Tang Lab

07/18/2022

New publication out on / . Work from my postdoc at University of Pennsylvania School of Medicine, finally making it to print. We conducted a pilot study examining technology use in young adults with , clinical high risk, and typically developing peers.

We found no differences in access to technology or use of social media - suggesting a baseline level of feasibility for digital health and social-media based interventions.

However, young people with were less likely to actively post on social media, perhaps reflecting impaired social functioning, and/or presenting an opportunity for intervention.

Check out this paper in JMIR Publications. Congrats to our team - Olivia Franco, Monica Calkins, Sal Giorgi, Lyle Ungar, Raquel Gur, and Christian Kohler.

Background: Digital technology, the internet, and social media are increasingly investigated as promising means for monitoring symptoms and delivering mental health treatment. These apps and interventions have demonstrated preliminary acceptability and feasibility, but previous reports suggest that....

07/12/2022

Our publication is now out in Schizophrenia Research! It can be downloaded for free until 8/30. Here we evaluated the relationships between (emotion processing, mentalizing & attribution bias) and via clinical ratings and computational linguistic features in N=63 people with schizophrenia spectrum disorders.

Computational features included speech graph metrics, modal verbs, first-person pronouns, cosine similarity of adjacent utterances, and sentiment; these were represented by 4 principal components capturing content-rich speech, insular speech, local coherence, and affirmative speech.

Higher clinical ratings for disorganized speech predicted greater impairments in emotion processing and mentalizing. This remained true when accounting for demographic variables, overall psychosis symptoms, and verbal ability. Not true for attribution bias.

Computational features reflecting insular speech also consistently predicted greater impairment in emotion processing. There were notable trends for underproductive speech and decreased content-rich speech predicting mentalizing ability.

Exploratory longitudinal analyses in a small subset of participants (n=17) found that improvements in both emotion processing and mentalizing were predicted by improvements in disorganized speech.

Beyond the interesting relationships, we also interpreted the strong relationships between language and emotion processing + mentalization but not attribution bias to be consistent with an active inference model of discourse.

Many thanks and congrats to our co-authors and collaborators! Yan Cong, Amir Hossein Nikzad, Sunghye Cho, Katrin Hänsel, Aarush Mehta, Aamina Dhar, Sarah Berretta, John M. Kane, and Anil Malhotra, with support from Brain & Behavior Research Foundation (BBRF) and Winterlight Labs.

In this study, we compared three domains of social cognition (emotion processing, mentalizing, and attribution bias) to clinical and computational lan…

06/14/2022

Our publication just came out in Translational Psychiatry! Open access link below. Brief summary: In N=245 participants with , we found that social cognition was strongly related to metabolic functioning.

The relationship between social cognition and metabolic burden was present for cumulative metabolic burden and high waist circumference. The strongest effect was for hemoglobin A1c (HbA1c).

The relationship between HbA1c and social cognition was present for each social cognition domain, both males and females, and was present even when accounting for age, number of hospitalizations, and use of antipsychotics.

The relationship between HbA1c and social cognition was 1) stronger for people with schizophrenia than healthy volunteers, 2) stronger for social cognition than for non-social neurocognition, and 3) partially mediated by functional connectivity between social cog networks.

To our knowledge, this is the 1st report of a robust relationship between social cog and metabolic burden in people with . We cannot conclude causality at this time, but it begs the question - Can we improve social cognition by targeting HbA1c? 5/6

Work was done at , , and . Many thanks to our participants, and my amazing coauthors: Lindsay Oliver, Katrin Hänsel, Pam DeRosse, Majnu John, Ammar Khairullah, Jim Gold, Bob Buchanan, Aristotle Voineskos, and Anil Malhotra.

https://doi.org/10.1038/s41398-022-02002-z

01/04/2022

Our work was featured among the Top 5 Innovations of 2021 at Northwell Health! Way to go, team!

Listen to Northwell experts and some of the brightest minds in health care breaking down the latest news and developments in this podcast.

07/22/2021

Excited that our work has highlighted in a recent article from Mental Health Weekly!

The mental health field continues to rely largely on decades-old tools for assessment of psychotic disorders, in a diagnostic interview process that can prove time-consuming and subjective. Yet advan...

05/14/2021

Excited to share our recent publication in
Schizophrenia! Key takeaway: Natural language processing methods can detect language disturbance in individuals with schizophrenia which cannot be picked up by traditional clinical ratings.

Computerized natural language processing (NLP) allows for objective and sensitive detection of speech disturbance, a hallmark of schizophrenia spectrum disorders (SSD). We explored several methods for characterizing speech changes in SSD (n = 20) compared to healthy control (HC) participants (n....

05/06/2021

Recent article in WSJ highlighting one of our areas of focus - using natural language processing as a biomarker for psychiatric illness.

The information captured by our smartphones, as well as new speech- and facial-recognition technologies, can yield invaluable insights for mental health professionals.

04/20/2021

Proud to (belatedly) share our recent conference paper: we used lexical and acoustic speech features to separately model different aspects of speech disturbance, e.g. poverty, perseveration. Thanks, team!

PDF | There is potential to leverage lexical and acoustic features as predictors of clinical ratings used to measure thought disorder and negative... | Find, read and cite all the research you need on ResearchGate

03/01/2021

Awesome, innovative work from our collaborator Michael Birnbaum:

For the 20 percent of people with a mental illness, early identification of the condition is key to getting the best treatment. But people often suffer symptoms for months, even years, without receiving clinical attention. Part of the problem is that psychiatrists have few tools to identify mental i...

02/27/2021

Some awesome recent papers from our collaborators at the Linguistic Data Consortium / Penn looking at speech and language features of normal aging as well as frontotemporal degeneration.

https://pubs.asha.org/doi/10.1044/2020_JSLHR-19-00384

https://www.medrxiv.org/content/10.1101/2020.09.10.20192054v2

We implemented an automated analysis of lexical aspects of semi-structured speech produced by healthy elderly controls (n=37) and three patient groups with frontotemporal degeneration (FTD): behavioral variant FTD (n=74), semantic variant primary progressive aphasia (svPPA, n=42), and nonfluent/agra...

02/27/2021

Check it out! Spoke about our work in this blog-interview by Michael Birnbaum.

Q&A with Sunny Tang, M.D.