Transferability of artificial neural networks for clinical document classification across hospitals: A case study on abnormality detection from radiology reports.

Application of machine learning techniques for automatic and reliable classification of clinical documents have shown promising results. However, machine learning models require abundant training data specific to each target hospital and may not be able to benefit from available labeled data from each of the hospitals due to data variations. Such training data limitations have presented one of the major obstacles for maximising potential application of machine learning approaches in the healthcare domain. We investigated transferability of artificial neural network models across hospitals from different domains representing various age demographic groups (i.e., children, adults, and mixed) in order to cope with such limitations. We explored the transferability of artificial neural networks for clinical document classification. Our case study was to detect abnormalities from limb X-ray reports obtained from the emergency department (ED) of three hospitals within different domains. Different transfer learning scenarios were investigated in order to employ a source hospital's trained model for addressing a target hospital's abnormality detection problem. A Convolutional Neural Network (CNN) model exhibited the best effectiveness compared to other networks when employing an embedding model trained on a large corpus of clinical documents. Furthermore, CNN models derived from a source hospital outperformed a conventional machine learning approach based on Support Vector Machines (SVM) when applied to a different (target) hospital. These models were further improved by leveraging available training data in target hospitals and outperformed the models that used only the target hospital data with F1-Score of 0.92-0.96 across three hospitals. Our transfer learning model used only simple vector representations of documents without any task-specific feature engineering. Transferring the CNN model significantly improved (approx.10% in F1-Score) the state-of-the-art approach for clinical document classification based on a trivial transferred model. In addition, the results showed that transfer learning techniques can further improve a CNN model that is trained only on either a source or target hospital's data. Transferring a pre-trained CNN model generated in one hospital to another facilitates application of machine learning approaches that alleviate both hospital-specific feature engineering and training data.

Hassanzadeh H ，Nguyen A ，Karimi S ，Chu K ... -  《-》

Automatic extraction of cancer registry reportable information from free-text pathology reports using multitask convolutional neural networks.

We implement 2 different multitask learning (MTL) techniques, hard parameter sharing and cross-stitch, to train a word-level convolutional neural network (CNN) specifically designed for automatic extraction of cancer data from unstructured text in pathology reports. We show the importance of learning related information extraction (IE) tasks leveraging shared representations across the tasks to achieve state-of-the-art performance in classification accuracy and computational efficiency. Multitask CNN (MTCNN) attempts to tackle document information extraction by learning to extract multiple key cancer characteristics simultaneously. We trained our MTCNN to perform 5 information extraction tasks: (1) primary cancer site (65 classes), (2) laterality (4 classes), (3) behavior (3 classes), (4) histological type (63 classes), and (5) histological grade (5 classes). We evaluated the performance on a corpus of 95 231 pathology documents (71 223 unique tumors) obtained from the Louisiana Tumor Registry. We compared the performance of the MTCNN models against single-task CNN models and 2 traditional machine learning approaches, namely support vector machine (SVM) and random forest classifier (RFC). MTCNNs offered superior performance across all 5 tasks in terms of classification accuracy as compared with the other machine learning models. Based on retrospective evaluation, the hard parameter sharing and cross-stitch MTCNN models correctly classified 59.04% and 57.93% of the pathology reports respectively across all 5 tasks. The baseline models achieved 53.68% (CNN), 46.37% (RFC), and 36.75% (SVM). Based on prospective evaluation, the percentages of correctly classified cases across the 5 tasks were 60.11% (hard parameter sharing), 58.13% (cross-stitch), 51.30% (single-task CNN), 42.07% (RFC), and 35.16% (SVM). Moreover, hard parameter sharing MTCNNs outperformed the other models in computational efficiency by using about the same number of trainable parameters as a single-task CNN. The hard parameter sharing MTCNN offers superior classification accuracy for automated coding support of pathology documents across a wide range of cancers and multiple information extraction tasks while maintaining similar training and inference time as those of a single task-specific model.

Alawad M ，Gao S ，Qiu JX ，Yoon HJ ，Blair Christian J ，Penberthy L ，Mumphrey B ，Wu XC ，Coyle L ，Tourassi G ... -  《-》

Deep Learning versus Conventional Machine Learning for Detection of Healthcare-Associated Infections in French Clinical Narratives.

The objective of this article was to compare the performances of health care-associated infection (HAI) detection between deep learning and conventional machine learning (ML) methods in French medical reports. The corpus consisted in different types of medical reports (discharge summaries, surgery reports, consultation reports, etc.). A total of 1,531 medical text documents were extracted and deidentified in three French university hospitals. Each of them was labeled as presence (1) or absence (0) of HAI. We started by normalizing the records using a list of preprocessing techniques. We calculated an overall performance metric, the F1 Score, to compare a deep learning method (convolutional neural network [CNN]) with the most popular conventional ML models (Bernoulli and multi-naïve Bayes, k-nearest neighbors, logistic regression, random forests, extra-trees, gradient boosting, support vector machines). We applied the hyperparameter Bayesian optimization for each model based on its HAI identification performances. We included the set of text representation as an additional hyperparameter for each model, using four different text representations (bag of words, term frequency-inverse document frequency, word2vec, and Glove). CNN outperforms all other conventional ML algorithms for HAI classification. The best F1 Score of 97.7% ± 3.6% and best area under the curve score of 99.8% ± 0.41% were achieved when CNN was directly applied to the processed clinical notes without a pretrained word2vec embedding. Through receiver operating characteristic curve analysis, we could achieve a good balance between false notifications (with a specificity equal to 0.937) and system detection capability (with a sensitivity equal to 0.962) using the Youden's index reference. The main drawback of CNNs is their opacity. To address this issue, we investigated CNN inner layers' activation values to visualize the most meaningful phrases in a document. This method could be used to build a phrase-based medical assistant algorithm to help the infection control practitioner to select relevant medical records. Our study demonstrated that deep learning approach outperforms other classification learning algorithms for automatically identifying HAIs in medical reports.

Rabhi S ，Jakubowicz J ，Metzger MH 《-》

Clinical Document Classification Using Labeled and Unlabeled Data Across Hospitals.

Hassanzadeh H ，Kholghi M ，Nguyen A ，Chu K ... -  《-》

Transfer learning for classification of cardiovascular tissues in histological images.

Automatic classification of healthy tissues and organs based on histology images is an open problem, mainly due to the lack of automated tools. Solutions in this regard have potential in educational medicine and medical practices. Some preliminary advances have been made using image processing techniques and classical supervised learning. Due to the breakthrough performance of deep learning in various areas, we present an approach to recognise and classify, automatically, fundamental tissues and organs using Convolutional Neural Networks (CNN). We adapt four popular CNNs architectures - ResNet, VGG19, VGG16 and Inception - to this problem through transfer learning. The resulting models are evaluated at three stages. Firstly, all the transferred networks are compared to each other. Secondly, the best resulting fine-tuned model is compared to an ad-hoc 2D multi-path model to outline the importance of transfer learning. Thirdly, the same model is evaluated against the state-of-the-art method, a cascade SVM using LBP-based descriptors, to contrast a traditional machine learning approach and a representation learning one. The evaluation task consists of separating six classes accurately: smooth muscle of the elastic artery, smooth muscle of the large vein, smooth muscle of the muscular artery, cardiac muscle, loose connective tissue, and light regions. The different networks are tuned on 6000 blocks of 100 × 100 pixels and tested on 7500. Our proposal yields F-score values between 0.717 and 0.928. The highest and lowest performances are for cardiac muscle and smooth muscle of the large vein, respectively. The main issue leading to limited classification scores for the latter class is its similarity with the elastic artery. However, this confusion is evidenced during manual annotation as well. Our algorithm reached improvements in F-score between 0.080 and 0.220 compared to the state-of-the-art machine learning approach. We conclude that it is possible to classify healthy cardiovascular tissues and organs automatically using CNNs and that deep learning holds great promise to improve tissue and organs classification. We left our training and test sets, models and source code publicly available to the research community.

Mazo C ，Bernal J ，Trujillo M ，Alegre E ... -  《-》