Clinical Concept-Based Radiology Reports Classification Pipeline for Lung Carcinoma

Rising incidence and mortality of cancer have led to an incremental amount of research in the field. To learn from preexisting data, it has become important to capture maximum information related to disease type, stage, treatment, and outcomes. Medical imaging reports are rich in this kind of information but are only present as free text. The extraction of information from such unstructured text reports is labor-intensive. The use of Natural Language Processing (NLP) tools to extract information from radiology reports can make it less time-consuming as well as more effective. In this study, we have developed and compared different models for the classification of lung carcinoma reports using clinical concepts. This study was approved by the institutional ethics committee as a retrospective study with a waiver of informed consent. A clinical concept-based classification pipeline for lung carcinoma radiology reports was developed using rule-based as well as machine learning models and compared. The machine learning models used were XGBoost and two more deep learning model architectures with bidirectional long short-term neural networks. A corpus consisting of 1700 radiology reports including computed tomography (CT) and positron emission tomography/computed tomography (PET/CT) reports were used for development and testing. Five hundred one radiology reports from MIMIC-III Clinical Database version 1.4 was used for external validation. The pipeline achieved an overall F1 score of 0.94 on the internal set and 0.74 on external validation with the rule-based algorithm using expert input giving the best performance. Among the machine learning models, the Bi-LSTM_dropout model performed better than the ML model using XGBoost and the Bi-LSTM_simple model on internal set, whereas on external validation, the Bi-LSTM_simple model performed relatively better than other 2. This pipeline can be used for clinical concept-based classification of radiology reports related to lung carcinoma from a huge corpus and also for automated annotation of these reports.

     

Chronic pharmacological blockade of the Na+/Ca2+ exchanger modulates the growth and development of the Purkinje cell dendritic arbor in mouse cerebellar slice cultures

The Na⁺/Ca²⁺ exchanger (NCX) is a bidirectional plasma membrane antiporter involved in Ca²⁺ homeostasis in eukaryotes. NCX has three isoforms, NCX1–3, and all of them are expressed in the cerebellum. Immunostaining on cerebellar slice cultures indicates that NCX is widely expressed in the cerebellum, including expression in Purkinje cells. The pharmacological blockade of the forward mode of NCX (Ca²⁺ efflux mode) by bepridil moderately inhibited growth and development of Purkinje cell dendritic arbor in cerebellar slice cultures. However, the blockade of the reverse mode (Ca²⁺ influx mode) by KB‐R7943 severely reduced the dendritic arbor and induced a morphological change with thickened distal dendrites. The effect of KB‐R7943 on dendritic growth was unrelated to the activity of voltage‐gated calcium channels and was also apparent in the absence of bioelectrical activity indicating that it was mediated by NCX expressed in Purkinje cells. We have used additional NCX inhibitors including CB‐DMB, ORM‐10103, SEA0400, YM‐244769, and SN‐6 which have higher specificity for NCX isoforms and target either the forward, reverse, or both modes. These inhibitors caused a strong dendritic reduction similar to that seen with KB‐R7943, but did not elicit thickening of distal dendrites. Our findings indicate that disturbance of the NCX‐dependent calcium transport in Purkinje cells induces a reduction of dendritic arbor, which is presumably caused by changes in the calcium handling, and underline the importance of the calcium equilibrium for the dendritic development in cerebellar Purkinje cells.     

Isogenic autosomes to be applied in optimal screening for novel mutants with viable phenotypes in Drosophila melanogaster

Most insertional mutagenesis screens of Drosophila performed to date have not used target chromosomes that have been checked for their suitability for phenotypic screens for viable phenotypes. To address this, we have generated a selection of stocks carrying either isogenized second chromosomes or isogenized third chromosomes, in a genetic background derived from a Canton-S wild-type strain. We have tested these stocks for a range of behavioral and other viable phenotypes. As expected, most lines are statistically indistinguishable from Canton-S in most phenotypes tested. The lines generated are now being used as target chromosomes in mutagenesis screens, and the characterization reported here will facilitate their use in screens of these lines for behavioral and other viable phenotypes.