Text Classification For Mental Disorder

Description

The PHS-BERT Mental Disorder Classifier Model classifies text based on the following mental disorders. (Schizophrenia, Depression, Bipolar disorder, Anxiety disorder).

The Mental Disorder Classifier Model is a specialized text classification system, engineered to accurately identify and categorize textual mentions of four prominent respiratory diseases: Schizophrenia, Depression, Bipolar disorder, Anxiety disorder. More detailed information about classes as follows:

Schizophrenia: This category is designated for text mentions that correspond to Schizophrenia, a mental disorder characterized by thought disorders, hallucinations, and often involving delusional beliefs and cognitive challenges. Example: “My brother’s hallucinations, a symptom of his schizophrenia, make daily tasks challenging for him.”

Depression: Textual content which implicates Depression, a pervasive mental health disorder marked by persistent feelings of sadness, loss of interest in activities, and potential physical health alterations, is classified here. Example: “The heavy weight of depression often leaves me struggling to find joy in activities I once loved.”

Bipolar Disorder: Entries here incorporate text alluding to Bipolar Disorder, a mental health condition distinguished by stark fluctuations in mood, transitioning between manic highs and depressive lows. Example: “Bipolar disorder has my mood swinging from exhilarating happiness to debilitating lows without warning.”

Anxiety Disorder: Textual mentions signifying Anxiety Disorder, which involves persistent, excessive worry, nervousness, and physiological effects, are classified under this label. Example: “My anxiety disorder frequently sets my heart racing and mind spiralling over mere hypothetical scenarios.”

Predicted Entities

Anxiety Disorder, Bipolar disorder, No, Schizophrenia, Depression, Other/Unknown

Live Demo Open in Colab Download Copy S3 URI

How to use

# 1. Document assembler
document_assembler = DocumentAssembler() \
    .setInputCol("text") \
    .setOutputCol("document")

tokenizer = Tokenizer()\
    .setInputCols(["document"])\
    .setOutputCol("token")

word_embeddings = WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")\
    .setInputCols(["document", "token"])\
    .setOutputCol("word_embeddings")

sentence_embeddings = SentenceEmbeddings()\
    .setInputCols(["document", "word_embeddings"])\
    .setOutputCol("sentence_embeddings")\
    .setPoolingStrategy("AVERAGE")

# 3. Mental Disorder MultiClassifier (Using the provided model name)
multilabel_classifier = MultiClassifierDLModel.pretrained("multiclassifierdl_mental_disorder", "en", "clinical/models") \
    .setInputCols(["sentence_embeddings"]) \
    .setOutputCol("prediction")\
    .setThreshold(0.999)

# Creating the pipeline
clf_pipeline = Pipeline(stages=[
    document_assembler,
    tokenizer,
    word_embeddings,
    sentence_embeddings,
    multilabel_classifier
])

# Sample text list
text_list = [
    """The patient verbalizes feelings of worthlessness and recurrent thoughts of death, necessitating immediate intervention for severe depression.""",
    """Monitoring the patient for potential side effects of lithium therapy, initiated to manage the mood fluctuations of bipolar disorder.""",
    """Commenced CBT to equip the patient with coping strategies for persistent irrational fears and nervousness stemming from their anxiety disorder.""",
    """Michael Anderson takes medications for schizophrenia, hypertension, hyperlipidemia, and migraines.""",
    """With a clinical presentation of persistent wheezing and respiratory discomfort, the patient has been conclusively diagnosed with Asthma, necessitating immediate commencement of anti-inflammatory medications."""
]

# Creating a DataFrame
df = spark.createDataFrame(text_list, StringType()).toDF("text")


result = clf_pipeline.fit(df).transform(df)

val documentAssembler = new DocumentAssembler()
    .setInputCol("text")
    .setOutputCol("document")

val tokenizer = new Tokenizer()
    .setInputCols(Array("document"))
    .setOutputCol("token")

val wordEmbeddings = WordEmbeddingsModel.pretrained("embeddings_clinical", "en", "clinical/models")
    .setInputCols(Array("document", "token"))
    .setOutputCol("embeddings")

val sentence_embeddings = new SentenceEmbeddings()\
    .setInputCols(Array("document", "embeddings")) \
    .setOutputCol("sentence_embeddings") \
    .setPoolingStrategy("AVERAGE")

val multiclassifierdl = MultiClassifierDLModel.pretrained("multiclassifierdl_mental_disorder", "en", "clinical/models")\
    .setInputCols("sentence_embeddings")\
    .setOutputCol("predicted_class")

val clf_pipeline = new Pipeline().setStages(Array(
    documentAssembler,
    tokenizer,
    wordEmbeddings,
    sentence_embeddings,
    multiclassifierdl
))

val data = Seq(Array(
    """The patient verbalizes feelings of worthlessness and recurrent thoughts of death, necessitating immediate intervention for severe depression.""",
    """Monitoring the patient for potential side effects of lithium therapy, initiated to manage the mood fluctuations of bipolar disorder.""",
    """Commenced CBT to equip the patient with coping strategies for persistent irrational fears and nervousness stemming from their anxiety disorder.""",
    """Michael Anderson takes medications for schizophrenia, hypertension, hyperlipidemia, and migraines.""",
    """With a clinical presentation of persistent wheezing and respiratory discomfort, the patient has been conclusively diagnosed with Asthma, necessitating immediate commencement of anti-inflammatory medications.""",
)).toDS.toDF("text")

val result = clf_pipeline.fit(data).transform(data)

Results

+------------------------------------------------------------------------------------------------------------------------------------------------------+------------------+
|                                                                                                                                                  text|            result|
+------------------------------------------------------------------------------------------------------------------------------------------------------+------------------+
|         The patient verbalizes feelings of worthlessness and recurrent thoughts of death, necessitating immediate intervention for severe depression.|      [Depression]|
|                  Monitoring the patient for potential side effects of lithium therapy, initiated to manage the mood fluctuations of bipolar disorder.|[Bipolar disorder]|
|       Commenced CBT to equip the patient with coping strategies for persistent irrational fears and nervousness stemming from their anxiety disorder.|[Anxiety disorder]|
|                                                    Michael Anderson takes medications for schizophrenia, hypertension, hyperlipidemia, and migraines.|   [Schizophrenia]|
|With a clinical presentation of persistent wheezing and respiratory discomfort, the patient has been conclusively diagnosed with Asthma, necessitat...|              [No]|
+------------------------------------------------------------------------------------------------------------------------------------------------------+------------------+

Model Information

References

Trained with the in-house dataset

Sample text from the training dataset

Schizophrenia: The patient was diagnosed with schizophrenia at the age of 22 after a series of intense psychotic episodes. The patient often describes hearing voices that others don’t hear and harbors a persistent belief in unfounded conspiracies, indicative of schizophrenia. The patient has been prescribed second-generation antipsychotics and will be undergoing cognitive behavioral therapy to help manage the symptoms of schizophrenia. Depression: Mrs. Thompson first sought treatment for depression during her late 20s, shortly after a series of personal losses. She often expresses feelings of hopelessness, has a persistent low mood, and lacks interest in activities she once enjoyed, classic signs of depression. Mrs. Thompson will continue her selective serotonin reuptake inhibitor (SSRI) medication and engage in weekly psychotherapy sessions to address her depressive symptoms. Bipolar Disorder: Mr. Gray was diagnosed with bipolar disorder at the age of 27 following an episode of mania that was preceded by a deep depressive phase. He cycles between periods of elevated mood, increased energy, and impulsiveness, and phases of intense sadness and lethargy, consistent with bipolar disorder. The management approach for Mr. Gray includes mood stabilizers like lithium, coupled with cognitive-behavioral therapy to monitor and address the mood swings of bipolar disorder. Anxiety Disorder: The patient has been experiencing generalized anxiety disorder symptoms for the past six years, often linked to work-related stress. The patient describes persistent feelings of restlessness, muscle tension, and overwhelming worry about everyday events and activities, common to those with an anxiety disorder. A combination of cognitive behavioral therapy and an anxiolytic medication has been advised to alleviate the distressing symptoms of the anxiety disorder.

Benchmarking

label              tp   fp      fn    prec          rec         f1
Other/Unknown      39   10      22    0.79591835   0.6393443    0.7090909
Depression         120  18      19    0.8695652    0.8633093    0.866426
No                 28   21      27    0.93203884   0.9142857    0.9230769
Anxiety_disorder   136  15      18    0.90066224   0.8831169    0.8918033
Bipolar_disorder   101  5       10    0.9528302    0.9099099    0.9308756
Schizophrenia      66   4       23    0.94285715   0.74157304   0.8301887
Macro-average      750  73     119    0.89897865   0.8252565    0.8605415
Micro-average      750  73     119    0.9113001    0.863061     0.886524