Investigation of an efficient multi-modal convolutional neural network for multiple sclerosis lesion detection.

In this study, an automated 2D machine learning approach for fast and precise segmentation of MS lesions from multi-modal magnetic resonance images (mmMRI) is presented. The method is based on an U-Net like convolutional neural network (CNN) for automated 2D slice-based-segmentation of brain MRI volumes. The individual modalities are encoded in separate downsampling branches without weight sharing, to leverage the specific features. Skip connections input feature maps to multi-scale feature fusion (MSFF) blocks at every decoder stage of the network. Those are followed by multi-scale feature upsampling (MSFU) blocks which use the information about lesion shape and location. The CNN is evaluated on two publicly available datasets: The ISBI 2015 longitudinal MS lesion segmentation challenge dataset containing 19 subjects and the MICCAI 2016 MSSEG challenge dataset containing 15 subjects from various scanners. The proposed multi-input 2D architecture is among the top performing approaches in the ISBI challenge, to which open-access papers are available, is able to outperform state-of-the-art 3D approaches without additional post-processing, can be adapted to other scanners quickly, is robust against scanner variability and can be deployed for inference even on a standard laptop without a dedicated GPU.

Application of A U-Net for Map-like Segmentation and Classification of Discontinuous Fibrosis Distribution in Gd-EOB-DTPA-Enhanced Liver MRI.

We aimed to evaluate whether U-shaped convolutional neuronal networks can be used to segment liver parenchyma and indicate the degree of liver fibrosis/cirrhosis at the voxel level using contrast-enhanced magnetic resonance imaging. This retrospective study included 112 examinations with histologically determined liver fibrosis/cirrhosis grade (Ishak score) as the ground truth. The T1-weighted volume-interpolated breath-hold examination sequences of native, arterial, late arterial, portal venous, and hepatobiliary phases were semi-automatically segmented and co-registered. The segmentations were assigned the corresponding Ishak score. In a nested cross-validation procedure, five models of a convolutional neural network with U-Net architecture (nnU-Net) were trained, with the dataset being divided into stratified training/validation (n = 89/90) and holdout test datasets (n = 23/22). The trained models precisely segmented the test data (mean dice similarity coefficient = 0.938) and assigned separate fibrosis scores to each voxel, allowing localization-dependent determination of the degree of fibrosis. The per voxel results were evaluated by the histologically determined fibrosis score. The micro-average area under the receiver operating characteristic curve of this seven-class classification problem (Ishak score 0 to 6) was 0.752 for the test data. The top-three-accuracy-score was 0.750. We conclude that determining fibrosis grade or cirrhosis based on multiphase Gd-EOB-DTPA-enhanced liver MRI seems feasible using a 2D U-Net. Prospective studies with localized biopsies are needed to evaluate the reliability of this model in a clinical setting.

Regulation of activity in vitro and in vivo of three phospholipases B from Saccharomyces cerevisiae.

The genome of the yeast, Saccharomyces cerevisiae, contains three highly similar genes coding for phospholipases B/lysophospholipases. These enzymes behave differently with respect to substrate preferences in vitro and relative contributions to phospholipid catabolism in vivo [Merkel, Fido, Mayr, Pruger, Raab, Zandonella, Kohlwein and Paltauf (1999) J. Biol. Chem. 274, 28121-28127]. It is shown in the present study that, in vitro, pH markedly affects the substrate preference of Plb1p and Plb2p, but not of Plb3p. At the pH optimum of 2.5-3.5, the order of substrate preference of Plb1p and Plb2p is PtdSer (phosphatidylserine)>PtdIns>PtdCho (phosphatidylcholine>PtdEtn (phosphatidylethanolamine). At pH values of 5 and above, the substrate preferences change to PtdCho=PtdEtn for Plb1p and PtdSer=PtdEtn for Plb2p. Accordingly, with cultured cells the ratio of PtdIns/PtdCho breakdown, as reflected in the ratio of GroPIns (glycerophosphoinositol)/GroPCho (glycerophosphocholine) released into the culture medium, is inversely related to the pH of the growth medium. This effect is ascribed to the pH response of Plb1p, because Plb2p does not contribute to the degradation of PtdIns and PtdCho in vivo. Bivalent and tervalent cations activate phospholipases B at pH 5.5, but are inhibitory at pH 2.5. Al3+ at a concentration of 20 mM increases Plb1p activity in vitro by 8-fold and leads to a 9-fold increase in GroPCho release by whole cells. In vivo, cycloheximide strongly inhibits the breakdown of PtdIns, and to a lesser extent PtdCho. However, Al3+-stimulated GroPCho release is almost completely inhibited by cycloheximide. Deletion of PLB3 leads to increased sensitivity to toxic Al3+. Addition of SDS or melittin to cultured cells leads to a significant increase in phospholipid degradation, which is insensitive to inhibition by cycloheximide. Deletion mutants defective in the PLB1 gene are significantly more resistant to SDS than are wild-type cells.