Yes / Big data and machine learning result in agile and robust healthcare by expanding raw data into useful patterns for data-enhanced decision support. The available datasets are mostly small and unbalanced, resulting in non-optimal classification when the algorithms are implemented. In this study, five novel machine learning experiments are conducted to address the challenges of small datasets by expanding these into big data and then utilising Random Forests. The experiments are based on personalised adaptable strategies for both balanced and unbalanced datasets. Multiple datasets from cryotherapy and immunotherapy are considered, however, hereby only immunotherapy is used. In the first experiment, artificially generated data is presented by increasing the observations of the dataset, each new data is four-time larger than the previous one, resulting in better classification. In the second experiment, the effect of volume on classification is considered based on the number of attributes. The attributes of each new dataset are built based on conditional probabilities. It did not make any difference, in obtained classification, when the number of attributes is increased to more than 879. In the third simulation experiment, classes of data are classified manually by dividing the data into a two-dimensional plane. This experiment is first performed on small data and then on expanded big data: by increasing observations, an accuracy of 73.68% is attained. In the fourth experiment, the visualisation of the enlarged data did not provide better insights. In the fifth experiment, the impact of correlations among datasets’ attributes on classification is observed, however, no improvements in performance are achieved. The experiments generally improved performance by comparing the classification results using the original and artificial data.

Background: Persistent infection with high risk HPV is a necessary but insufficient cause of cervical cancer. Behavioural, viral and host factors modulate the risk of HPV persistence. In this thesis, I explore the role of the vaginal microbiota, a host factor and the presence of multiple HPV infections, a viral factor in HPV persistence. Considering the limited data on the epidemiology of HPV related diseases in low and middle-countries (LMIC), and the limited success of cervical cancer screening strategies in many LMIC, I provide data on the distribution of HPV related diseases in Nigeria and evaluate the acceptability of innovative strategies to increase cervical cancer screening uptake. Methods/Results: To achieve my aims, I implemented a longitudinal cohort study of 1,020 women in Nigeria. I begin my results chapters with two methodological papers. Attrition is an important consideration for every longitudinal cohort, particularly in LMIC, therefore, I present my findings on attrition, determinants of attrition and practical strategies to ensure low attrition in studies conducted in LMIC. Considering that sexual behaviour is an important potential confounder in all HPV studies, and the reliability of self-reported history is often questioned, I present findings on the test-retest reliability of self-reported sexual behaviour history collected in my study. Having found that attrition levels were low and that self-reported sexual behaviour history was generally reliable within my cohort, I present my findings on the association between the vaginal microbiota and persistent hrHPV; and the role of multiple HPV infections in viral persistence. I found that the vaginal microbiota was associated with persistent hrHPV in HIV negative women, but not in HIV positive women; and that multiple HPV infections did not increase the risk of viral persistence when compared to single HPV infections. Next, I present my findings on the prevalence and incidence of anogenital warts in Nigeria, with additional reports on the prevalence of cervical cancer and other HPV associated cancers using data from two population based cancer registries. Finally, I present my findings on the acceptability of innovative strategies to improve cervical cancer screening uptake in Nigeria. I found that Nigerian women had a favorable attitude to the use of HPV DNA based screening as part of routine antenatal care, however attitudes towards the use of self-sampling techniques for HPV based cervical cancer screening varied by religious affiliations. Conclusion: In my thesis, I was able to systematically investigate the epidemiology of HPV infections in a LMIC. I considered the distribution of HPV related diseases, host and viral determinants of HPV persistence and investigated control strategies to reduce the burden of cervical cancer in a LMIC. My results provide useful data for surveillance, monitoring and evaluation of control programs on HPV and cervical cancer in Nigeria and may be useful to cervical cancer control programs in other LMIC.

Introduction. Sexually transmitted infections (STIs) continue to be highly prevalent among young women, and STIs continue to be a challenging health issue on college campuses. Studies have shown that the highest prevalence of human papillomavirus (HPV) is among young adult women, ages 20 to 24, including female college students. While the HPV vaccine has proven to be highly effective in preventing certain high-risk types of HPV, it is not effective in preventing all types of HPV or other STIs. Practicing other safer sex behaviors, in addition to condom use, also can help individuals protect themselves and their partners from acquiring HPV and other STIs. Purpose. Relationships between knowledge about HPV and the HPV vaccine, beliefs about HPV, acceptance of the HPV vaccine, and intentions to practice safer sex behaviors were assessed to determine if female college students who had a high acceptance of the HPV vaccine were also more likely to practice safer sex behaviors. Methods. A convenience sample of 2,706 undergraduate female college students, ages 18 to 24, from three U.S. public universities, completed an online survey that measured the following study variables: knowledge about HPV; knowledge about the HPV vaccine; acceptance of the HPV vaccine, specifically influential factors and barriers to vaccination; as well as attitudes, normative beliefs, control beliefs and intentions for practicing safer sex behaviors. Results. The majority of participants had a high level of knowledge of HPV (70.4%) and the HPV vaccine (73.7%). Over one-third of the participants (37.3%) received at least one dose of the HPV vaccine. Most participants thought they had a low susceptibility to HPV, as 54% thought they were unlikely to contract HPV; however, most (53.6%) thought that HPV would be a serious problem for them. The safer sex behavior that participants thought was the easiest was refusing to have sex with a partner that would not use a condom (51.8% “Strongly agree”) and the safer sex behavior that participants thought to be the most difficult was asking a partner to get tested for STIs (54.5% “Disagree”). Attitudes, normative beliefs, and control beliefs combined to strongly predict intentions (R = .730, p < .001), and attitudes was the strongest predictor for intentions to practice safer sex behaviors (β = .666), 95% CI [.649, .711]. There was a significant positive correlation between vaccine acceptance and intentions to practice safer sex behaviors (r = .087, p < .001), including likelihood of getting vaccinated against HPV (r = .098, p < .001). Conclusions. Findings from this study demonstrate the need for university student health centers to provide information about the HPV vaccine as part of broader STI prevention and social marketing campaigns, targeting female college students who are single, as well as those in monogamous relationships. Furthermore, because young men can be carriers of HPV and the HPV vaccine is now available to them, follow-up studies are needed to determine acceptance of the HPV vaccine among male college students as it relates to HPV knowledge, knowledge of the vaccine, and intentions to practice safer sex behaviors.

CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / CNPq - Conselho Nacional de Desenvolvimento Científico e Tecnológico / FAPEMIG - Fundação de Amparo a Pesquisa do Estado de Minas Gerais / UFU - Universidade Federal de Uberlândia / UNIPAM - Centro Universitário de Patos de Minas / A metformina (MET) é um fármaco antidiabético utilizado para prevenir a liberação de glicose hepática e aumentar a sensibilidade à insulina nos tecidos. Pacientes diabéticos com câncer têm, em adição, terapia medicamentosa antineoplásica. A doxorrubicina (DXR) é um agente quimioterápico antineoplásico que interfere com enzimas topoisomerase II e gera radicais livres. A MET isolada (2,5; 5,0; 10,0; 25,0 ou 50,0 mM) foi avaliada quanto à mutagenicidade, recombinogenicidade e carcinogenicidade e associado com DXR (0,4 mM) para antimutagenicidade, antirecombinogenicidade e anticarcinogenicidade, utilizando o “Teste para Detecção de Mutação e Recombinação Somática” e o “Teste para Detecção de Clones de Tumores Epiteliais” em Drosophila melanogaster. A MET isolada não induziu mutação ou recombinação, mas foram observados efeitos moduladores da MET sobre as lesões de DNA induzidas pela DXR nas concentrações mais elevadas. Na avaliação da carcinogênese, a MET isolada não induziu tumores, mas quando associado com DXR, MET também reduziu os tumores induzidos por DXR nas concentrações mais elevadas. Sendo assim, nas presentes condições experimentais a MET isolada não apresentou efeitos mutagênicos, recombinogênicos e carcinogênicos, mas foi capaz de modular o efeito da DXR na indução de danos ao DNA e tumores em D. melanogaster. Acredita-se que este efeito modulador esteja relacionado principalmente aos efeitos antioxidantes, anti-inflamatórios e apoptóticos deste medicamento, embora tais efeitos não tenham sido avaliados diretamente. / Metformin (MET) is an anti-diabetic drug used to prevent hepatic glucose release and increase tissue insulin sensitivity. Diabetic cancer patients are on additional therapy with anticancer drugs. Doxorubicin (DXR) is a cancer chemotherapeutic agent that interferes with the topoisomerase II enzyme and generates free radicals. MET (2.5, 5, 10, 25 or 50 mM) alone was examined for mutagenicity, recombinogenicity and carcinogenicity, and combined with DXR (0.4 mM) for antimutagenicity, antirecombinogenicity and anticarcinogenicity, using the Somatic Mutation and Recombination Test and the Test for Detecting Epithelial Tumor Clones in Drosophila melanogaster. MET alone did not induce mutation or recombination. Modulating effects of MET on DXR-induced DNA damage were observed at the highest concentrations. In the evaluation of carcinogenesis, MET alone did not induce tumors. When combined with DXR, MET also reduced the DXR-induced tumors at the highest concentrations. Therefore, in the present experimental conditions, MET alone did not present mutagenic/recombinogenic/carcinogenic effects, but it was able to modulate the effect of DXR in the induction of DNA damage and of tumors in D. melanogaster. It is believed that this modulating effect is mainly related to the antioxidant, anti-inflammatory and apoptotic effects of this drug, although such effects have not been directly evaluated. / Dissertação (Mestrado)

The thesis deals with design and implementation of a tool for simulating marks of chosen skin diseases into a synthetic fingerprint. The diseases selected to work with are warts and atopic eczema. The marks of diseases are generated into a synthetic fingerprint image created by the SFinGe application. Existing disesase-affected fingerprints from the STRaDe database are analysed in detail. Then, methods for simulating the diseases into a synthetic fingerprint are proposed, implemented, and the results are evaluated.