1
|
Marmar CR, Brown AD, Qian M, Laska E, Siegel C, Li M, Abu-Amara D, Tsiartas A, Richey C, Smith J, Knoth B, Vergyri D. Speech-based markers for posttraumatic stress disorder in US veterans. Depress Anxiety 2019; 36:607-616. [PMID: 31006959 PMCID: PMC6602854 DOI: 10.1002/da.22890] [Citation(s) in RCA: 48] [Impact Index Per Article: 9.6] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 12/04/2018] [Revised: 02/14/2019] [Accepted: 03/08/2019] [Indexed: 01/01/2023] Open
Abstract
BACKGROUND The diagnosis of posttraumatic stress disorder (PTSD) is usually based on clinical interviews or self-report measures. Both approaches are subject to under- and over-reporting of symptoms. An objective test is lacking. We have developed a classifier of PTSD based on objective speech-marker features that discriminate PTSD cases from controls. METHODS Speech samples were obtained from warzone-exposed veterans, 52 cases with PTSD and 77 controls, assessed with the Clinician-Administered PTSD Scale. Individuals with major depressive disorder (MDD) were excluded. Audio recordings of clinical interviews were used to obtain 40,526 speech features which were input to a random forest (RF) algorithm. RESULTS The selected RF used 18 speech features and the receiver operating characteristic curve had an area under the curve (AUC) of 0.954. At a probability of PTSD cut point of 0.423, Youden's index was 0.787, and overall correct classification rate was 89.1%. The probability of PTSD was higher for markers that indicated slower, more monotonous speech, less change in tonality, and less activation. Depression symptoms, alcohol use disorder, and TBI did not meet statistical tests to be considered confounders. CONCLUSIONS This study demonstrates that a speech-based algorithm can objectively differentiate PTSD cases from controls. The RF classifier had a high AUC. Further validation in an independent sample and appraisal of the classifier to identify those with MDD only compared with those with PTSD comorbid with MDD is required.
Collapse
Affiliation(s)
- Charles R. Marmar
- Department of Psychiatry, New York University School of Medicine, New York, New York; Steven and Alexandra Cohen Veterans Center for the Study of Post-Traumatic Stress and Traumatic Brain Injury, New York, New York,Corresponding Author: Charles R. Marmar, MD - Department of Psychiatry, New York University School of Medicine, 1 Park Avenue, New York, NY 10016,
| | - Adam D. Brown
- Department of Psychiatry, New York University School of Medicine, New York, New York; Steven and Alexandra Cohen Veterans Center for the Study of Post-Traumatic Stress and Traumatic Brain Injury, New York, New York,Department of Psychology, New School for Social Research, New York, New York
| | - Meng Qian
- Department of Psychiatry, New York University School of Medicine, New York, New York; Steven and Alexandra Cohen Veterans Center for the Study of Post-Traumatic Stress and Traumatic Brain Injury, New York, New York
| | - Eugene Laska
- Department of Psychiatry, New York University School of Medicine, New York, New York; Steven and Alexandra Cohen Veterans Center for the Study of Post-Traumatic Stress and Traumatic Brain Injury, New York, New York
| | - Carole Siegel
- Department of Psychiatry, New York University School of Medicine, New York, New York; Steven and Alexandra Cohen Veterans Center for the Study of Post-Traumatic Stress and Traumatic Brain Injury, New York, New York
| | - Meng Li
- Department of Psychiatry, New York University School of Medicine, New York, New York; Steven and Alexandra Cohen Veterans Center for the Study of Post-Traumatic Stress and Traumatic Brain Injury, New York, New York
| | - Duna Abu-Amara
- Department of Psychiatry, New York University School of Medicine, New York, New York; Steven and Alexandra Cohen Veterans Center for the Study of Post-Traumatic Stress and Traumatic Brain Injury, New York, New York
| | | | | | | | | | | |
Collapse
|
2
|
Peintner B, Jarrold W, Vergyriy D, Richey C, Tempini MLG, Ogar J. Learning diagnostic models using speech and language measures. Annu Int Conf IEEE Eng Med Biol Soc 2009; 2008:4648-51. [PMID: 19163752 DOI: 10.1109/iembs.2008.4650249] [Citation(s) in RCA: 16] [Impact Index Per Article: 1.1] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/08/2022]
Abstract
We describe results that show the effectiveness of machine learning in the automatic diagnosis of certain neurodegenerative diseases, several of which alter speech and language production. We analyzed audio from 9 control subjects and 30 patients diagnosed with one of three subtypes of Frontotemporal Lobar Degeneration. From this data, we extracted features of the audio signal and the words the patient used, which were obtained using our automated transcription technologies. We then automatically learned models that predict the diagnosis of the patient using these features. Our results show that learned models over these features predict diagnosis with accuracy significantly better than random. Future studies using higher quality recordings will likely improve these results.
Collapse
|
3
|
Bendele A, Seely J, Richey C, Sennello G, Shopp G. Short communication: renal tubular vacuolation in animals treated with polyethylene-glycol-conjugated proteins. Toxicol Sci 1998; 42:152-7. [PMID: 9579027 DOI: 10.1006/toxs.1997.2396] [Citation(s) in RCA: 63] [Impact Index Per Article: 2.4] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [MESH Headings] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/22/2022] Open
Abstract
During toxicologic evaluation of a dimeric PEG-linked protein, tumor necrosis factor binding protein (TNF-bp), vacuolation of renal cortical tubular epithelium was seen in male and female Sprague-Dawley rats (200-300 g) given i.v. doses of 40, 20, or 10 mg/kg every other day for 3 months. Tubular lesions in rats treated with 20 or 40 mg/kg for 3 months were only partially reversible after a 2-month recovery period. Despite the presence of marked vacuolation, there were no changes in BUN, creatinine, urinalysis parameters, urinary NAG, urinary B2-microglobulin, or fractional sodium excretion. Single i.v. doses > or = 20 mg/kg TNF-bp caused similar but milder changes. However, equivalent doses of PEG alone or the non-PEG-linked TNF-bp did not cause light microscopic evidence of vacuolation. Treatment of rats with another PEG-linked protein of similar molecular weight resulted in similar changes. Immunostaining for TNF-bp revealed positivity in the apical cytoplasm of renal tubular epithelium within 1 h of i.v. dosing. Immunostaining of kidneys from chronically dosed rats indicated that protein was present in some vacuoles as long as dosing continued; however, kidneys from animals on a reversibility study had vacuoles but no immunostaining for TNF-bp. These results, along with a study that showed more severe lesions with PEG-linked proteins of lower molecular weight and minimal if any lesions with PEG-linked proteins > 70 kDa, suggest that TNF-bp is filtered through the glomerulus and that the protein with attached PEG is reabsorbed by the proximal tubules. Vacuolation may be a result of fluid distension of lysosomes due to the hygroscopic nature of PEG. These studies demonstrated that PEG-linked proteins have the capacity to induce renal tubular vacuolation at high doses. However, the change was not associated with alteration of clinical pathology or functional markers.
Collapse
Affiliation(s)
- A Bendele
- Amgen Inc., Boulder, Colorado 80301, USA
| | | | | | | | | |
Collapse
|