Related Experiment Video
Updated: Jun 11, 2025
16:13
Modeling Neural Immune Signaling of Episodic and Chronic Migraine Using Spreading Depression In Vitro
Published on: June 13, 2011
20.0K
Application Value of a Machine Learning Model in Predicting Mild Depression Associated with Migraine without Aura
Sheng-Wei Cui1, Pei Pei1, Wen-Ming Yang1,2,3
1Department of Neurology, The First Affiliated Hospital of Anhui University of Traditional Chinese Medicine, Hefei, Anhui, China.
British Journal of Hospital Medicine (London, England : 2005)
|September 30, 2024
Summary
A machine learning model effectively predicts mild depression in migraine without aura patients. Key factors include gender, disease duration, attack frequency, headache duration, and specific questionnaire scores (MIDAS, HIT-6).
Area of Science:
- Neurology
- Artificial Intelligence
- Psychiatry
Background:
- Migraine without aura (MwoA) frequently co-occurs with mild depression, impacting patient quality of life.
- Accurate prediction of comorbid depression in MwoA patients is crucial for timely intervention.
Purpose of the Study:
- To evaluate the efficacy of a machine learning model in predicting mild depression among patients with MwoA.
- To identify key clinical factors associated with mild depression in MwoA.
Main Methods:
- A cohort of 178 MwoA patients was retrospectively analyzed.
- Patients were divided into modeling (n=140) and validation (n=38) groups.
- Machine learning and logistic regression analyses were employed to identify predictors and build a predictive model.
Main Results:
- Gender, disease duration, attack frequency, headache duration, Migraine Disability Assessment Questionnaire (MIDAS), and Headache Impact Test-6 (HIT-6) scores were significant predictors (p < 0.05).
- The prediction model demonstrated high accuracy in both the modeling group (AUC=0.982) and validation group (AUC=0.901).
- High sensitivity and specificity were observed in both groups, indicating robust predictive performance.
Conclusions:
- Clinical factors such as gender, disease duration, attack frequency, headache duration, MIDAS, and HIT-6 scores are independent predictors of mild depression in MwoA.
- The developed machine learning model shows strong potential for predicting mild depression in MwoA patients.
