We aim at creating a stable and quality driven research cluster in Biomedicine.

Biostatistics Research Group (IO1)

Main lines of research:

- Regression Models for Survival Analysis in non-standard contexts: non-proportional risks, additive structures, specification contrasts, location-scale model
- Models and methods for time analysis events successively recorded under dependent censoring, including the caseloads of the missing censoring indicators, and not Markovian multi-state models.
- Kaplan-Meier Survival in complicated contexts: an observational bias, random truncation, informative censoring, sampling of group items (families, repeated measurements ...)
- Biostatistical methods in high dimension: functional data processing, multi-testing problems
- Flexible statistical inference (methods of smoothing, additive and generalized additive models including interactions), and for biometrical targets: ROC curve, odds ratio, risk functions, k-sample tests.
- Applications in oncology (breast cancer, childhood cancer, leukemia ...), neuroscience (neural activity), surgery (risk of postoperative infection)

Description:

The group of Prof. Jacobo de Uña (http://sidor.uvigo.es) develops statistical methodology in an integrated way (motivation, method definition, implementation, suitability, user friendliness, and applicability to solve the stated problem), in various fields that are constantly in need of quantitative techniques. Precisely, Biomedical Sciences is highly represented in their activity, funded with competitive projects under the National Plan, and involved with SERGAS (Galician Health Service) and the CIBER-ESP (Network of Biomedical Research Centers for Epidemiology and Public Health) in San Sebastian and Barcelona, developing epidemiological software; and also, agreements with the IPO (Portuguese Oncology Institute), for statistical analysis of cancer data.

Items such as survival analysis, ROC curves, clinical trials, epidemiology, Odds Ratio, etc. are all framed in the field of biometric techniques that have increased its practical potential for efficient statistical analysis of biomedical data of a complex nature.

The field of biostatistics has grown exponentially in recent years thanks to the massive collection of quantitative information (genes, neural function etc.), so that computational statistics is becoming essential in biomedicine. New challenges in health sciences are also concerned with the efficient analysis of quantitative information of complex nature, for which classical statistical approaches based on rigid structures (eg normal populations) and designed to moderate scale can not give an adequate response.