Key result
Support vector machine classification using 12 clinical features achieved 73.5% accuracy in distinguishing patients with both hypertension and depression from those with hypertension alone.
Why the study?
Can support vector machine learning applied to blood tests and vital signs accurately classify patients with hypertension complicated by depression versus hypertension alone?
Case-Control (n=294)
Can support vector machine learning applied to blood tests and vital signs accurately classify patients with hypertension complicated by depression versus hypertension alone?
Machine learning applied to routine blood tests and vital signs can identify depression in hypertensive patients with moderate accuracy, suggesting potential novel biomarkers.
Answers are grounded in this paper and cite their sources.
Machine learning on routine data may aid depression detection in hypertension; hypothesis-generating and requires prospective validation before clinical use.
Hypertension and depression, as 2 major public health issues, are closely related. For patients having hypertension, in particular, depression is a risk factor for mortality and jeopardizes their wellbeing. The aim of the study is to apply support vector machine (SVM) learning to blood tests and vital signs to classify patients having hypertension complicated by depression and patients having hypertension alone for the identification of novel markers.Data on patients having both hypertension and depression (n = 147) and patients having hypertension alone (n = 147) were obtained from electronic medical records of admissions containing the records on blood tests and vital signs. Using SVM, we distinguished patients having both hypertension and depression from gender- and age-matched patients having hypertension alone.SVM-based classification achieved 73.5% accuracy by 10-fold cross-validation between patients having both hypertension and depression and those having hypertension alone. Twelve features were selected to compose the optimal feature sets, including body temperature (T), glucose (GLU), creatine kinase (CK), albumin (ALB), hydroxybutyrate dehydrogenase (HBDH), blood urea nitrogen (BUN), uric Acid (UA), creatinine (Crea), cholesterol (TC), total protein (TP), pulse (P), and respiration (R).SVM can be used to distinguish patients having both hypertension and depression from those having hypertension alone. A significant association was identified between depression and blood tests and vital signs. This approach can be helpful for clinical diagnosis of depression, but further studies are needed to verify the role of these candidate markers for depression diagnosis.
No takes yet. Share an insight, caveat, or question.
Song et al. (2018) conducted a case-control in Hypertension and depression (n=294). Blood tests and vital signs (12 features) vs. Hypertension alone was evaluated on Classification accuracy. Support vector machine classification using 12 clinical features achieved 73.5% accuracy in distinguishing patients with both hypertension and depression from those with hypertension alone.
Synapse has enriched 10 related papers. Consider them for comparative context: