RESUMO
This study evaluated deep learning algorithms for semantic segmentation and quantification of intracerebral hemorrhage (ICH), perihematomal edema (PHE), and intraventricular hemorrhage (IVH) on noncontrast CT scans of patients with spontaneous ICH. Models were assessed on 1732 annotated baseline noncontrast CT scans obtained from the Tranexamic Acid for Hyperacute Primary Intracerebral Haemorrhage (ie, TICH-2) international multicenter trial (ISRCTN93732214), and different loss functions using a three-dimensional no-new-U-Net (nnU-Net) were examined to address class imbalance (30% of participants with IVH in dataset). On the test cohort (n = 174, 10% of dataset), the top-performing models achieved median Dice similarity coefficients of 0.92 (IQR, 0.89-0.94), 0.66 (0.58-0.71), and 1.00 (0.87-1.00), respectively, for ICH, PHE, and IVH segmentation. U-Net-based networks showed comparable, satisfactory performances on ICH and PHE segmentations (P > .05), but all nnU-Net variants achieved higher accuracy than the Brain Lesion Analysis and Segmentation Tool for CT (BLAST-CT) and DeepLabv3+ for all labels (P < .05). The Focal model showed improved performance in IVH segmentation compared with the Tversky, two-dimensional nnU-Net, U-Net, BLAST-CT, and DeepLabv3+ models (P < .05). Focal achieved concordance values of 0.98, 0.88, and 0.99 for ICH, PHE, and ICH volumes, respectively. The mean volumetric differences between the ground truth and prediction were 0.32 mL (95% CI: -8.35, 9.00), 1.14 mL (-9.53, 11.8), and 0.06 mL (-1.71, 1.84), respectively. In conclusion, U-Net-based networks provide accurate segmentation on CT images of spontaneous ICH, and Focal loss can address class imbalance. International Clinical Trials Registry Platform (ICTRP) no. ISRCTN93732214 Supplemental material is available for this article. © RSNA, 2022 Keywords: Head/Neck, Brain/Brain Stem, Hemorrhage, Segmentation, Quantification, Convolutional Neural Network (CNN), Deep Learning Algorithms, Machine Learning Algorithms.
RESUMO
Chirality, the ability of some molecules to exist as two non-superimposable mirror images, profoundly influences both chemistry and biology. Advances in deep learning enable the automatic recognition of chemical structure diagrams, however, studies on discovering the molecule chirality are scarce and the machine-readable molecular representations are not always sufficient to fully support the encoding of this important property. Here, we pretrained networks on a ChEMBL+ dataset (79641 molecules) and fine-tuned them for the binary classification of chirality (achiral/chiral) or multilabel chirality type classifications (none/centre/axial/planar). To address the label combination imbalanced problem in the multilabel task, the study proposed a Formulated Imbalanced Dataset Sampler (FIDS) to sample a formulated amount of minority label combinations on top of the training set. On a 10-fold cross validation experiment using our CHIRAL dataset (1142 manually curated molecules), our models achieved up to an accuracy of 90 % in the binary task. In the multilabel task incorporated with FIDS, the overall performance increases from 87 % to 89 % and the accuracy per label combination can attained up to a 50 % increase. Through the study of heatmaps, our work also exemplified the potential of deep neural network to make predictions based on the actual location of chirality elements.
