BIOGRAFIA
Professor White has been the Medical Director of the Institute for Addressing Strangulation, UK since it was established in October 2022. Prior to that she was the Clinical Director of the Saint Mary's Sexual Assault Referral Centre, Manchester, UK for 19 years where she continues to undertake clinical forensic medical examinations.
Professor White was the Sexual Offence Medicine Lead for the Faculty of Forensic Medicine, having held that role for over ten years.

RESUMO
Non-fatal strangulation and the challenges of making invisible harm visible.

Why NFS is considered a red flag in terms of future lethality in the context of interpersonal violence.
The establishment of IFAS and its role in the UK
What is known regarding the impact of NFS on the body
The evidence base regarding injuries and NFS
The rise of NFS in consensual sex
Guidelines for the clinical management of NFS

BIOGRAFIA
Daniel Martins de Barros é médico psiquiatra, professor colaborador do Departamento de Psiquiatria da Faculdade de Medicina da USP (FMUSP). Doutor em Ciências e bacharel em Filosofia pela USP. Coordenador do Núcleo de Psiquiatria Forense no Instituto de Psiquiatria do Hospital das Clínicas da FMUSP.

RESUMO
A loucura e as leis - o dilema da Psiquiatria Forense num conto de Machado de Assis

A determinação da responsabilidade é o dilema central da Psiquiatria Forense. A partir de uma crônica de Machado de Assis sobre a fuga do Hospício Nacional de Alienados e do caso de Custódio Alves Serrão, esta apresentação discute a diferença entre ter um transtorno mental e ser inimputável. Machado percebe, com ironia, que atos aparentemente lúcidos podem ser praticados por internos, enquanto crimes graves podem ser apressadamente atribuídos à loucura. O caso de Serrão evidencia que o diagnóstico psiquiátrico, por si só, não basta para afastar a responsabilidade penal. É necessário avaliar se, no momento do ato, havia prejuízo relevante da capacidade de entender o caráter ilícito do fato ou de determinar-se conforme esse entendimento. A crônica antecipa, assim, o critério biopsicológico moderno e mostra o papel da perícia como auxiliar da justiça.

BIOGRAFIA
Dr. Katharina Grafinger is a Senior Researcher at the Institute of Forensic Medicine in Bern, Switzerland. She studied Forensic Science in Germany, Spain, Portugal, and England and got her PhD in 2019 at the University of Bern. Dr. Grafinger conducted research in Germany, Belgium, Finland, and Switzerland and has authored more than 30 peer-reviewed publications.
Her research focuses on new psychoactive substances, receptor pharmacology, and the application of machine learning in toxicological research. Her current work combines experimental pharmacology with statistical and machine-learning approaches to improve the assessment of emerging new psychoactive substances.

RESUMO
Machine Learning in Forensic Toxicology- Concepts, applications and challenges in bioanalysis, ADME, and toxicodynamics

Advances in analytical technologies have dramatically increased the amount and complexity of data generated in forensic toxicology, creating new opportunities for machine learning (ML).
This presentation will introduce fundamental ML concepts and illustrate applications across bioanalysis, toxicokinetics, and toxicodynamics. Examples will include alternative screening approaches, retention-time prediction, metabolomics, estimation of time since drug consumption, and prediction of pharmacological and toxicological properties. Particular attention will be given to applications involving new psychoactive substances, where rapidly evolving drug markets and limited experimental data present both opportunities and challenges for predictive approaches. The presentation will also critically address current limitations of ML, including data quality, model validation, generalizability, and translation into forensic practice. Finally, it will highlight the importance of interdisciplinary collaboration, combining forensic and toxicological domain expertise with data science to develop meaningful, robust, and applicable ML models.