|
--- |
|
tags: |
|
- spacy |
|
- token-classification |
|
language: |
|
- en |
|
license: mit |
|
model-index: |
|
- name: en_biobert_ner_symptom |
|
results: |
|
- task: |
|
name: NER |
|
type: token-classification |
|
metrics: |
|
- name: NER Precision |
|
type: precision |
|
value: 0.9997017596 |
|
- name: NER Recall |
|
type: recall |
|
value: 0.9994036971 |
|
- name: NER F Score |
|
type: f_score |
|
value: 0.9995527061 |
|
widget: |
|
- text: "Patient X reported coughing and sneezing." |
|
example_title: "Example 1" |
|
- text: "There was a case of rash and inflammation." |
|
example_title: "Example 2" |
|
- text: "He complained of dizziness during the trip." |
|
example_title: "Example 3" |
|
- text: "I felt distressed , giddy and nauseous during my stay in Florida." |
|
example_title: "Example 4" |
|
- text: "Mr. Y complained of breathlesness and chest pain when he was driving back to his house." |
|
example_title: "Example 5" |
|
--- |
|
BioBERT based NER model for medical symptoms |
|
|
|
| Feature | Description | |
|
| --- | --- | |
|
| **Name** | `en_biobert_ner_symptom` | |
|
| **Version** | `1.0.0` | |
|
| **spaCy** | `>=3.5.1,<3.6.0` | |
|
| **Default Pipeline** | `transformer`, `ner` | |
|
| **Components** | `transformer`, `ner` | |
|
| **Vectors** | 0 keys, 0 unique vectors (0 dimensions) | |
|
| **Sources** | n/a | |
|
| **License** | `MIT` | |
|
| **Author** | [Sena Chae, Pratik Maitra, Padmini Srinivasan]() | |
|
|
|
|
|
|
|
<b> <u> Model Description </u> </b> |
|
|
|
The model was trained on a combined maccrobat and i2c2 dataset and is based on biobert. If you use this model kindly cite the paper below: |
|
|
|
<b> |
|
<i> |
|
Developing a BioBERT-based Natural Language Processing Algorithm for Acute Myeloid Leukemia Symptoms Identification from Clinical Notes - Sena Chae , Pratik Maitra , Padmini Srinivasan |
|
</i> |
|
</b> |
|
|
|
<b> <u> How to use the Model </u> </b> |
|
|
|
<div class="wrapper"> |
|
<span class="inner"> |
|
|
|
from transformers import pipeline, AutoTokenizer, AutoModelForTokenClassification |
|
|
|
tokenizer = AutoTokenizer.from_pretrained("d4data/biomedical-ner-all") |
|
model = AutoModelForTokenClassification.from_pretrained("d4data/biomedical-ner-all") |
|
|
|
pipe = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="simple") # pass device=0 if using gpu |
|
pipe("""The patient reported no recurrence of palpitations at follow-up 6 months after the ablation.""") |
|
|
|
</span> |
|
</div> |
|
|
|
### Accuracy |
|
|
|
| Type | Score | |
|
| --- | --- | |
|
| `ENTS_F` | 99.96 | |
|
| `ENTS_P` | 99.97 | |
|
| `ENTS_R` | 99.94 | |
|
| `TRANSFORMER_LOSS` | 20456.83 | |
|
| `NER_LOSS` | 38920.06 | |
