Sign In

Michele Ferrante

Computational Neuroscience & Computational Psychiatry

Michele Ferrante is the Computational Psychiatry Program Director at the National Institute of Mental Health (NIMH).5 He recently returned to this position at NIMH.5 Ferrante has a background in neuroscience, computational psychiatry, and digital health.1 He has been involved in various research areas, including explainable artificial intelligence for neuroscience, computational modeling of neuronal systems, and advancing translational research using computational methods.1

Ferrante's work focuses on fostering innovative computational approaches to identify and validate novel mechanisms, biomarkers, and treatment targets relevant to the prevention and treatment of psychiatric disorders.3 He supports research projects that use advanced computational methods with behavioral, biological, and clinical data to decipher complex mechanisms involved in mental disorders.3

Prior to his current role, Ferrante has held positions such as NIMH Program Director for Computational Neuroscience and Computational Psychiatry.2 He has contributed to numerous scientific publications and has been involved in workshops and initiatives to advance the field of computational psychiatry.14

Highlights

Aug 9 · linkedin

Alexa, Print Me a Better Brain.

The article in the comments focuses on regenerative cell therapy w/ 3D bioprinting & how In situ additive biomanufacturing of structures may boost regenerative medicine.

Regenerative medicine utilizes viable cells to repair or replace damaged tissues & organs. However, efficient delivery, survival, & functional integration of therapeutic cells at the target site remain significant hurdles.

3D Bioprinting involves the layer-by-layer deposition of cells & biomaterials to create complex tissue structures. It allows precise spatial organization of cells & biomaterials, closely mimicking the native tissue architecture.

In Situ Bioprinting can directly print tissues at the site of injury or defect. Custom-made grafts that fit precisely to the target defects can be made to improve therapeutic outcomes.

Applications

Skin & Muscle Repair: Matching the tissue architecture is crucial for achieving structural, functional, & aesthetic outcomes.

Other Applications: The approach can be adapted for various cell therapy scenarios where conforming to the native geometries of the therapeutic site is essential, like neuroscience.

Techniques

  1. Extrusion Bioprinting uses a nozzle to deposit bioinks (cells & other components) layer by layer.

  2. Inkjet Bioprinting ejects tiny droplets of bioinks into precise patterns.

  3. Vat Polymerization Bioprinting uses vats preloaded w/ bioinks to produce 3D constructs w/ photochemistry.

Relevance to Computational Psychiatry

  1. 3D bioprinting can create brain tissue models that can be used to study psychiatric disorders. Computational models can simulate these disorders, allowing researchers to understand the underlying mechanisms & test new treatments.

  2. Personalized brain tissue models created through 3D bioprinting can be used w/ computational models to test the efficacy & safety of psychiatric medications. This can lead to more effective & tailored treatments.

  3. 3D bioprinting might be used to develop regenerative therapies for brain injuries or neurodegenerative diseases. Computational psychiatry can help in predicting the outcomes of these therapies & optimizing treatment plans.

  4. Computational psychiatry often involves integrating various types of data (e.g., genetic, imaging, clinical). 3D bioprinting can provide additional data points, such as the structural & functional properties of bioprinted brain tissues, enhancing the overall understanding of psychiatric conditions.

3D bioprinting holds promise for advancing regenerative medicine by overcoming current limitations & improving the precision & effectiveness of cell therapies. However, significant challenges remain, including the complexity of replicating the intricate architecture of brain tissues & ensuring the long-term viability & functionality of bioprinted constructs. Further research & technological advancements are essential to fully realize the potential of 3D bioprinting in psychiatry & other medical fields.

May 1 · linkedin

LLMs Evaluations Metrics: A New Frontier in Mental Health Diagnostics?

As we witness the rapid development of computational technology and metrics designed to evaluate and assess the performance and accuracy of LLMs in various contexts, it raises an interesting question.

Could these technologies be adapted and applied to evaluate language output in individuals with mental health conditions?

This could potentially open up new avenues for understanding and treating mental health disorders. For instance, there is a hallucination index and many other potentially useful metrics, such as:

Perplexity: This is a crucial metric used to evaluate the efficacy of LLMs. It measures the uncertainty of a language model’s predictions. In the context of mental health, it could potentially be used to measure the uncertainty or variability in a patient’s speech patterns.

ROUGE (Recall-Oriented Understudy for Gissing Evaluation): This is a set of metrics used for evaluating automatic summarization and machine translation. It could potentially be used to evaluate the ability of a patient to summarize information or recall details.

Diversity: This metric evaluates the range and variety of the LLM’s outputs. It could potentially be used to assess the range and variety of a patient’s speech, which might be indicative of certain mental health conditions linked to mood and thought disorders.

Zero-shot Evaluation: This is a method of evaluating the ability of an LLM to generalize to tasks it has not been explicitly trained on. In a mental health context, it could potentially be used to assess a patient’s ability to generalize or apply learned information to new situations.

Coherence: This metric evaluates the logical consistency of the LLM’s outputs. In a mental health context, it could potentially be used to assess the coherence in a patient’s speech, which might be indicative of certain cognitive disorders.

Creativity: This metric evaluates the novelty and originality of the LLM’s outputs. In a mental health context, it could potentially be used to assess the creativity in a patient’s speech or thought patterns, which might be indicative of certain personality traits or mental health conditions.

In conclusion, the parallels drawn between Large Language Models and human cognitive processes not only provide a fascinating perspective on AI development but also open up potential avenues for understanding and treating mental health disorders.

By adapting computational technologies and metrics used in AI, we could potentially develop innovative methods to evaluate language output in individuals with mental health conditions. While these are early days and much research is needed, the possibilities are intriguing and hold promise for the future.

As we continue to advance in our understanding of both AI and human cognition, it’s exciting to envision a future where these two fields intersect to create transformative solutions for mental health.

Related Questions

What are Michele Ferrante's main research interests?
How has Michele Ferrante contributed to computational psychiatry?
What projects is Michele Ferrante currently working on at NIMH?
Can you provide examples of Michele Ferrante's notable publications?
What is the significance of Michele Ferrante's work on computational phenotyping?
Michele Ferrante
Michele Ferrante, photo 1
Michele Ferrante, photo 2
Add to my network

Location

Washington DC-Baltimore Area