Download This Paper
Preprints with The Lancet is a collaboration between The Lancet Group of journals and SSRN to facilitate the open sharing of preprints for early engagement, community comment, and collaboration. Preprints available here are not Lancet publications or necessarily under review with a Lancet journal. These preprints are early-stage research papers that have not been peer-reviewed. The usual SSRN checks and a Lancet-specific check for appropriateness and transparency have been applied. The findings should not be used for clinical or public health decision-making or presented without highlighting these facts. For more information, please see the FAQs.
Machine Learning for Exploring and Evaluating Clinical Diagnosis Model in Primary Biliary Cholangitis
30 Pages Posted: 10 Aug 2020
More...Abstract
Background: To explore new indicators and improve predictive models for early diagnosis of primary biliary cholangitis (PBC) based on large samples of clinical data in the real world.
Methods: We used different machine learning technologies to explore and analyze the clinical diagnosis models of PBC based on 1355 large sample data.
Findings: The explored Supported Vector Machine (SVM) and the validated Generalized Linear Machine (GLM) models with anti-nuclear antibody (ANA), alkaline phosphatase (ALP) and antimitochondrial antibodies (AMA)-positive showed to be more excellent than the baseline model: In SVM model, negative predictive rate was 8% higher than the baseline, and an 13% increase on sensitivity; in GLM model, the chi-square test for coefficients were all significant, it indicated that variables ANA and ALP had a significant impact on the target variable. ALP value less than 127.0 in males would predict preclinical PBC with the confidence 0.99, and the one greater than 127.0 clinical PBC with the confidence 0.88, and for females, ALP value less than 131.5 would predict preclinical PBC with the confidence 0.90, and the one greater than 131.5 clinical PBC with the confidence 0.96 respectively.
Interpretation: Diverse supervised and unsupervised machine learning algorithms can be effectively applied to explore new clinic knowledge hiding in real-world clinical data to help improving clinic diagnosis, and ANA, found by machine learning, is another most important indicator as well as AMA and ALP on diagnosing clinical PBC, especially in male patients with PBC.
Funding: Financial support: This work was supported by the National S&T Major Projects for Infectious Diseases Control (2017ZX10302201-004-001 and 2017ZX10203202-003-003).
Declaration of Interests: The authors declare no competing interest.
Ethics Approval Statement: The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Peking University People’s Hospital (2019PHB279-01).
Keywords: Machine learning; Primary biliary cholangitis; Antinuclear antibody; Diagnosis
Suggested Citation: Suggested Citation