The method requires no assumptions about the baseline excess hazard and thus enables more reliable estimates in many practical situations when the assumptions of the more standard approaches are not met or are impossible to check. The EM algorithm based approach enables the user to obtain additional information and thus enriches the interpretation of the data.
COBISS.SI-ID: 24416217
Psuedo-observations present a generalized approach to fitting various survival models and checking their assumptions. The advantages of the method have been studied in various situations.
COBISS.SI-ID: 26578137
The paper evaluates the childhood cancer survival in Slovenia between 1957 and 2007. The 5 and 10-year survival is estimated. The trends in survival were analysed using the Cox model with restricted cubic splines to allow non-linear effects of covariates. The improvement in hazard is thus the most pronounced in case of leukemia, that represents more than a quarter of all patients. On the other hand, it seems to have almost stopped since 1990 in case of non-hodgkin limphoma and intracranial and intraspinal neoplasms.
COBISS.SI-ID: 25676249
We used microarrays to measure the gene expression of breast cancer patients included in a phase II neoadjuvant treatment study. We compared the gene-expression before and after chemotherapy and then evaluated if there were any differences in terms of gene expression between patients that responded to chemotherapy and those who did not. We determined the Gene Ontology pathways of each gene, and compared the pathways rather than genes. With this approach we were able to increase the power of our analysis and to identify some pathways pertinent to therapeutic response.
COBISS.SI-ID: 26386393
Reliability of the F-Scan plantar pressure measurement system is confirmed. Proper methodology, including intraclass correlations, different uses of coefficient of variation, Hotelling T2 test and individual sequence plots, is demonstrated and discussed with reference to statistical process control.
COBISS.SI-ID: 893801