Search

Background: Many methodologies for selecting histopathological images, such as sample image patches or segment histology from regions of interest (ROIs) or whole-slide images (WSIs), have been utilized to develop survival models. With gigapixel WSIs exhibiting diverse histological appearances, obtaining clinically prognostic and explainable features remains challenging. Therefore, we propose a novel deep learning-based algorithm combining tissue areas with histopathological features to predict cancer survival., Methods: The Cancer Genome Atlas Colon Adenocarcinoma (TCGA-COAD) dataset was used in this investigation. A deep convolutional survival model (DeepConvSurv) extracted histopathological information from the image patches of nine different tissue types, including tumors, lymphocytes, stroma, and mucus. The tissue map of the WSIs was segmented using image processing techniques that involved localizing and quantifying the tissue region. Six survival models with the concordance index (C-index) were used as the evaluation metrics., Results: We extracted 128 histopathological features from four histological types and five tissue area features from WSIs to predict colorectal cancer survival. Our method performed better in six distinct survival models than the Whole Slide Histopathological Images Survival Analysis framework (WSISA), which adaptively sampled patches using K-means from WSIs. The best performance using histopathological features was 0.679 using LASSO-Cox. Compared to histopathological features alone, tissue area features increased the C-index by 2.5%. Based on histopathological features and tissue area features, our approach achieved performance of 0.704 with RIDGE-Cox., Conclusions: A deep learning-based algorithm combining histopathological features with tissue area proved clinically relevant and effective for predicting cancer survival., (© 2023. BioMed Central Ltd., part of Springer Nature.)

Background: Biomedical named entity recognition (NER) is a fundamental task of biomedical text mining that finds the boundaries of entity mentions in biomedical text and determines their entity type. To accelerate the development of biomedical NER techniques in Spanish, the PharmaCoNER organizers launched a competition to recognize pharmacological substances, compounds, and proteins. Biomedical NER is usually recognized as a sequence labeling task, and almost all state-of-the-art sequence labeling methods ignore the meaning of different entity types. In this paper, we investigate some methods to introduce the meaning of entity types in deep learning methods for biomedical NER and apply them to the PharmaCoNER 2019 challenge. The meaning of each entity type is represented by its definition information., Material and Method: We investigate how to use entity definition information in the following two methods: (1) SQuad-style machine reading comprehension (MRC) methods that treat entity definition information as query and biomedical text as context and predict answer spans as entities. (2) Span-level one-pass (SOne) methods that predict entity spans of one type by one type and introduce entity type meaning, which is represented by entity definition information. All models are trained and tested on the PharmaCoNER 2019 corpus, and their performance is evaluated by strict micro-average precision, recall, and F1-score., Results: Entity definition information brings improvements to both SQuad-style MRC and SOne methods by about 0.003 in micro-averaged F1-score. The SQuad-style MRC model using entity definition information as query achieves the best performance with a micro-averaged precision of 0.9225, a recall of 0.9050, and an F1-score of 0.9137, respectively. It outperforms the best model of the PharmaCoNER 2019 challenge by 0.0032 in F1-score. Compared with the state-of-the-art model without using manually-crafted features, our model obtains a 1% improvement in F1-score, which is significant. These results indicate that entity definition information is useful for deep learning methods on biomedical NER., Conclusion: Our entity definition information enhanced models achieve the state-of-the-art micro-average F1 score of 0.9137, which implies that entity definition information has a positive impact on biomedical NER detection. In the future, we will explore more entity definition information from knowledge graph., (© 2021. The Author(s).)

Background: Whole brain radiotherapy (WBRT) can impair patients' cognitive function. Hippocampal avoidance during WBRT can potentially prevent this side effect. However, manually delineating the target area is time-consuming and difficult. Here, we proposed a credible approach of automatic hippocampal delineation based on convolutional neural networks., Methods: Referring to the hippocampus contouring atlas proposed by RTOG 0933, we manually delineated (MD) the hippocampus on the MRI data sets (3-dimensional T1-weighted with slice thickness of 1 mm, n = 175), which were used to construct a three-dimensional convolutional neural network aiming for the hippocampus automatic delineation (AD). The performance of this AD tool was tested on three cohorts: (a) 3D T1 MRI with 1-mm slice thickness (n = 30); (b) non-3D T1-weighted MRI with 3-mm slice thickness (n = 19); (c) non-3D T1-weighted MRI with 1-mm slice thickness (n = 11). All MRIs confirmed with normal hippocampus has not been violated by any disease. Virtual radiation plans were created for AD and MD hippocampi in cohort c to evaluate the clinical feasibility of the artificial intelligence approach. Statistical analyses were performed using SPSS version 23. P < 0.05 was considered significant., Results: The Dice similarity coefficient (DSC) and Average Hausdorff Distance (AVD) between the AD and MD hippocampi are 0.86 ± 0.028 and 0.18 ± 0.050 cm in cohort a, 0.76 ± 0.035 and 0.31 ± 0.064 cm in cohort b, 0.80 ± 0.015 and 0.24 ± 0.021 cm in cohort c, respectively. The DSC and AVD in cohort a were better than those in cohorts b and c (P < 0.01). There is no significant difference between the radiotherapy plans generated using the AD and MD hippocampi., Conclusion: The AD of the hippocampus based on a deep learning algorithm showed satisfying results, which could have a positive impact on improving delineation accuracy and reducing work load.

Background: Family history (FH) information, including family members, side of family of family members (i.e., maternal or paternal), living status of family members, observations (diseases) of family members, etc., is very important in the decision-making process of disorder diagnosis and treatment. However FH information cannot be used directly by computers as it is always embedded in unstructured text in electronic health records (EHRs). In order to extract FH information form clinical text, there is a need of natural language processing (NLP). In the BioCreative/OHNLP2018 challenge, there is a task regarding FH extraction (i.e., task1), including two subtasks: (1) entity identification, identifying family members and their observations (diseases) mentioned in clinical text; (2) family history extraction, extracting side of family of family members, living status of family members, and observations of family members. For this task, we propose a system based on deep joint learning methods to extract FH information. Our system achieves the highest F1- scores of 0.8901 on subtask1 and 0.6359 on subtask2, respectively.

Background: Extracting relationships between chemicals and diseases from unstructured literature have attracted plenty of attention since the relationships are very useful for a large number of biomedical applications such as drug repositioning and pharmacovigilance. A number of machine learning methods have been proposed for chemical-induced disease (CID) extraction due to some publicly available annotated corpora. Most of them suffer from time-consuming feature engineering except deep learning methods. In this paper, we propose a novel document-level deep learning method, called recurrent piecewise convolutional neural networks (RPCNN), for CID extraction., Results: Experimental results on a benchmark dataset, the CDR (Chemical-induced Disease Relation) dataset of the BioCreative V challenge for CID extraction show that the highest precision, recall and F-score of our RPCNN-based CID extraction system are 65.24, 77.21 and 70.77%, which is competitive with other state-of-the-art systems., Conclusions: A novel deep learning method is proposed for document-level CID extraction, where domain knowledge, piecewise strategy, attention mechanism, and multi-instance learning are combined together. The effectiveness of the method is proved by experiments conducted on a benchmark dataset.

Backgrounds: Knowledge discovery from breast cancer treatment records has promoted downstream clinical studies such as careflow mining and therapy analysis. However, the clinical treatment text from electronic health data might be recorded by different doctors under their hospital guidelines, making the final data rich in author- and domain-specific idiosyncrasies. Therefore, breast cancer treatment entity normalization becomes an essential task for the above downstream clinical studies. The latest studies have demonstrated the superiority of deep learning methods in named entity normalization tasks. Fundamentally, most existing approaches adopt pipeline implementations that treat it as an independent process after named entity recognition, which can propagate errors to later tasks. In addition, despite its importance in clinical and translational research, few studies directly deal with the normalization task in Chinese clinical text due to the complexity of composition forms.Methods: To address these issues, we propose PASCAL, an end-to-end and accurate framework for breast cancer treatment entity normalization (TEN). PASCAL leverages a gated convolutional neural network to obtain a representation vector that can capture contextual features and long-term dependencies. Additionally, it treats treatment entity recognition (TER) as an auxiliary task that can provide meaningful information to the primary TEN task and as a particular regularization to further optimize the shared parameters. Finally, by concatenating the context-aware vector and probabilistic distribution vector from TEN, we utilize the conditional random field layer (CRF) to model the normalization sequence and predict the TEN sequential results.Results: To evaluate the effectiveness of the proposed framework, we employ the three latest sequential models as baselines and build the model in single- and multitask on a real-world database. Experimental results show that our method achieves better accuracy and efficiency than state-of-the-art approaches.Conclusions: The effectiveness and efficiency of the presented pseudo cascade learning framework were validated for breast cancer treatment normalization in clinical text. We believe the predominant performance lies in its ability to extract valuable information from unstructured text data, which will significantly contribute to downstream tasks, such as treatment recommendations, breast cancer staging and careflow mining. [ABSTRACT FROM AUTHOR]