Description
The National Institutes of Health Stroke Scale (NIHSS) is a 15-item neurologic examination stroke scale. It quantifies the physical manifestations of neurological deficits and provides crucial support for clinical decision making and early-stage emergency triage.
Predicted Entities
11_ExtinctionInattention, 6b_RightLeg, 1c_LOCCommands, 10_Dysarthria, NIHSS, 5_Motor, 8_Sensory, 4_FacialPalsy, 6_Motor, 2_BestGaze, Measurement, 6a_LeftLeg, 5b_RightArm, 5a_LeftArm, 1b_LOCQuestions, 3_Visual, 9_BestLanguage, 7_LimbAtaxia, 1a_LOC
Live Demo Open in Colab Copy S3 URI
How to use
document_assembler = DocumentAssembler()\
.setInputCol("text")\
.setOutputCol("document")
sentence_detector = SentenceDetectorDLModel.pretrained("sentence_detector_dl_healthcare", "en", "clinical/models") \
.setInputCols(["document"]) \
.setOutputCol("sentence")
tokenizer = Tokenizer()\
.setInputCols(["sentence"])\
.setOutputCol("token")
embeddings_clinical = WordEmbeddingsModel.pretrained('embeddings_clinical', 'en', 'clinical/models') \
.setInputCols(['sentence', 'token']) \
.setOutputCol('embeddings')
clinical_ner = MedicalNerModel.pretrained("ner_nihss", "en", "clinical/models") \
.setInputCols(["sentence", "token", "embeddings"]) \
.setOutputCol("ner")
ner_converter = NerConverter() \
.setInputCols(["sentence", "token", "ner"]) \
.setOutputCol("entities")
nlpPipeline = Pipeline(stages=[document_assembler, sentence_detector, tokenizer, embeddings_clinical, clinical_ner, ner_converter])
model = nlpPipeline.fit(spark.createDataFrame([[""]]).toDF("text"))
results = model.transform(spark.createDataFrame([["Abdomen , soft , nontender . NIH stroke scale on presentation was 23 to 24 for , one for consciousness , two for month and year and two for eye / grip , one to two for gaze , two for face , eight for motor , one for limited ataxia , one to two for sensory , three for best language and two for attention . On the neurologic examination the patient was intermittently"]], ["text"]))
val document_assembler = new DocumentAssembler()
.setInputCol("text")
.setOutputCol("document")
val sentence_detector = SentenceDetectorDLModel.pretrained("sentence_detector_dl_healthcare", "en", "clinical/models")
.setInputCols("document")
.setOutputCol("sentence")
val tokenizer = new Tokenizer()
.setInputCols("sentence")
.setOutputCol("token")
val embeddings_clinical = WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")
.setInputCols(Array("sentence", "token"))
.setOutputCol("embeddings")
val ner = MedicalNerModel.pretrained("ner_nihss", "en", "clinical/models")
.setInputCols(Array("sentence", "token", "embeddings"))
.setOutputCol("ner")
val ner_converter = new NerConverter()
.setInputCols(Array("sentence", "token", "ner"))
.setOutputCol("entities")
val pipeline = new Pipeline().setStages(Array(document_assembler, sentence_detector, tokenizer, embeddings_clinical, ner, ner_converter))
val data = Seq("""Abdomen , soft , nontender . NIH stroke scale on presentation was 23 to 24 for , one for consciousness , two for month and year and two for eye / grip , one to two for gaze , two for face , eight for motor , one for limited ataxia , one to two for sensory , three for best language and two for attention . On the neurologic examination the patient was intermittently""").toDS.toDF("text")
val result = pipeline.fit(data).transform(data)
import nlu
nlu.load("en.med_ner.nihss").predict("""Abdomen , soft , nontender . NIH stroke scale on presentation was 23 to 24 for , one for consciousness , two for month and year and two for eye / grip , one to two for gaze , two for face , eight for motor , one for limited ataxia , one to two for sensory , three for best language and two for attention . On the neurologic examination the patient was intermittently""")
Results
| | chunk | entity |
|---:|:-------------------|:-------------------------|
| 0 | NIH stroke scale | NIHSS |
| 1 | 23 to 24 | Measurement |
| 2 | one | Measurement |
| 3 | consciousness | 1a_LOC |
| 4 | two | Measurement |
| 5 | month and year and | 1b_LOCQuestions |
| 6 | two | Measurement |
| 7 | eye / grip | 1c_LOCCommands |
| 8 | one to | Measurement |
| 9 | two | Measurement |
| 10 | gaze | 2_BestGaze |
| 11 | two | Measurement |
| 12 | face | 4_FacialPalsy |
| 13 | eight | Measurement |
| 14 | one | Measurement |
| 15 | limited | 7_LimbAtaxia |
| 16 | ataxia | 7_LimbAtaxia |
| 17 | one to two | Measurement |
| 18 | sensory | 8_Sensory |
| 19 | three | Measurement |
| 20 | best language | 9_BestLanguage |
| 21 | two | Measurement |
| 22 | attention | 11_ExtinctionInattention |
Model Information
| Model Name: | ner_nihss |
| Compatibility: | Healthcare NLP 3.3.2+ |
| License: | Licensed |
| Edition: | Official |
| Input Labels: | [sentence, token, embeddings] |
| Output Labels: | [ner] |
| Language: | en |
Data Source
@article{wangnational, title={National Institutes of Health Stroke Scale (NIHSS) Annotations for the MIMIC-III Database}, author={Wang, Jiayang and Huang, Xiaoshuo and Yang, Lin and Li, Jiao} }
Benchmarking
label tp fp fn prec rec f1
I-NIHSS 126 5 14 0.96183205 0.9 0.92988926
B-NIHSS 152 9 14 0.94409937 0.91566265 0.9296636
B-5b_RightArm 33 0 7 1.0 0.825 0.90410954
I-Measurement 17 1 69 0.9444444 0.19767442 0.3269231
I-5_Motor 12 3 2 0.8 0.85714287 0.82758623
I-1a_LOC 134 1 3 0.9925926 0.9781022 0.9852941
I-9_BestLanguage 85 3 0 0.96590906 1.0 0.982659
B-7_LimbAtaxia 39 0 3 1.0 0.9285714 0.9629629
B-4_FacialPalsy 53 4 4 0.9298246 0.9298246 0.9298246
B-1a_LOC 39 0 5 1.0 0.8863636 0.939759
B-6a_LeftLeg 35 0 4 1.0 0.8974359 0.945946
B-10_Dysarthria 51 2 4 0.9622642 0.92727274 0.94444454
B-8_Sensory 43 3 7 0.9347826 0.86 0.8958333
I-6b_RightLeg 149 5 0 0.96753246 1.0 0.9834984
B-3_Visual 40 0 6 1.0 0.8695652 0.9302325
B-5_Motor 4 1 4 0.8 0.5 0.61538464
I-6a_LeftLeg 157 1 6 0.9936709 0.9631902 0.9781932
B-5a_LeftArm 37 1 4 0.9736842 0.902439 0.93670887
I-4_FacialPalsy 129 5 4 0.96268654 0.9699248 0.96629214
B-2_BestGaze 41 1 2 0.97619045 0.95348835 0.9647058
I-8_Sensory 78 0 10 1.0 0.8863636 0.939759
I-5b_RightArm 153 1 12 0.9935065 0.92727274 0.95924765
B-9_BestLanguage 45 1 5 0.9782609 0.9 0.9375
I-5a_LeftArm 159 3 10 0.9814815 0.9408284 0.96072507
I-1c_LOCCommands 109 1 0 0.9909091 1.0 0.9954338
I-6_Motor 12 4 4 0.75 0.75 0.75
B-1b_LOCQuestions 43 0 2 1.0 0.95555556 0.97727275
B-6b_RightLeg 32 1 2 0.969697 0.9411765 0.9552239
I-2_BestGaze 112 0 2 1.0 0.98245615 0.99115044
I-7_LimbAtaxia 113 1 0 0.99122804 1.0 0.9955947
I-11_ExtinctionInattention 142 0 7 1.0 0.95302016 0.97594506
B-1c_LOCCommands 40 0 1 1.0 0.9756098 0.9876543
B-6_Motor 4 1 5 0.8 0.44444445 0.57142854
I-3_Visual 103 1 5 0.99038464 0.9537037 0.9716981
B-11_ExtinctionInattention 44 1 3 0.9777778 0.9361702 0.9565217
B-Measurement 787 23 13 0.97160494 0.98375 0.97763973
I-10_Dysarthria 76 0 1 1.0 0.987013 0.99346405
I-1b_LOCQuestions 114 0 6 1.0 0.95 0.9743589
Macro-average 3542 83 250 0.9606412 0.8876058 0.9226804
Micro-average 3542 83 250 0.9771035 0.9340717 0.9551032