Automated Affect Detection in Deep Brain Stimulation for Obsessive-Compulsive Disorder

Abstract

Automated measurement of affective behavior in psychopathology has been limited primarily to screening and diagnosis. While useful, clinicians more often are concerned with whether patients are improving in response to treatment. Are symptoms abating, is affect becoming more positive, are unanticipated side effects emerging? When treatment includes neural implants, need for objective, repeatable biometrics tied to neurophysiology becomes especially pressing. We used automated face analysis to assess treatment response to deep brain stimulation (DBS) in two patients with intractable obsessive-compulsive disorder (OCD). One was assessed intraoperatively following implantation and activation of the DBS device. The other was assessed three months post-implantation. Both were assessed during DBS on and o conditions. Positive and negative valence were quantified using a CNN trained on normative data of 160 non-OCD participants. Thus, a secondary goal was domain transfer of the classifiers. In both contexts, DBS-on resulted in marked positive affect. In response to DBS-off, affect flattened in both contexts and alternated with increased negative affect in the outpatient setting. Mean AUC for domain transfer was 0.87. These findings suggest that parametric variation of DBS is strongly related to affective behavior and may introduce vulnerability for negative affect in the event that DBS is discontinued.

D-PAttNet: Dynamic Patch-Attentive Deep Network for Action Unit Detection

“…To be consistent with previous work, we selected patch sequences of length T = 20 within that range. The CNN architecture used in this study has been shown to be successful in previous studies (Cohn et al, 2018;Onal Ertugrul et al, 2019a,c). Two differences from previous work may be noted.…”

Crossing Domains for AU Coding: Perspectives, Approaches, and Measures

“…We used CNN architecture given in Fig. 1 in our experiments since it has been shown to perform well and provide meaningful results in a cross-domain setting [54]. One of the limitations of our work is, we do not know whether a deeper CNN would perform better.…”

“…To compare AU-specific within-and cross-domain transfer, we use both deep and shallow approaches in two databases that represent different domains. The deep approach is a CNN architecture [54]; the shallow approach is a support vector machine (SVM) with handcrafted features. One database used for training is an expanded version of BP4D+ [6] (which we refer to as EB+).…”

Deep Brain Stimulation for Intractable Obsessive-Compulsive Disorder: Progress and Opportunities

“…A study using single-photon emission computed tomography found that nucleus accumbens DBS induced striatal dopamine release in patients with OCD (17). Using AFAR, we have been able to demonstrate that ventral striatum DBS can turn up the gain on positive valence affect (8). For the most part, patients with OCD are driven by harm avoidance (18,19), and their compulsions are aimed at preventing negative outcomes and are not inherently pleasurable, thus differentiating compulsions from behavioral addictions.…”