Hooman Rokham1, Godfrey Pearlson2, Anees Abrol3, Haleh Falakshahi4, Sergey Plis3, Vince D Calhoun5. 1. Department of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, Georgia; Tri-institutional Center of Translational Research in Neuroimaging and Data Science (TReNDS), Georgia State University, Georgia Institute of Technology, and Emory University, Atlanta, Georgia. Electronic address: hrokham@gatech.edu. 2. Department of Psychiatry, Yale University, New Haven, Connecticut; Department of Neuroscience, Yale University, New Haven, Connecticut; Olin Neuropsychiatry Research Center, Hartford Hospital, Hartford, Connecticut. 3. Tri-institutional Center of Translational Research in Neuroimaging and Data Science (TReNDS), Georgia State University, Georgia Institute of Technology, and Emory University, Atlanta, Georgia; Department of Computer Science, Georgia State University, Atlanta, Georgia. 4. Department of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, Georgia; Tri-institutional Center of Translational Research in Neuroimaging and Data Science (TReNDS), Georgia State University, Georgia Institute of Technology, and Emory University, Atlanta, Georgia. 5. Department of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, Georgia; Tri-institutional Center of Translational Research in Neuroimaging and Data Science (TReNDS), Georgia State University, Georgia Institute of Technology, and Emory University, Atlanta, Georgia; Department of Computer Science, Georgia State University, Atlanta, Georgia; Department of Psychology, Georgia State University, Atlanta, Georgia; Department of Psychiatry, Yale University, New Haven, Connecticut.
Abstract
BACKGROUND: Mental health diagnostic approaches are seeking to identify biological markers to work alongside advanced machine learning approaches. It is difficult to identify a biological marker of disease when the traditional diagnostic labels themselves are not necessarily valid. METHODS: We worked with T1 structural magnetic resonance imaging data collected from 1493 individuals comprising healthy control subjects, patients with psychosis, and their unaffected first-degree relatives. Specifically, the dataset included 176 bipolar disorder probands, 134 schizoaffective disorder probands, 240 schizophrenia probands, 362 control subjects, and 581 patient relatives. We assumed that there might be noise in the diagnostic labeling process. We detected label noise by classifying the data multiple times using a support vector machine classifier, and then we flagged those individuals in which all classifiers unanimously mislabeled those subjects. Next, we assigned a new diagnostic label to these individuals, based on the biological data (magnetic resonance imaging), using an iterative data cleansing approach. RESULTS: Simulation results showed that our method was highly accurate in identifying label noise. Both diagnostic and biotype categories showed about 65% and 63% of noisy labels, respectively, with the largest amount of relabeling occurring between the healthy control subjects and individuals with bipolar disorder and schizophrenia as well as in unaffected close relatives. The extraction of imaging features highlighted regional brain changes associated with each group. CONCLUSIONS: This approach represents an initial step toward developing strategies that need not assume that existing mental health diagnostic categories are always valid but rather allows us to leverage this information while also acknowledging that there are misassignments.
BACKGROUND: Mental health diagnostic approaches are seeking to identify biological markers to work alongside advanced machine learning approaches. It is difficult to identify a biological marker of disease when the traditional diagnostic labels themselves are not necessarily valid. METHODS: We worked with T1 structural magnetic resonance imaging data collected from 1493 individuals comprising healthy control subjects, patients with psychosis, and their unaffected first-degree relatives. Specifically, the dataset included 176 bipolar disorder probands, 134 schizoaffective disorder probands, 240 schizophrenia probands, 362 control subjects, and 581 patient relatives. We assumed that there might be noise in the diagnostic labeling process. We detected label noise by classifying the data multiple times using a support vector machine classifier, and then we flagged those individuals in which all classifiers unanimously mislabeled those subjects. Next, we assigned a new diagnostic label to these individuals, based on the biological data (magnetic resonance imaging), using an iterative data cleansing approach. RESULTS: Simulation results showed that our method was highly accurate in identifying label noise. Both diagnostic and biotype categories showed about 65% and 63% of noisy labels, respectively, with the largest amount of relabeling occurring between the healthy control subjects and individuals with bipolar disorder and schizophrenia as well as in unaffected close relatives. The extraction of imaging features highlighted regional brain changes associated with each group. CONCLUSIONS: This approach represents an initial step toward developing strategies that need not assume that existing mental health diagnostic categories are always valid but rather allows us to leverage this information while also acknowledging that there are misassignments.
Authors: Brett A Clementz; John A Sweeney; Jordan P Hamm; Elena I Ivleva; Lauren E Ethridge; Godfrey D Pearlson; Matcheri S Keshavan; Carol A Tamminga Journal: Am J Psychiatry Date: 2015-12-07 Impact factor: 18.112
Authors: Elena I Ivleva; Brett A Clementz; Anthony M Dutcher; Sara J M Arnold; Haekyung Jeon-Slaughter; Sina Aslan; Bradley Witte; Gaurav Poudyal; Hanzhang Lu; Shashwath A Meda; Godfrey D Pearlson; John A Sweeney; Matcheri S Keshavan; Carol A Tamminga Journal: Biol Psychiatry Date: 2016-08-31 Impact factor: 13.382
Authors: Elena I Ivleva; Anup S Bidesi; Matcheri S Keshavan; Godfrey D Pearlson; Shashwath A Meda; Darko Dodig; Amanda F Moates; Hanzhang Lu; Alan N Francis; Neeraj Tandon; David J Schretlen; John A Sweeney; Brett A Clementz; Carol A Tamminga Journal: Am J Psychiatry Date: 2013-11 Impact factor: 18.112
Authors: Zheng Wang; Shashwath A Meda; Matcheri S Keshavan; Carol A Tamminga; John A Sweeney; Brett A Clementz; David J Schretlen; Vince D Calhoun; Su Lui; Godfrey D Pearlson Journal: Front Psychiatry Date: 2015-12-21 Impact factor: 4.157
Authors: Haleh Falakshahi; Hooman Rokham; Zening Fu; Armin Iraji; Daniel H Mathalon; Judith M Ford; Bryon A Mueller; Adrian Preda; Theo G M van Erp; Jessica A Turner; Sergey Plis; Vince D Calhoun Journal: Netw Neurosci Date: 2022-07-01