Your browser doesn't support javascript.
loading
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 36.101
Filtrar
1.
Recurso na Internet em Português | LIS - Localizador de Informação em Saúde | ID: lis-47914

RESUMO

Pacientes que sofrem com atrasos em exames e tratamento do câncer na pandemia


Assuntos
Neoplasias da Mama , Pandemias , Neoplasias da Mama/diagnóstico
2.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 4369-4372, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018963

RESUMO

In this report, we introduce a flexible board combining a custom switching circuit and 16 integrated antennas for a time-domain ultrawideband radar for breast health monitoring in one device. The goal of this study is to assess the suitability of the flexible prototype for tumor detection using carbon-polyurethane experimental breast models and comparing the performance to an earlier prototype with a rigid switching circuit and 16 separate antennas. The flexible antenna array allows direct contact with the patient skin while reducing the number of RF and DC cables needed in the previously reported system. We evaluate that the introduced flexible board successfully transmits and receives the microwave signals, and isolates tumor responses using a simple method. However, we observe that the board exhibits an early signal in the recordings of all antenna pairs which corresponds to direct cross-talk on the board and is not part of the signal that has passed through the phantom.


Assuntos
Neoplasias da Mama , Mama , Neoplasias da Mama/diagnóstico , Humanos , Micro-Ondas , Modelos Teóricos , Imagens de Fantasmas
3.
Nihon Koshu Eisei Zasshi ; 67(9): 593-602, 2020.
Artigo em Japonês | MEDLINE | ID: mdl-33041285

RESUMO

Objective The purpose of this study was to investigate the influence of age, period, birth cohort, and regional differences in the detection of breast cancer using screening data.Method Data from the Japan Cancer Society's breast cancer screening program, collected from 21 prefectural branches between 2004 to 2015, were used to generate age-specific estimates of cancer detection for women aged between 40 to 79 years. We used Bayesian age-period-cohort (APC) analyses based on the cohort table to describe the simultaneous effects of age, period, and cohort on breast cancer detection rates to understand the population dynamics underlying the detection patterns. We also incorporated region as a random effect to examine regional characteristics.Results The age effect showed bimodality in the late 40s and late 50s. The period effect decreased from 2004 to 2007 and remained constant thereafter. The cohort effect showed that the detection rate for women born between 1943 and 1958 was high. Furthermore, we found regional differences in the breast cancer detection rate: Miyazaki, Fukui, Tochigi, and Hokkaido prefectures showed higher detection rates, while Kagoshima and Chiba prefecture had lower rates.Conclusion Age effect has the strongest influence on the secular trend of breast cancer detection, and there is a regional difference in the detection rate. The present study that used screening data presented similar results to those of previous studies. The National Cancer Registry, based on the Cancer Registry Act of 2016, reports accurate national data. Similar to the National Cancer Registry data, analysis using screening data has immediacy and could be used for disease prevention.


Assuntos
Neoplasias da Mama/diagnóstico , Neoplasias da Mama/epidemiologia , Detecção Precoce de Câncer/estatística & dados numéricos , Programas de Rastreamento , Fatores Etários , Teorema de Bayes , Neoplasias da Mama/prevenção & controle , Estudos de Coortes , Interpretação Estatística de Dados , Japão/epidemiologia , Sistema de Registros , Fatores de Tempo
4.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 1136-1139, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018187

RESUMO

Computerized parenchymal analysis has shown potential to be utilized as an imaging biomarker to estimate the risk of breast cancer. Parenchymal analysis of digital mammograms is based on the extraction of computerized measures to build machine learning-based models for the prediction of breast cancer risk. However, the choice of the region of interest (ROI) for feature extraction within the breast remains an open problem. In this work we perform a comparison between five different methods suggested in the literature for automated ROI selection, including the whole breast (WB), the maximum squared (MS), the retro-areolar region (RA), the lattice-based (LB), and the polar-based (PB) selection methods. For the experiments, we built a retrospective dataset of 896 screening mammograms from 224 women (112 cases and 112 healthy controls). The performance of each ROI selection method was measured in terms of the area under the curve (AUC) values. The AUC values varied between 0.55 and 0.79 depending on the method and experimental settings. The best performance on an independent test set was achieved by the MS method (AUC of 0.59, 95% CI: 0.55-0.64). This method is fully-automated and does not require adjusting hyper-parameters. Based on our results, we prompt the use of the MS method for ROI selection in the computerized parenchymal analysis for breast cancer risk assessment.


Assuntos
Neoplasias da Mama , Área Sob a Curva , Neoplasias da Mama/diagnóstico , Feminino , Humanos , Mamografia , Estudos Retrospectivos , Medição de Risco
5.
Annu Int Conf IEEE Eng Med Biol Soc ; 2020: 2178-2181, 2020 07.
Artigo em Inglês | MEDLINE | ID: mdl-33018438

RESUMO

Cancer has affected the human community to a large extent due to its low survival rate towards the end stage of the disease. It is asymptomatic in many cases during the initial stage. Thus the dependency on early diagnosis and regular check up increases manifold. Computer Aided Diagnostic Model is the need of the hour which will increase the diagnostic efficiency. A total of 400 images acquired from the Digital Database for Screening Mammography have been used here for analysis. This paper proposes a novel technique to differentiate benign and malignant breast lesions in mammograms using multiresolution analysis and Schmid Filter Bank, which were not reported earlier. A three level Haar wavelet decomposed image(L1, L2, L3) is obtained for each Region of Interest. In each level Texton based analysis is further investigated through Schmid filter bank. Statistical features and Haralick's Features are obtained from filter response and Gray Level Cooccurence Matrix respectively. Partition Membership Filter is further applied to the feature matrix for feature partitioning. The method shows maximum accuracy of 98.63% and Area under Curve of 0.981 using Random Forest Classifier and ten fold cross validation.


Assuntos
Neoplasias da Mama , Análise de Ondaletas , Mama/diagnóstico por imagem , Neoplasias da Mama/diagnóstico , Detecção Precoce de Câncer , Humanos , Mamografia
6.
Tumour Biol ; 42(10): 1010428320963811, 2020 Oct.
Artigo em Inglês | MEDLINE | ID: mdl-33028151

RESUMO

This study aimed at investigating the expression of candidate microRNAs (miRs), at initial diagnosis, during neoadjuvant chemotherapy, and after the tumor resection in locally advanced breast cancer patients. Plasma samples were collected from locally advanced breast cancer patients (n = 30) and healthy subjects (n = 20) for the detection of candidate miRs' expression using the real-time quantitative polymerase chain reaction. At initial locally advanced breast cancer diagnosis, the expression of miR-21, miR-181a, and miR-10b was significantly increased, whereas that of miR-145 and let-7a was significantly decreased, compared to the healthy individuals. The diagnostic accuracy of miR-21 was superior to both carcinoembryonic antigen and carcinoma antigen 15-3 as diagnostic biomarkers for locally advanced breast cancer. By the end of the treatment, the expression of altered miRs rebound to control values. The expression levels of candidate plasma miRs are useful diagnostic biomarkers, as well as monitoring a proper response for locally advanced breast cancer patients to the treatment. Furthermore, miR-10b and miR-21 can be considered as predictive biomarkers for progression-free survival.


Assuntos
Biomarcadores Tumorais , Neoplasias da Mama/diagnóstico , Neoplasias da Mama/genética , MicroRNA Circulante , MicroRNAs , Adulto , Neoplasias da Mama/mortalidade , Neoplasias da Mama/terapia , Terapia Combinada , Feminino , Regulação Neoplásica da Expressão Gênica , Humanos , Estimativa de Kaplan-Meier , Pessoa de Meia-Idade , Gradação de Tumores , Metástase Neoplásica , Estadiamento de Neoplasias , Prognóstico , Reação em Cadeia da Polimerase em Tempo Real , Sensibilidade e Especificidade
10.
Medicine (Baltimore) ; 99(33): e20996, 2020 Aug 14.
Artigo em Inglês | MEDLINE | ID: mdl-32871976

RESUMO

INTRODUCTION: Pure mucinous carcinoma is a rare type of breast carcinoma, but it usually has a favorable prognosis. Tumors of pure mucinous carcinoma are typically positive for both estrogen receptor (ER) and progesterone receptor (PR), and they do not commonly overexpress human epidermal growth factor receptor 2 (HER2). However, when tumors have HER2 overexpression and are progesterone receptor negative, the prognosis is worse. PATIENT CONCERNS: A 59-year-old female reported a slow growth mass of 3 years, which was radiologically diagnosed as fibroadenoma at another institution. The patient came to our institution for treatment and follow-up. She had no salient past history. DIAGNOSIS: Excisional biopsy revealed a pure mucinous breast carcinoma that was ER (100%, moderate-strong intensity), PR(-), 5% Ki-67 (+), and HER2(3+) by immunohistochemistry. The HER2 gene was found to be amplified by fluorescence in situ hybridization (FISH). The clinical staging was T2N0M0, with pathological grade I, subtype luminal B. INTERVENTIONS: After a modified radical mastectomy, she received four 21-day cycles of intravenous docetaxel (75 mg/m), intravenous cyclophosphamide (600 mg/m), and intravenous trastuzumab (8 mg/kg) (loading dose) on day 1 followed by 6 mg/kg every 3 weeks to complete a full year of treatment. She then received 2.5 mg of letrozole daily for 5 years. OUTCOMES: After following up for 2 years, the patient's outcome was survival without recurrence. Cardiac ultrasounds were performed every 3 months and there was no change in the left ventricular ejection fraction (LEVF). CONCLUSION: It is essential to correctly diagnose the ER(+), PR(-) HER2(+) subtype in mucinous carcinoma. This type should be treated with chemotherapy and anti-HER2 therapy, as well as aromatase inhibitor endocrine therapy.


Assuntos
Adenocarcinoma Mucinoso/genética , Adenocarcinoma Mucinoso/terapia , Neoplasias da Mama/genética , Neoplasias da Mama/terapia , Receptor ErbB-2/genética , Adenocarcinoma Mucinoso/diagnóstico , Adenocarcinoma Mucinoso/patologia , Neoplasias da Mama/diagnóstico , Neoplasias da Mama/patologia , Diagnóstico Diferencial , Feminino , Humanos , Pessoa de Meia-Idade , Receptores Estrogênicos/genética , Receptores de Progesterona/genética
11.
In Vivo ; 34(5): 3047-3053, 2020.
Artigo em Inglês | MEDLINE | ID: mdl-32871851

RESUMO

BACKGROUND/AIM: Coronavirus disease is spreading worldwide. Due to fast transmission and high fatality rate drastic emergency restrictions were issued. During the lockdown, only urgent medical services are guaranteed. All non-urgent services, as breast cancer (BC) screening, are temporarily suspended. The potential of breast cancer screening programs in increasing the survival rate and decreasing the mortality rate has been widely confirmed. Suspension could lead to worse outcomes for breast cancer patients. Our study aimed to analyse the data and provide estimates regarding the temporary BC screening suspension. PATIENTS AND METHODS: Data regarding breast cancer and respective screening programs were achieved through literature research and analysis. RESULTS: Considering three different scenarios with respect to the lockdown's impact on breast cancer screening, we estimate that approximately 10,000 patients could have a missed diagnosis during these 3 months. Considering a 6-month period, as suggested by the Imperial college model, the number of patients who will not receive a diagnosis will rise to 16,000. CONCLUSION: Breast cancer screening should be resumed as soon as possible in order to avoid further breast cancer missed diagnosis and reduce the impact of delayed diagnosis.


Assuntos
Neoplasias da Mama/diagnóstico , Infecções por Coronavirus/epidemiologia , Detecção Precoce de Câncer , Programas de Rastreamento/tendências , Pneumonia Viral/epidemiologia , Idoso , Betacoronavirus/patogenicidade , Neoplasias da Mama/complicações , Neoplasias da Mama/epidemiologia , Infecções por Coronavirus/complicações , Infecções por Coronavirus/virologia , Feminino , Humanos , Mamografia , Pessoa de Meia-Idade , Pandemias , Pneumonia Viral/complicações , Pneumonia Viral/virologia
12.
Chirurgia (Bucur) ; 115(4): 511-519, 2020.
Artigo em Inglês | MEDLINE | ID: mdl-32876025

RESUMO

Breast adenoid cystic carcinoma (BACC) is a rarely encountered malignant breast neoplasm with a favourable outcome, despite its triple-negative receptor status. It is comprising less than 0.1% of all breast cancers. The more usual primary site of the adenoid cystic carcinoma is seen in the salivary glands, however BACC prognosis is better than the one for salivary gland and also than that of other breast invasive carcinoma. BACC also known to have fewer lymph node metastases as well as fewer distant metastases, hereby we present 73 years old female with previous history of breast ductal carcinoma in-situ, then developed contralateral breast adenoid cystic carcinoma.


Assuntos
Neoplasias da Mama/diagnóstico , Carcinoma Adenoide Cístico/diagnóstico , Carcinoma Intraductal não Infiltrante/diagnóstico , Idoso , Feminino , Humanos , Resultado do Tratamento
13.
BMC Public Health ; 20(1): 1467, 2020 Sep 29.
Artigo em Inglês | MEDLINE | ID: mdl-32993596

RESUMO

BACKGROUND: Breast cancer is one of the leading public health problem globally, especially in low-resource countries (LRCs). Breast cancer screening (BCS) services are an effective strategy for early determining of breast cancer. Hence, it is imperative to understand the utilisation of BCS services and their correlated predictors in LRCs. This study aims to determine the distribution of predictors that significantly influence the utilisation of BCS services among women in LRCs. METHODS: The present study used data on 140,974 women aged 40 years or over from 14 LRCs. The data came from country Demographic and Health Surveys (DHS) between 2008 and 2016. Multivariate logistic regression analysis was employed to investigate the significant predictors that influence the use of BCS services. RESULTS: The utilisation of BCS services was 15.41%, varying from 81.10% (95% CI: 76.85-84.73%) in one European country, to 18.61% (95% CI: 18.16 to 19.06%) in Asian countries, 14.30% (95% CI: 13.67-14.96%) in American countries, and 14.29% (95% CI: 13.87-14.74%). Factors that were significantly associated to increase the use of BCS services include a higher level of education (OR = 2.48), advanced age at first birth (> 25 years) (OR = 1.65), female-headed households (OR = 1.65), access to mass media communication (OR = 1.84), health insurance coverage (OR = 1.09), urban residence (OR = 1.20) and highest socio-economic status (OR = 2.01). However, obese women shown a significantly 11% (OR = 0.89) lower use of BSC services compared to health weight women. CONCLUSION: The utilisation of BCS services is low in many LRCs. The findings of this study will assist policymakers in identifying the factors that influence the use of BCS services. To increase the national BCS rate, more attention should be essential to under-represented clusters; in particular women who have a poor socioeconomic clusters, live in a rural community, have limited access to mass media communication, and are have a low level educational background. These factors highlight the necessity for a new country-specific emphasis of promotional campaigns, health education, and policy targeting these underrepresented groups in LRCs.


Assuntos
Neoplasias da Mama/diagnóstico , Detecção Precoce de Câncer/estatística & dados numéricos , Acesso aos Serviços de Saúde/estatística & dados numéricos , Serviços de Saúde da Mulher/estatística & dados numéricos , Saúde da Mulher/estatística & dados numéricos , Adulto , Idoso , Ásia , Neoplasias da Mama/epidemiologia , Europa (Continente) , Feminino , Educação em Saúde/estatística & dados numéricos , Humanos , Pessoa de Meia-Idade , Estados Unidos
15.
PLoS One ; 15(9): e0237925, 2020.
Artigo em Inglês | MEDLINE | ID: mdl-32911496

RESUMO

BACKGROUND: Invasive Breast Cancer (IBC) risk estimates continue to be based on data collated from cancer registries, i.e., retrospective research that excludes disease-free women. For women without a prior diagnosis, these estimates inflate both risk and screening frequency recommendations and inadvertently increase recently recognized harms from overdiagnosis and overtreatment. OBJECTIVE: To estimate the likelihood that pre or postmenopausal women with no prior diagnosis will remain free of IBC in order to enable evidence-based screening recommendations. METHODS: Prospective data from 21 studies of 2,402,672 women were analyzed, updating our previously published systematic search of 19 studies. This second systematic search included PubMed and The Cochrane Library from 2012 through April 2019. Inclusion criteria: only studies reporting the number of women enrolled, length of follow-up, and number of women diagnosed with IBC. Linear regression was used to estimate the percentage of women expected to remain free from an IBC diagnosis based on follow-up duration. To minimize non-response bias and selective outcome bias, only studies reporting outcomes for all enrolled women followed for similar, specific lengths of time were included. Sensitivity analyses confirm that the overall findings were unchanged by age at enrollment, menopausal status, screened women, variation in sample size, duration of follow-up, and heteroskedasticity. RESULTS: The calculated percentage of women remaining IBC-free after follow-ups of 5, 10, 15, 20 and 25 years decreases uniformly by about one-fourth of one percent per year, i.e., 0.255% (95% CI: -0.29, -0.22; p < .0001). At 25 years, the expected percentage of women with no invasive breast cancer is 93.41% (95% CI: 92.75, 94.07). CONCLUSIONS: Over 99.7% of pre/postmenopausal women with no prior diagnosis continued with no IBC each year, with 93.41% still free after 25 years. Our study supports the medical justification for reducing the frequency of mammograms for menopausal women with no prior IBC diagnosis.


Assuntos
Neoplasias da Mama/diagnóstico , Neoplasias da Mama/patologia , Pós-Menopausa/fisiologia , Viés , Detecção Precoce de Câncer , Feminino , Seguimentos , Humanos , Invasividade Neoplásica , Probabilidade , Fatores de Risco , Tamanho da Amostra , Reino Unido
16.
Value Health ; 23(9): 1246-1255, 2020 09.
Artigo em Inglês | MEDLINE | ID: mdl-32940243

RESUMO

OBJECTIVES: Low uptake of cancer screening services is a global concern. Our aim was to understand factors that influence the screening decision, including screening and treatment subsidies and a gain-frame message designed to present screening as a win-win. METHODS: We analyzed preferences for mammography and Pap smear among women in Singapore by means of discrete choice experiments while randomly exposing half of respondents to a gain-framed public health message promoting the benefits of screening. RESULTS: Results showed that the message did not influence stated uptake, and given the levels shown, respondents were influenced more by treatment attributes, including effectiveness and out-of-pocket cost should they test positive, than by screening attributes, including the offer of a monetary incentive for screening. Respondents also underestimated the survival chances of screen-detected breast and cervical cancers. CONCLUSIONS: Combined, these findings suggest that correcting misconceptions about screen-detected cancer prognosis or providing greater financial protection for those who test positive could be more effective and more cost-effective than subsidizing screening directly in increasing screening uptakes.


Assuntos
Detecção Precoce de Câncer/psicologia , Mamografia/psicologia , Programas de Rastreamento/psicologia , Teste de Papanicolaou/psicologia , Adulto , Neoplasias da Mama/diagnóstico , Comportamento de Escolha , Detecção Precoce de Câncer/economia , Detecção Precoce de Câncer/métodos , Feminino , Conhecimentos, Atitudes e Prática em Saúde , Necessidades e Demandas de Serviços de Saúde , Humanos , Mamografia/economia , Programas de Rastreamento/economia , Programas de Rastreamento/métodos , Pessoa de Meia-Idade , Teste de Papanicolaou/economia , Singapura , Inquéritos e Questionários , Neoplasias do Colo do Útero/diagnóstico
17.
BMC Bioinformatics ; 21(1): 401, 2020 Sep 10.
Artigo em Inglês | MEDLINE | ID: mdl-32912137

RESUMO

BACKGROUND: As an important non-coding RNA, microRNA (miRNA) plays a significant role in a series of life processes and is closely associated with a variety of Human diseases. Hence, identification of potential miRNA-disease associations can make great contributions to the research and treatment of Human diseases. However, to our knowledge, many existing computational methods only utilize the single type of known association information between miRNAs and diseases to predict their potential associations, without focusing on their interactions or associations with other types of molecules. RESULTS: In this paper, we propose a network embedding-based method for predicting miRNA-disease associations by preserving behavior and attribute information. Firstly, a heterogeneous network is constructed by integrating known associations among miRNA, protein and disease, and the network representation method Learning Graph Representations with Global Structural Information (GraRep) is implemented to learn the behavior information of miRNAs and diseases in the network. Then, the behavior information of miRNAs and diseases is combined with the attribute information of them to represent miRNA-disease association pairs. Finally, the prediction model is established based on the Random Forest algorithm. Under the five-fold cross validation, the proposed NEMPD model obtained average 85.41% prediction accuracy with 80.96% sensitivity at the AUC of 91.58%. Furthermore, the performance of NEMPD is also validated by the case studies. Among the top 50 predicted disease-related miRNAs, 48 (breast neoplasms), 47 (colon neoplasms), 47 (lung neoplasms) were confirmed by two other databases. CONCLUSIONS: The proposed NEMPD model has a good performance in predicting the potential associations between miRNAs and diseases, and has great potency in the field of miRNA-disease association prediction in the future.


Assuntos
Neoplasias da Mama/diagnóstico , Neoplasias do Colo/diagnóstico , Biologia Computacional/métodos , Neoplasias Pulmonares/diagnóstico , MicroRNAs/metabolismo , Algoritmos , Área Sob a Curva , Neoplasias da Mama/genética , Neoplasias do Colo/genética , Feminino , Humanos , Neoplasias Pulmonares/genética , MicroRNAs/genética , Curva ROC
19.
J Am Coll Surg ; 231(4): 434-447.e2, 2020 10.
Artigo em Inglês | MEDLINE | ID: mdl-32771654

RESUMO

BACKGROUND: During the COVID-19 pandemic, surgical delays have been common for patients with ductal carcinoma in situ (DCIS) and early-stage estrogen receptor-positive (ER+) breast cancer, often in favor of neoadjuvant endocrine therapy (NET). To understand possible ramifications of these delays, we examined the association between time to operation and pathologic staging and overall survival (OS). STUDY DESIGN: Patients with DCIS or ER+ cT1-2N0 breast cancer treated from 2010 through 2016 were identified in the National Cancer Database. Time to operation was recorded. Factors associated with pathologic upstaging were examined using logistic regression analyses. Cox proportional hazard models were used to analyze OS. Analyses were stratified by disease stage and initial treatment strategy. RESULTS: There were 378,839 patients identified. Among those undergoing primary surgical procedure, time to operation was within 120 days in > 98% in all groups. Among cT1-2N0 patients selected for NET, operations were performed within 120 days in 59.6% of cT1N0 and 30.9% of cT2N0 patients. Increased time to operation was associated with increased odds of pathologic upstaging in DCIS patients (ER+: 60 to 120 days: odds ratio 1.15; 95% CI, 1.08 to 1.22; more than 120 days: odds ratio 1.44; 95% CI, 1.24 to 1.68; ER-: 60 to 120 days: NS; more than 120 days: odds ratio 1.36; 95% CI, 1.01 to 1.82; 60 days or less: reference), but not in patients with invasive cancer, irrespective of initial treatment strategy. No difference in OS was seen by time to operation in DCIS or NET patients. CONCLUSIONS: Increased time to operation was associated with a small increase in pathologic upstaging in DCIS patients, but did not impact OS. In patients with cT1-2N0 disease, NET use did not impact stage or OS, supporting the safety of delay strategies in ER+ breast cancer patients during the pandemic.


Assuntos
Betacoronavirus , Neoplasias da Mama/diagnóstico , Carcinoma Ductal de Mama/diagnóstico , Infecções por Coronavirus/epidemiologia , Mastectomia/métodos , Estadiamento de Neoplasias , Pneumonia Viral/epidemiologia , Receptores Estrogênicos/metabolismo , Adolescente , Adulto , Idoso , Idoso de 80 Anos ou mais , Biomarcadores Tumorais/metabolismo , Neoplasias da Mama/epidemiologia , Neoplasias da Mama/cirurgia , Carcinoma Ductal de Mama/epidemiologia , Carcinoma Ductal de Mama/cirurgia , Comorbidade , Feminino , Seguimentos , Humanos , Pessoa de Meia-Idade , Pandemias , Prognóstico , Estudos Retrospectivos , Fatores de Risco , Taxa de Sobrevida/tendências , Fatores de Tempo , Tempo para o Tratamento , Estados Unidos/epidemiologia , Adulto Jovem
20.
BMC Bioinformatics ; 21(1): 374, 2020 Aug 28.
Artigo em Inglês | MEDLINE | ID: mdl-32859146

RESUMO

BACKGROUND: In this era of data science-driven bioinformatics, machine learning research has focused on feature selection as users want more interpretation and post-hoc analyses for biomarker detection. However, when there are more features (i.e., transcripts) than samples (i.e., mice or human samples) in a study, it poses major statistical challenges in biomarker detection tasks as traditional statistical techniques are underpowered in high dimension. Second and third order interactions of these features pose a substantial combinatoric dimensional challenge. In computational biology, random forest (RF) classifiers are widely used due to their flexibility, powerful performance, their ability to rank features, and their robustness to the "P > > N" high-dimensional limitation that many matrix regression algorithms face. We propose binomialRF, a feature selection technique in RFs that provides an alternative interpretation for features using a correlated binomial distribution and scales efficiently to analyze multiway interactions. RESULTS: In both simulations and validation studies using datasets from the TCGA and UCI repositories, binomialRF showed computational gains (up to 5 to 300 times faster) while maintaining competitive variable precision and recall in identifying biomarkers' main effects and interactions. In two clinical studies, the binomialRF algorithm prioritizes previously-published relevant pathological molecular mechanisms (features) with high classification precision and recall using features alone, as well as with their statistical interactions alone. CONCLUSION: binomialRF extends upon previous methods for identifying interpretable features in RFs and brings them together under a correlated binomial distribution to create an efficient hypothesis testing algorithm that identifies biomarkers' main effects and interactions. Preliminary results in simulations demonstrate computational gains while retaining competitive model selection and classification accuracies. Future work will extend this framework to incorporate ontologies that provide pathway-level feature selection from gene expression input data.


Assuntos
Algoritmos , Biomarcadores/metabolismo , Biomarcadores Tumorais/metabolismo , Neoplasias da Mama/diagnóstico , Biologia Computacional/métodos , Feminino , Humanos , Neoplasias Renais/diagnóstico
SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA