Aconteceu nesta terça-feira, dia 12 de março de 2024, um encontro de estudantes do NIATS com o Prof. Steffen Walter da Universidade de Ulm na Alemanha. O encontro ocorreu das 13:30 às 13:30 no anfiteatro do Bloco 1E da Faculdade de Engenharia Elétrica, no campus Santa Mônica. Durante o encontro o Prof. Steffen Walter ministrou uma palestra sobre suas pesquisas na área do reconhecimento automático de dor e os estudantes do NIATS tiveram a oportunidade de apresentar suas pesquisas.

Além do Prof. Adriano de Oliveira Andrade e do Prof. Steffen Walter, os seguintes estudantes participaram do encontro: Ana Carolina Miziara (graduanda em Engenharia Biomédica), Luanne Cardoso Mendes, Camille Marques Alves, Ariana Moura Cabral, Pedro Henrique Bernardes Caetano, Letícia de Queiroz Martins, Eduardo de Moura Neto, José Renato Munari Nardo, Rodrigo Gomide Gonzaga, Sheida Mehrpour, Walter Rodrigues de Andrade, Maria Olivia Domingos Rezio Ramos, Leandro Rodrigues da Silva Souza e Alice de Oliveira Barreto Suassuna.

A participação do Prof. Steffen Walter nesse encontro aconteceu no âmbito de uma missão de trabalho internacional financiada pelo Projeto CAPES-PRINT UFU (P9-Tecnologias convergentes aplicadas à saúde e bem-estar). Durante a missão de trabalho o Prof. Steffen executou ações no âmbito de uma pesquisa internacionacional sobre a doença de Parkinson, que é coordenada pelo Prof. Adriano Andrade e apoiada pelo CNPq (Processos 442150/2023-7 e 302942/2022-0).

Resumo da palestra ministrada pelo Prof. Steffen Walter

Short CV: Prof. Steffen Walter has carried out and published preliminary work in the field of automated content analysis of psychotherapy conversations. Since 2008, he has been involved in several projects concerning the optimization of so-called "affective computing"/ “automated pain recognition”, in which the recognition of emotions and pain is investigated multimodally (biosignals, video, paralinguistics). Machine learning methods are used to classify basic emotions or valence/arousal and pain intensities. The BioVid, SenseEmotion and X-ITE datasets - which offers the possibility of training algorithms for pain and emotion recognition - were recorded by Walter and are used by working groups worldwide to optimize machine learning.

Title: Automated Pain Recognition - Where we are and where we want to go:  A perspective

Pain typically is measured by patient self-report, but self-reported pain is difficult to interpret and may be impaired or in some circumstances not possible to obtain. For instance, in patients with restricted verbal abilities such as neonates, young children, and in patients with certain neurological or psychiatric impairments (e.g., dementia). Additionally, the subjectively experienced pain may be partly or even completely unrelated to the somatic pathology of tissue damage and other disorders. Therefore, the standard self-assessment of pain does not always allow for an objective and reliable assessment of the quality and intensity of pain. Given individual differences among patients, their families, and healthcare providers, pain often is poorly assessed, underestimated, and inadequately treated. To improve assessment of pain, objective, valid, and efficient assessment of the onset, intensity, and pattern of occurrence of pain is necessary. To address these needs, several efforts are being made in machine learning for automated and objective assessment intensity of pain from multimodal approach (bio-signal, video [face, mimic], voice [paralinguistic]) as a powerful alternative to self-reported pain.