ORIGINAL
A Lesão Cerebral Traumática (LCT) continua a ser uma crise de saúde pública global com elevadas taxas de morbilidade e mortalidade. O estudo recente de Silveira et al., utilizando modelos de aprendizado de máquina como Máquinas de Vetores de Suporte e Gradient Boosting, para prever desfechos de LCT, é um passo à frente, louvável e oportuno, demonstrando com sucesso o potencial da inteligência artificial (IA) em melhorar a precisão prognóstica além dos sistemas de pontuação tradicionais. No entanto, uma limitação significativa foi o âmbito restrito do estudo, concentrado exclusivamente na previsão da mortalidade intra-hospitalar. O potencial total do aprendizado de máquina em neurotrauma estende-se muito além desta métrica. Portanto, pesquisas futuras devem alargar o horizonte preditivo incluindo desfechos críticos, tais como complicações secundárias (pneumonia, sepsis), tempo de permanência na UCI, recuperação funcional a longo prazo usando escalas como a Escala de Resultados de Glasgow Estendida, e desfechos neurocognitivos. Além disso, o papel da IA na gestão da LCT abrange o diagnóstico, incluindo a interpretação de neuroimagem e a previsão da necessidade cirúrgica. Dada a crescente incidência global de LCT, este estudo serve como alerta para a comunidade neurocirúrgica abraçar e implementar as vastas e inexploradas possibilidades do aprendizado de máquina para fornecer cuidados mais abrangentes e holísticos a pacientes em todo o mundo.
Traumatic Brain Injury (TBI) remains a formidable global public health crisis with high morbidity and mortality rates. The recent study by Silveira et al., which utilized machine learning models such as Support Vector Machines and Gradient Boosting to predict TBI outcomes, is a commendable and timely step forward. The authors successfully demonstrated the potential of artificial intelligence (AI) in enhancing prognostic accuracy beyond traditional scoring systems. However, a significant limitation was the narrow scope of the study, which exclusively focused on predicting in-hospital mortality. The full potential of machine learning in neurotrauma extends far beyond this single metric. Therefore, future research must broaden its predictive horizon to include critical outcomes such as secondary complications (pneumonia, sepsis), length of intensive care unit stay, long-term functional recovery using scales like the Glasgow Outcome Score Extended, and neurocognitive outcomes. Furthermore, the role of artificial intelligence in TBI management encompasses diagnostics, including the interpretation of neuroimaging and the prediction of surgical necessity. Given the increasing global incidence of TBI, this study should serve as a clarion call for the neurosurgical community to embrace and deploy the vast, untapped possibilities of machine learning to provide more comprehensive and holistic care for patients worldwide.
1. Silveira SPP, Correia MM, Alberto GB, et al. Using machine learning to predict outcomes after traumatic brain injury. J Bras Neurocir. 2025;36(3):301-7. https://doi.org/10.22290/jbnc.2025.360308.
2. Ahmed M, Karaboue A, Ministeri F, et al. Traumatic brain injury as a public health issue : epidemiology, prognostic factors and useful data from forensic practice. Health Care. 2024;12(2266):e78900.
3. Andraos C, Siddiqi A, Brazdzionis J, Siddiqi J. Limitations of the glasgow coma scale : challenges and considerations. Cureus. 2025;17(2): 1-7. https://doi.org/10.7759/cureus.78900. PMid:40091938.
4. Chaimae T, Imad H, Metrane A. Machine learning and deep learning in health informatics: advancements, applications, and challenges. Eng Lett. 2025;33(5):1448-61.
11. Wang Y. Global, regional, and national burdens of traumatic brain Ilorin, Kwara State, Nigeria injury, spinal cord injury, and skull fracture and their attributable risk E-mail: ooadekunle@rocketmail.com factors from 1990 to 2021 : a systematic analysis of the global burden of disease study 2021. Front Public Health. 2025;13:1622693. https:// doi.org/10.3389/fpubh.2025.1622693. PMid:40910052. Funding: nothing to disclose.
1Division of Neurosurgery, Department of Surgery, University of Ilorin Teaching Hospital, Ilorin, Kwara State, Nigeria.
Received Dec 14, 2025
Accepted Jan 12, 2026
