CCN 2026 | Computational Psychiatry & Development (CT)

CCN 2026 | Computational Psychiatry & Development (CT)

by Cognitive Computational Neuroscience

Life & Natural Systems Medicine & Health MMedicine and NursingMKLPsychiatry
🎙 Cognitive Computational Neuroscience 👥 4K 📅 August 12, 2026 ⏱ 61 min 👁 70 📄 original study 🧭 2026-08-15
Available in: English (current) Français

Keywords

excitation-inhibition imbalancelatent cause inferencevisual representationsEEGcomputational modeling

Summary

This session from the Cognitive Computational Neuroscience (CCN) 2026 conference features five contributed talks on computational psychiatry and development. The first talk by Kailin Zhu presents a study using virtual brains informed by white-matter microstructure to localize excitation-inhibition (EI) imbalance in schizophrenia. The model, based on a reduced Wong-Wang model, uses structural connectivity metrics (ADC, GFA) to improve functional connectivity predictions and identifies regions like posterior cingulate and paracentral as key in EI differences. Classification using model parameters achieves around 0.7 accuracy, suggesting clinical relevance. The second talk by Camilla van Geen investigates distinct computational pathways to persistent fear across development, using a latent cause inference model. They find that children show more generalization and less lasting memory of latent causes, while young adults show selective maintenance of aversive causes. The third talk by Chun-Hui Li examines the developmental trajectory of temporal dynamics in human visual representations using EEG data from 206 participants across seven age cohorts. They find that visual representations become stronger, peak earlier, and persist longer with age, with category-specific developmental trajectories. Tensor decomposition reveals two latent patterns: an early category-separating pattern and a late scene-dominant pattern. The fourth talk by Thibaut Chataing presents a self-supervised model of social primitives in human dyads, though details are not provided in the transcript. The fifth talk by Domenic Bersch and Elizabeth Jiwon Im are also mentioned but not detailed. Overall, the session highlights the use of computational models to understand psychiatric and developmental phenomena, with a focus on mechanistic insights and potential clinical applications.

255 words

Critical Evaluation

Value of the Information & Strength of the Argument

The talks present valuable computational approaches to understanding psychiatric and developmental phenomena. Zhu’s work demonstrates the importance of white-matter microstructure metrics in improving model fit and identifying clinically relevant biomarkers. Van Geen’s latent cause inference framework provides a formal theory for persistent fear, with clear behavioral and computational predictions. Li’s large-scale EEG study offers a comprehensive view of visual development, revealing distinct trajectories. The argumentation is generally solid, with quantitative results and clear hypotheses. However, some talks are brief and lack detailed statistical reporting, and the lack of peer-reviewed publication details limits the ability to fully assess the robustness of the findings.

Scientific Rigor, Source Quality, Title Accuracy

The session is part of the CCN 2026 conference, and the talks present original research. The sources cited are primarily the conference presentation itself, with a link to the contributed talk page. The title accurately reflects the session’s content. The scientific rigor appears high, with clear methodological descriptions and quantitative analyses. However, the lack of detailed references to prior work in the transcript makes it difficult to assess the novelty and context fully. The adequacy between title and content is good, as the session indeed focuses on computational psychiatry and development.

208 words

Title / Content Match

The title accurately reflects the session's focus on computational psychiatry and development, featuring contributed talks on these topics.

Quality & Reliability

8/10

The session presents original research from multiple labs, with clear methodological descriptions and quantitative results. However, the lack of peer-reviewed publication details and the brevity of some talks limit the ability to fully verify claims.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

Concurring Sources

Contribution & Novelties

This session provides novel computational approaches to understanding psychiatric and developmental phenomena. Zhu’s work introduces the use of white-matter microstructure metrics (ADC, GFA) in virtual brain models, improving model fit and identifying potential biomarkers for schizophrenia. Van Geen’s latent cause inference framework offers a formal theory for persistent fear, distinguishing between generalization and selective maintenance pathways across development. Li’s large-scale EEG study reveals multi-dimensional developmental trajectories in visual representations, with distinct latent patterns. These contributions advance the field by linking computational models to clinical and developmental data.

Pour aller plus loin :

  • Wong-Wang model — Foundational model for local cortical dynamics used in Zhu’s work.
  • Latent cause inference — General concept underlying Van Geen’s model.
  • EEG decoding — Methods used in Li’s study to analyze visual representations.

127 words

Radar Profile

The radar profile shows high scores across all dimensions, indicating a well-rounded scientific session with strong information content, technical depth, and reliability. The session excels in providing original research with clear methodologies and quantitative results.

Reliability 8/10

💬 No comments were provided for analysis.