Timofey's picture
Update README.md
799069a
---
tags:
- ANDDigest
- ANDSystem
extra_gated_fields:
I agree to share my contact information: checkbox
widget:
- text: "Intermittent obstruction of jejunostomy tube due to Ascaris lumbricoides infection. A 45-year-old Costa Rican woman was seen for a jejunostomy tube malfunction. There was no evidence of tube malposition or intestinal obstruction. During endoscopy, a long worm was retrieved from the distal duodenum; it was later confirmed to be Ascaris lumbricoides. After treatment with <andsystem-candidate>, no further episodes of tube occlusion were observed. This case reminds us of the importance of considering helminthic infections and their atypical manifestations in patients from endemic regions."
example_title: "mebendazole"
- text: "Neurobrucellosis Presenting with Features of Demyelinating Disorder in a Pediatric Patient. Brucellosis is an endemic disease in Saudi Arabia, which can present with variable clinical manifestations. It is a zoonotic disease transmitted from animals to humans. Brucellosis is a multisystemic disease that can present with any system involvement; And neurobrucellosis is a serious complication, sometimes leading to permanent neurological deficit, if treatment is not started promptly. Herein, we present a 6-year boy with neurobrucellosis, who developed demyelination of cerebral white matter and presented with <andsystem-candidate> and seizures."
example_title: "fever"
---
This model is a fine-tuned model of [BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext](https://arxiv.org/abs/2007.15779) ([hugging-face card](https://huggingface.co./microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext)). The current model was developed for the web-based [ANDDigest](https://anddigest.sysbio.ru/) system for the classification of the short names of drugs and metabolites in texts on the basis of their context (the name considered to be short if it\'s length is 4 symbols or less). The analyzed name should be replaced in text with <andsystem-candidate> tag.<br>
<br>
<b>Input:</b><br>
Any biomedical text where a name of classified object is replaced with <andsystem-candidate> tag, for example, [this](https://pubmed.ncbi.nlm.nih.gov/18475230/) pubmed abstract:<br>
<i>Intermittent obstruction of jejunostomy tube due to Ascaris lumbricoides infection. A 45-year-old Costa Rican woman was seen for a jejunostomy tube malfunction. There was no evidence of tube malposition or intestinal obstruction. During endoscopy, a long worm was retrieved from the distal duodenum; it was later confirmed to be Ascaris lumbricoides. After treatment with <b>\<andsystem\-candidate\></b>, no further episodes of tube occlusion were observed. This case reminds us of the importance of considering helminthic infections and their atypical manifestations in patients from endemic regions.</i>
<br>
<br>In this example <i>mebendazole</i> was replaced with <i>\<andsystem\-candidate\></i>. Please keep in mind that maximum length of input sequence for BERT is limited to 512 tokens.
<br>
<b>Output:</b><br>
<i>LABEL_0</i> refers to the probability of the <i>FALSE</i> recognition, i.e. if the context of \<andsystem\-candidate\> doesn't corresponds to the context specific for drugs or metabolites.<br>
<i>LABEL_1</i> refers to the probability of the <i>TRUE</i> recognition, i.e. when the context of \<andsystem\-candidate\> corresponds to the context specific for drugs or metabolites.<br>
<br>
The optimal threshold value for the short names of drugs or metabolites for the LABEL_1, was calculated using a gold standard (add link). It is<b> >= 0.999992847442627</b>.<br>
<br>
The Mathew Correlation Coefficient of the model for the long names (\>= 15 symbols) is 0.983.<br>
The ROC AUC value of the model, calculated for the short names (\<\= 4 symbols) is 0.907.