Nuclei panoptic segmentation and composition regression with multi-task deep neural networks

Nuclear segmentation, classification and quantification within Haematoxylin & Eosin stained histology images enables the extraction of interpretable cell-based features that can be used in downstream explainable models in computational pathology. The Colon Nuclei Identification and Counting (CoNIC) Challenge is held to help drive forward research and innovation for automatic nuclei recognition in computational pathology. This report describes our proposed method submitted to the CoNIC challenge. Our method employs a multi-task learning framework, which performs a panoptic segmentation task and a regression task. For the panoptic segmentation task, we use encoder-decoder type deep neural networks predicting a direction map in addition to a segmentation map in order to separate neighboring nuclei into different instances

eess.IV↗

arXiv2005

Magnetization Losses in Multiply Connected YBa2Cu3O6+x Coated Conductors

We report the results of a magnetization losses study in experimental multifilament, multiply connected coated superconductors exposed to time-varying magnetic field. In these samples, the superconducting layer is divided into parallel stripes segregated by non-superconducting grooves. In order to facilitate the current sharing between the stripes and thus increase the reliability of the striated conductors, a sparse network of superconducting bridges is superimposed on the striated film. We find that the presence of the bridges does not substantially increase the magnetization losses, both hysteresis and coupling, as long as the number of bridges per length of the sample is not large. These results indicate that it is possible to find a reasonable compromise between the competing requirements of connectivity and loss reduction in an ac-tolerant version of the high temperature coated conductors specifically designed for ac power applications.

cond-mat.supr-con↗