71
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: not found
      • Article: not found

      Machine Learning in Medicine

      1
      Circulation
      Ovid Technologies (Wolters Kluwer Health)

      Read this article at

      ScienceOpenPublisherPMC
      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Spurred by advances in processing power, memory, storage, and an unprecedented wealth of data, computers are being asked to tackle increasingly complex learning tasks, often with astonishing success. Computers have now mastered a popular variant of poker, learned the laws of physics from experimental data, and become experts in video games - tasks that would have been deemed impossible not too long ago. In parallel, the number of companies centered on applying complex data analysis to varying industries has exploded, and it is thus unsurprising that some analytic companies are turning attention to problems in health care. The purpose of this review is to explore what problems in medicine might benefit from such learning approaches and use examples from the literature to introduce basic concepts in machine learning. It is important to note that seemingly large enough medical data sets and adequate learning algorithms have been available for many decades, and yet, although there are thousands of papers applying machine learning algorithms to medical data, very few have contributed meaningfully to clinical care. This lack of impact stands in stark contrast to the enormous relevance of machine learning to many other industries. Thus, part of my effort will be to identify what obstacles there may be to changing the practice of medicine through statistical learning approaches, and discuss how these might be overcome.

          Related collections

          Most cited references14

          • Record: found
          • Abstract: found
          • Article: not found

          T-helper type 2-driven inflammation defines major subphenotypes of asthma.

          T-helper type 2 (Th2) inflammation, mediated by IL-4, IL-5, and IL-13, is considered the central molecular mechanism underlying asthma, and Th2 cytokines are emerging therapeutic targets. However, clinical studies increasingly suggest that asthma is heterogeneous. To determine whether this clinical heterogeneity reflects heterogeneity in underlying molecular mechanisms related to Th2 inflammation. Using microarray and polymerase chain reaction analyses of airway epithelial brushings from 42 patients with mild-to-moderate asthma and 28 healthy control subjects, we classified subjects with asthma based on high or low expression of IL-13-inducible genes. We then validated this classification and investigated its clinical implications through analyses of cytokine expression in bronchial biopsies, markers of inflammation and remodeling, responsiveness to inhaled corticosteroids, and reproducibility on repeat examination. Gene expression analyses identified two evenly sized and distinct subgroups, "Th2-high" and "Th2-low" asthma (the latter indistinguishable from control subjects). These subgroups differed significantly in expression of IL-5 and IL-13 in bronchial biopsies and in airway hyperresponsiveness, serum IgE, blood and airway eosinophilia, subepithelial fibrosis, and airway mucin gene expression (all P < 0.03). The lung function improvements expected with inhaled corticosteroids were restricted to Th2-high asthma, and Th2 markers were reproducible on repeat evaluation. Asthma can be divided into at least two distinct molecular phenotypes defined by degree of Th2 inflammation. Th2 cytokines are likely to be a relevant therapeutic target in only a subset of patients with asthma. Furthermore, current models do not adequately explain non-Th2-driven asthma, which represents a significant proportion of patients and responds poorly to current therapies.
            Bookmark
            • Record: found
            • Abstract: not found
            • Article: not found

            Sparse coding with an overcomplete basis set: A strategy employed by V1?

              Bookmark
              • Record: found
              • Abstract: found
              • Article: not found

              Lebrikizumab treatment in adults with asthma.

              Many patients with asthma have uncontrolled disease despite treatment with inhaled glucocorticoids. One potential cause of the variability in response to treatment is heterogeneity in the role of interleukin-13 expression in the clinical asthma phenotype. We hypothesized that anti-interleukin-13 therapy would benefit patients with asthma who had a pretreatment profile consistent with interleukin-13 activity. We conducted a randomized, double-blind, placebo-controlled study of lebrikizumab, a monoclonal antibody to interleukin-13, in 219 adults who had asthma that was inadequately controlled despite inhaled glucocorticoid therapy. The primary efficacy outcome was the relative change in prebronchodilator forced expiratory volume in 1 second (FEV(1)) from baseline to week 12. Among the secondary outcomes was the rate of asthma exacerbations through 24 weeks. Patient subgroups were prespecified according to baseline type 2 helper T-cell (Th2) status (assessed on the basis of total IgE level and blood eosinophil count) and serum periostin level. At baseline, patients had a mean FEV(1) that was 65% of the predicted value and were taking a mean dose of inhaled glucocorticoids of 580 μg per day; 80% were also taking a long-acting beta-agonist. At week 12, the mean increase in FEV(1) was 5.5 percentage points higher in the lebrikizumab group than in the placebo group (P = 0.02). Among patients in the high-periostin subgroup, the increase from baseline FEV(1) was 8.2 percentage points higher in the lebrikizumab group than in the placebo group (P = 0.03). Among patients in the low-periostin subgroup, the increase from baseline FEV(1) was 1.6 percentage points higher in the lebrikizumab group than in the placebo group (P = 0.61). Musculoskeletal side effects were more common with lebrikizumab than with placebo (13.2% vs. 5.4%, P = 0.045). Lebrikizumab treatment was associated with improved lung function. Patients with high pretreatment levels of serum periostin had greater improvement in lung function with lebrikizumab than did patients with low periostin levels. (Funded by Genentech; ClinicalTrials.gov number, NCT00930163 .).
                Bookmark

                Author and article information

                Journal
                Circulation
                Circulation
                Ovid Technologies (Wolters Kluwer Health)
                0009-7322
                1524-4539
                November 17 2015
                November 17 2015
                : 132
                : 20
                : 1920-1930
                Affiliations
                [1 ]From Cardiovascular Research Institute, Department of Medicine and Institute for Human Genetics, University of California, San Francisco, and California Institute for Quantitative Biosciences, San Francisco.
                Article
                10.1161/CIRCULATIONAHA.115.001593
                5831252
                26572668
                dbedb3a2-62ee-4584-92c1-49625999f526
                © 2015
                History

                Comments

                Comment on this article