Biostatistics Commons^™

Statistical Handling Of Medical Data - An Ethical Perspective, Ajay Kumar Bansal Dr

COBRA Preprint Series

Medical Science is a delicate subject and the clinical data generated from the medical trials must be reliable and of good quality. Not only the quality of generated data is important, but the management is also crucial and is to be handled very carefully. In this paper, the ethical aspect of statistical handling of such data is discussed.

Every profession has some set of norms to follow to achieve its objectives. These norms are called professional ethics which shows the essence of human behaviour. Same way, the field of medical research is expected to follow ethical norms, to obtain reliable …

Semi-Parametric Estimation And Inference For The Mean Outcome Of The Single Time-Point Intervention In A Causally Connected Population, Oleg Sofrygin, Mark J. Van Der Laan

U.C. Berkeley Division of Biostatistics Working Paper Series

We study the framework for semi-parametric estimation and statistical inference for the sample average treatment-specific mean effects in observational settings where data are collected on a single network of connected units (e.g., in the presence of interference or spillover). Despite recent advances, many of the current statistical methods rely on estimation techniques that assume a particular parametric model for the outcome, even though some of the most important statistical assumptions required by these models are most likely violated in the observational network settings, often resulting in invalid and anti-conservative statistical inference. In this manuscript, we rely on the recent methodological …

Statistical Estimation Of White Matter Microstructure From Conventional Mri, Leah Suttner, Amanda Mejia, Blake Dewey, Pascal Sati, Daniel S. Reich, Russell T. Shinohara

UPenn Biostatistics Working Papers

Diffusion tensor imaging (DTI) has become the predominant modality for studying white matter integrity in multiple sclerosis (MS) and other neurological disorders. Unfortunately, the use of DTI-based biomarkers in large multi-center studies is hindered by systematic biases that confound the study of disease-related changes. Furthermore, the site-to-site variability in multi-center studies is significantly higher for DTI than that for conventional MRI-based markers. In our study, we apply the Quantitative MR Estimation Employing Normalization (QuEEN) model to estimate the four DTI measures: MD, FA, RD, and AD. QuEEN uses a voxel-wise generalized additive regression model to relate the normalized intensities of …

Correction Of Verication Bias Using Log-Linear Models For A Single Binaryscale Diagnostic Tests, Haresh Rochani, Hani M. Samawi, Robert L. Vogel, Jingjing Yin

Biostatistics: Faculty Publications

In diagnostic medicine, the test that determines the true disease status without an error is referred to as the gold standard. Even when a gold standard exists, it is extremely difficult to verify each patient due to the issues of costeffectiveness and invasive nature of the procedures. In practice some of the patients with test results are not selected for verification of the disease status which results in verification bias for diagnostic tests. The ability of the diagnostic test to correctly identify the patients with and without the disease can be evaluated by measures such as sensitivity, specificity and predictive …

A Generally Efficient Targeted Minimum Loss Based Estimator, Mark J. Van Der Laan

U.C. Berkeley Division of Biostatistics Working Paper Series

Suppose we observe n independent and identically distributed observations of a finite dimensional bounded random variable. This article is concerned with the construction of an efficient targeted minimum loss-based estimator (TMLE) of a pathwise differentiable target parameter based on a realistic statistical model.

The canonical gradient of the target parameter at a particular data distribution will depend on the data distribution through an infinite dimensional nuisance parameter which can be defined as the minimizer of the expectation of a loss function (e.g., log-likelihood loss). For many models and target parameters the nuisance parameter can be split up in two components, …

Estimated Probability Of Becoming Alcohol Dependent: Extending A Multiparametric Approach, Olga A. Vsevolozhskaya, James C. Anthony

Biostatistics Presentations

Background: United States (US) epidemiological studies suggest that for every 5-8 who start drinking alcoholic beverages, at least one drinker will develop an alcohol dependence (AD) syndrome within the first 10 years after onset of drinking (Lopez-Quintero et al., 2011; Wagner & Anthony, 2002). Recently, we described a multiparametric functional analysis approach for new research to estimate these transition probabilities with a one-dimensional function (1D; Vsevolozhskaya & Anthony, 2015). Here, we demonstrate extension of this analysis to two-dimensional (2D) functions that combine information about number of recent drinking days and number of drinks on the typical drinking day.

Methods: Data …

Inequality In Treatment Benefits: Can We Determine If A New Treatment Benefits The Many Or The Few?, Emily Huang, Ethan Fang, Daniel Hanley, Michael Rosenblum

Johns Hopkins University, Dept. of Biostatistics Working Papers

The primary analysis in many randomized controlled trials focuses on the average treatment effect and does not address whether treatment benefits are widespread or limited to a select few. This problem affects many disease areas, since it stems from how randomized trials, often the gold standard for evaluating treatments, are designed and analyzed. Our goal is to learn about the fraction who benefit from a treatment, based on randomized trial data. We consider the case where the outcome is ordinal, with binary outcomes as a special case. In general, the fraction who benefit is a non-identifiable parameter, and the best …

Factors Impacting Transgender Patients’ Discomfort With Their Family Physicians: A Respondent-Driven Sampling Survey, Greta R. Bauer, Xuchen Zong, Ayden I. Scheim, Rebecca Hammond, Amardeep Thind

Epidemiology and Biostatistics Publications

BACKGROUND: Representing approximately 0.5% of the population, transgender (trans) persons in Canada depend on family physicians for both general and transition-related care. However, physicians receive little to no training on this patient population, and trans patients are often profoundly uncomfortable and may avoid health care. This study examined factors associated with patient discomfort discussing trans health issues with a family physician in Ontario, Canada.

METHODS: 433 trans people age 16 and over were surveyed using respondent-driven sampling for the Trans PULSE Project; 356 had a family physician. Weighted logistic regression models were fit to produce prevalence risk ratios (PRRs) via …

Pesticide Exposure, Behavior Of Farmer, And Activity Of Cholinesterase Enzyme In Blood Of Fertile Women Farmers, Deni Abdul Rahman, Zakianis Zakianis, Laila Fitria

Kesmas

Petani perempuan usia subur berisiko mengalami penurunan aktivitas kolinesterase akibat pajanan pestisida. Penelitian ini bertujuan untuk menganalisis hubungan antara pajanan pestisida dan perilaku pemajan terhadap aktivitas kolinesterase petani perempuan usia subur di Desa Kedunguter. Penelitian ini menggunakan desain potong lintang pada 94 petani perempuan usia subur tahun 2015. Pengumpulan data dilakukan secara observasi, wawancara, dan uji kolinesterase. Analisis data menggunakan uji kai kuadrat dan hasil analisis menunjukkan hubungan signifikan antara jenis pestisida, waktu kerja, penggunaan sarung tangan, perilaku mencuci tangan terhadap aktivitas kolinesterase petani perempuan usia subur. Hasil analisis pada penelitian ini menunjukkan bahwa variabel waktu kerja memiliki nilai odds …

The Use Of Earthworms And Household Organic Waste Composting Length Of Time, Abdul Khair, Lucky Herawati, Noraida Noraida, Munawar Raharja

Kesmas

Pengomposan dengan cacing tanah merupakan proses pembuatan kompos dengan melibatkan organisme makro cacing tanah. Kerja sama antara cacing tanah dengan mikroorganisme dapat memberi dampak pada proses penguraian yang dilakukan oleh mikroorganisme tersebut dibantu dengan keberadaan cacing tanah. Oleh karena bahan-bahan yang akan diurai oleh mikroorganisme telah diurai lebih dahulu oleh cacing, maka kerja mikroorganisme lebih efektif dan lebih cepat. Penelitian ini bertujuan untuk mengetahui pengaruh penggunaan cacing tanah terhadap lama waktu pengomposan sampah organik dari rumah tangga dengan menggunakan desain penelitian eksperimen. Objek penelitian adalah seluruh sampah organik dari rumah tangga yang diambil secara acak pada satu rumah tangga. Variabel …

Riwayat Berat Badan Lahir Dengan Kejadian Stunting Pada Anak Usia Bawah Dua Tahun, Atikah Rahayu, Fahrini Yulidasari, Andini Octaviana Putri, Fauzie Rahman

Kesmas

Kabupaten Hulu Sungai Utara masih dihadapkan dengan permasalahan gizi pada anak bawah dua tahun (baduta). Salah satu masalah gizi hingga saat ini adalah stunting. Anak dengan riwayat berat badan lahir rendah (BBLR) merupakan salah satu faktor yang potensial memengaruhi pertumbuhan anak. Penelitian ini bertujuan untuk mengkaji risiko riwayat berat badan lahir dengan kejadian stunting pada anak baduta. Desain penelitian adalah potong lintang. Populasi penelitian ini merupakan ibu-ibu yang memiliki anak baduta dan besar sampel sejumlah 117 terdiri dari anak baduta. Pelaksanaan penelitian dilakukan selama tiga bulan pada bulan September – November 2014. Kategori BBLR jika riwayat berat badan lahir < 2.500 gram. Analisis data bivariat menggunakan uji kai kuadrat dan data multivariat menggunakan uji regresi logistik. Hasil analisis bivariat menunjukkan bahwa terdapat hubungan yang signifikan antara riwayat status BBLR (nilai p = 0,015) dengan stunting pada anak baduta. Berdasarkan hasil analisis multivariat, diperoleh bahwa BBLR merupakan faktor risiko yang paling dominan berhubungan dengan kejadian stunting. Anak dengan BBLR memiliki risiko 5,87 kali untuk mengalami stunting. Riwayat BBLR memiliki peranan penting dalam kejadian stunting anak baduta di wilayah Puskesmas Sungai Karias, Hulu Sungai Utara. North Hulu Sungai District is still facing nutrition problems among children under two years old. One of nutrition problems up to now is stunting. Child with low birthweight (LBW) record is one of potential factors influencing the growth of a child. This study aimed to assess any risk of LBW records with stunting incidence among children under two years old. This study used cross-sectional design. The population was mothers having children under two years old and samples amounted to 117 consisted of children under two years old. This study was conducted within three months on September – November 2014. Category of LBW was if birth weight records < 2,500 gram. Bivariate data analysis used chi-square test and multivariate data analysis used logistic regression test. The result of bivariate analysis showed a significant relation between LBW status records (p value = 0.015) with stunting incidence among children under two years old. Based on the result of multivariate analysis, LBW was the most dominating risk factor related to stunting incidence. Children with LBW had 5.87 times risk of suffering from stunting. LBW records take an important role in stunting incidence among children under two years old around Sungai Karias Primary Health Care area in North Hulu Sungai.

Penemuan Kasus Infeksi Kusta Subklinis Pada Anak Melalui Deteksi Kadar Antibodi (Igm) Anti Pgl-1, Arif Sujagat, Fardhiasih Dwi Astuti, Eva Muslimawati Saputri, Annisa Sani, Agil Dwi Prasetya

Kesmas

]Kusta merupakan penyakit menular kronis yang disebabkan Mycobacterium leprae. Indonesia menempati peringkat ketiga penyumbang kasus kusta terbanyak di dunia pada tahun 2012 – 2014. Kota Surabaya merupakan daerah endemis kusta. Penelitian ini bertujuan untuk deteksi dini kusta subklinis berdasarkan pengukuran kadar antibodi (IgM) anti PGL-1 dan mengetahui faktor-faktor yang berhubungan dengan kusta subklinis pada anak di Kota Surabaya. Penelitian analitik menggunakan desain potong lintang pada bulan April - Mei 2015. Populasi dilakukan penelitian adalah keluarga yang memiliki anggota penderita kusta. Sampel dipilih dengan teknik simple random sampling dan sebanyak 30 kepala keluarga terpilih melalui rapid school and village survey pada …

Pre-Eklampsia Berat Dan Kematian Ibu, Nova Muhani, Besral Besral

Kesmas

Pre-eklampsia berat, salah satu penyebab utama kematian ibu di Indonesia dan di RSUD Dr. H. Abdul Moeloek Lampung, merupakan penyebab kematian ibu nomor satu (47,25%). Penelitian ini bertujuan untuk mengetahui hubungan prediktor pre-eklampsi berat (PEB) yang dinilai dari tekanan darah sistolik, tekanan darah diastolik, proteiunuria, eklampsia, sindrom hemolysis, elevated liver enzymes, low platelets count (HELLP) dengan kematian ibu di RSUD Dr. H. Abdul Moeloek. Penelitian ini menggunakan desain kasus kontrol dengan jumlah sampel 60 kasus dan 120 kontrol. Data diolah dari rekam medis rumah sakit selama periode lima tahun (2010 – 2014). Hasil penelitian ini memperlihatkan bahwa sindrom HELLP memiliki …

Peran Tenaga Kesehatan Dan Keluarga Dalam Kehamilan Usia Remaja, Mery Ramadani, Dien Gusta Anggraini Nursal, Livia Ramli

Kesmas

Sebanyak 10,3% kematian tidak langsung pada ibu disebabkan kehamilan usia remaja (< 20 tahun). Di Kabupaten Tanah Datar, masih terjadi peningkatan kehamilan usia remaja dalam tiga tahun terakhir. Penelitian ini bertujuan untuk mengetahui peran tenaga kesehatan, dukungan keluarga, dan pengetahuan remaja dengan kehamilan usia remaja di wilayah kerja Puskesmas Singgalang, Kabupaten Tanah Datar tahun 2014. Penelitian dilaksanakan pada bulan Mei – Juni 2014 menggunakan desain potong lintang. Populasi adalah seluruh remaja putri berusia < 20 tahun yang telah menikah berjumlah 215 orang. Sampel berjumlah 68 orang dan pengambilan sampel dilakukan secara proporsional di delapan jorong/desa. Data dikumpulkan melalui wawancara menggunakan kuesioner. Kemudian, analisis bivariat dilakukan dengan uji kai kuadrat dan analisis multivariat dengan uji regresi logistik ganda. Hasil penelitian mendapatkan sebanyak 55,9% responden hamil di usia remaja. Sebanyak 52,9% responden kurang merasakan peran dari tenaga kesehatan, 66,2% kurang mendapat dukungan keluarga, dan 58,8% memiliki pengetahuan rendah. Didapatkan hubungan peran tenaga kesehatan (nilai p = 0,032), dukungan keluarga (nilai p = 0,025), dan tingkat pengetahuan (nilai p = 0,002) dengan kehamilan usia remaja. Dapat disimpulkan bahwa tenaga kesehatan, keluarga dan tingkat pengetahuan berperan dalam kehamilan remaja. Tenaga kesehatan perlu memberikan penyuluhan mengenai risiko kehamilan remaja kepada remaja serta keluarga. Worth 10.3% of indirect maternal death is due to teenage pregnancy (< 20 years old). In Tanah Datar District, the increase of teenage pregnancy has occured in the last three years. This study aimed to determine health worker's role, family's support and teenagers' knowledge with teenage pregnancy in work area of Singgalang Primary Health Care, Tanah Datar District in 2014. This study was conducted on May – June 2014 using cross-sectional design. Population was 215 married teenage girls < 20 years old. A total of sample was 68 selected proportionally in eight villages. Data were collected through interview using questionnaire. Then bivariate analysis was conducted using chi-square test and multivariate analysis using multiple logistic regression test. Results of study found 55.9% of respondents were pregnant in teen age. Respondents worth 52.9% got less health worker’s role, 66.2% got less family’s support and 58.8% had low level of knowledge. There was a relation found between health worker’s role (p value = 0.032), family’s support (p value = 0.025) and knowledge level (p value = 0.002) with teenage pregnancy. In conclusion, health workers, family and knowledge level play a role in teenage pregnancy. Health workers need to provide counseling concerning teenage pregnancy risks for both teenagers and families.

Pengaruh Musik Keroncong Selama Pelaksanaan Kangaroo Mother Care Terhadap Respons Fisiologis Dan Lama Rawat Bayi Dengan Berat Badan Lahir Rendah, Anita Rahmawati, Endah Marianingsih Theresia, Yuliasti Eka Purnamaningrum

Kesmas

Kangaroo mother care (KMC) merupakan metode merawat bayi berat badan lahir rendah (BBLR). Beberapa intervensi perawatan di neonatal intensive care unit seperti pijat bayi, KMC, dan mendengarkan musik bermanfaat untuk pertumbuhan bayi berupa respons fisiologis BBLR dan mengurangi lama rawat. Penelitian ini bertujuan untuk mengetahui manfaat musik keroncong terhadap respons BBLR selama KMC dan lama rawat. Rancangan penelitian adalah quasi eksperimental dengan pretest dan posttest dengan desain grup kontrol. Pada Juli - September 2014 populasi penelitian adalah ibu dan bayi BBLR yang melaksanakan KMC. Pengambilan sampel dengan purposive sampling sebanyak 60 bayi. Kriteria inklusi bayi BBLR yang ditetapkan adalah berat …

Disability Prevention Programs For Older People: Factors Associated With Medical And Nursing Care Costs, Takaaki Ikeda, Susumu Yanagawa, Kenichi Murakami, Toshiyuki Inoue, Takamasa Tsurumi

Kesmas

Penelitian ini bertujuan untuk mengklarifikasi faktor-faktor yang terkait dengan biaya medis dan perawatan lanjut usia (lansia) yang tinggal di panti sosial serta menyarankan program pencegahan disabilitas yang efektif. Total partisipan dalam penelitian ini adalah 83 orang (29 laki-laki dan 54 perempuan; usia rata-rata 81,2 ± 6,3 tahun) pada 1 November – 28 Desember 2014. Penelitian ini membandingkan biaya medis dan perawatan rata-rata per bulan dengan rata-rata nasional bagi yang berusia ≥ 65 tahun. Uji regresi logistik dilakukan untuk memeriksa keterkaitannya dengan biaya medis dan perawatan. Mereka yang pergi ke luar rumah ≥ 3 kali seminggu tiga kali lebih kecil kemungkinannya …

Monitoring For Adverse Events Post Marketing Approval Of Drugs, Karl E. Peace, Macaulay Okwuokenye

Biostatistics: Faculty Publications

This brief communication provides information to those developing monitoring plans for serious adverse events (SAE’s) following regulatory approval of a new drug. In addition, we (1) illustrate how many patients would need to be treated in order to have high confidence of seeing at least 1 pre-specified SAE, (2) show that absence of proof of a SAE is not proof of absence of that SAE, and (3) identify statistical methodology that could be used for formal statistical monitoring of SAE’s.

Meta-Analysis Of Genome-Wide Association Studies With Correlated Individuals: Application To The Hispanic Community Health Study/Study Of Latinos (Hchs/Sol), Tamar Sofer, John R. Shaffer, Misa Graff, Qibin Qi, Adrienne M. Stilp, Stephanie M. Gogarten, Kari E. North, Carmen R. Isasi, Cathy C. Laurie, Adam A. Szpiro

UW Biostatistics Working Paper Series

Investigators often meta-analyze multiple genome-wide association studies (GWASs) to increase the power to detect associations of single nucleotide polymorphisms (SNPs) with a trait. Meta-analysis is also performed within a single cohort that is stratified by, e.g., sex or ancestry group. Having correlated individuals among the strata may complicate meta-analyses, limit power, and inflate Type 1 error. For example, in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL), sources of correlation include genetic relatedness, shared household, and shared community. We propose a novel mixed-effect model for meta-analysis, “MetaCor", which accounts for correlation between stratum-specific effect estimates. Simulations show that MetaCor controls …

Nested Partially-Latent, Class Models For Dependent Binary Data, Estimating Disease Etiology, Zhenke Wu, Maria Deloria-Knoll, Scott L. Zeger

Johns Hopkins University, Dept. of Biostatistics Working Papers

The Pneumonia Etiology Research for Child Health (PERCH) study seeks to use modern measurement technology to infer the causes of pneumonia for which gold-standard evidence is unavailable. The paper describes a latent variable model designed to infer from case-control data the etiology distribution for the population of cases, and for an individual case given his or her measurements. We assume each observation is drawn from a mixture model for which each component represents one cause or disease class. The model addresses a major limitation of the traditional latent class approach by taking account of residual dependence among multivariate binary outcome …

Identification And Survival Outcomes Of A Cohort Of Patients With Cancer Of Unknown Primary In Ontario, Canada., Chong S Kim, Malek B Hannouf, Sisira Sarma, George B Rodrigues, Peter K Rogan, Salaheddin M Mahmud, Eric Winquist, Muriel Brackstone, Gregory S Zaric

Epidemiology and Biostatistics Publications

BACKGROUND: Cancer of unknown primary origin (CUP) is defined by the presence of pathologically identified metastatic disease without clinical or radiological evidence of a primary tumour. Our objective was to identify incident cases of CUP in Ontario, Canada, and determine the influence of histology and sites of metastases on overall survival (OS).

MATERIAL AND METHODS: We used the Ontario Cancer Registry (OCR) and the Same-Day Surgery and Discharge Abstract Database (SDS/DAD) to identify patients diagnosed with CUP in Ontario between 1 January 2000, and 31 December 2005. Patient diagnostic information, including histology and survival data, was obtained from the OCR. …

Abcc9/Sur2 In The Brain: Implications For Hippocampal Sclerosis Of Aging And A Potential Therapeutic Target, Peter T. Nelson, Gregory A. Jicha, Wang-Xia Wang, Eseosa T. Ighodaro, Sergey C. Artiushin, Colin G. Nichols, David W. Fardo

Sanders-Brown Center on Aging Faculty Publications

The ABCC9 gene and its polypeptide product, SUR2, are increasingly implicated in human neurologic disease, including prevalent diseases of the aged brain. SUR2 proteins are a component of the ATP-sensitive potassium (“K _ATP ”) channel, a metabolic sensor for stress and/or hypoxia that has been shown to change in aging. The K _ATP channel also helps regulate the neurovascular unit. Most brain cell types express SUR2, including neurons, astrocytes, oligodendrocytes, microglia, vascular smooth muscle, pericytes, and endothelial cells. Thus it is not surprising that ABCC9 gene variants are associated with risk for human brain diseases. For example, Cantu syndrome is …

Removing Inter-Subject Technical Variability In Magnetic Resonance Imaging Studies, Jean-Philippe Fortin, Elizabeth M. Sweeney, John Muschelli, Ciprian M. Crainiceanu, Russell T. Shinohara, Alzheimer’S Disease Neuroimaging Initiative

UPenn Biostatistics Working Papers

Magnetic resonance imaging (MRI) intensities are acquired in arbitrary units, making scans non-comparable across sites and between subjects. Intensity normalization is a first step for the improvement of comparability of the images across subjects. However, we show that unwanted inter-scan variability associated with imaging site, scanner effect and other technical artifacts is still present after standard intensity normalization in large multi-site neuroimaging studies. We propose RAVEL (Removal of Artificial Voxel Effect by Linear regression), a tool to remove residual technical variability after intensity normalization. As proposed by SVA and RUV [Leek and Storey, 2007, …

Functional Linear Models Extensions Uncover Pleiotropic Effects Of Chronic Pain Phenotypes, Dmitri V. Zaykin, L. Qing, G. D. Slade, R. Dubner, R. B. Fillingim, J. D. Greenspan, R. Ohrbach, W. Maixner, L. B. Diatchenko, Olga A. Vsevolozhskaya

Biostatistics Presentations

Growing scientific evidence suggests that intricate interactions of genetic risk factors with environmental exposures play a major role in the development of chronic pain conditions. In studies of relative contribution of an individual’s genetic composition to the perception of pain, the general characteristics of pain sensitivity are typically measured by a wide range of different, yet possibly etiologically related pain phenotypes. Testing each of these pain-perception traits individually is subject to problems of multiple testing and low statistical power. Furthermore, pain-related traits may share common etiology and comprise binary, categorical, and quantitative measurements. In the current study, we propose a …

Validation Of A New Predictive Risk Model: Measuring The Impact Of The Major Modifiable Risks Of Death For Patients And Populations, Stephen S. Lim, Emily Carnahan, Eugene C. Nelson, Catherine W. Gillespie, Ali H. Mokdad, Christopher J. L. Murray, Elliott S. Fisher

Dartmouth Scholarship

Background: Modifiable risks account for a large fraction of disease and death, but clinicians and patients lack tools to identify high risk populations or compare the possible benefit of different interventions.

Methods: We used data on the distribution of exposure to 12 major behavioral and biometric risk factors inthe US population, mortality rates by cause, and estimates of the proportional hazards of risk factor exposure from published systematic reviews to develop a risk prediction model that estimates an adult's 10 year mortality risk compared to a population with optimum risk factors. We compared predicted risk to observed mortality in 8,241 …

Patient-Specific Variations In Biomarkers Across Gingivitis And Periodontitis, Radhakrishnan Nagarajan, Craig S. Miller, Dolph Dawson, Mohanad Al-Sabbagh, J. L. Ebersole

Biostatistics Faculty Publications

This study investigates the use of saliva, as an emerging diagnostic fluid in conjunction with classification techniques to discern biological heterogeneity in clinically labelled gingivitis and periodontitis subjects (80 subjects; 40/group) A battery of classification techniques were investigated as traditional single classifier systems as well as within a novel selective voting ensemble classification approach (SVA) framework. Unlike traditional single classifiers, SVA is shown to reveal patient-specific variations within disease groups, which may be important for identifying proclivity to disease progression or disease stability. Salivary expression profiles of IL-1ß, IL-6, MMP-8, and MIP-1α from 80 patients were analyzed using four classification …

Addressing Confounding In Predictive Models With An Application To Neuroimaging, Kristin A. Linn, Bilwaj Gaonkar, Jimit Doshi, Christos Davatzikos, Russell T. Shinohara

UPenn Biostatistics Working Papers

Understanding structural changes in the brain that are caused by a particular disease is a major goal of neuroimaging research. Multivariate pattern analysis (MVPA) comprises a collection of tools that can be used to understand complex disease effects across the brain. We discuss several important issues that must be considered when analyzing data from neuroimaging studies using MVPA. In particular, we focus on the consequences of confounding by non-imaging variables such as age and sex on the results of MVPA. After reviewing current practice to address confounding in neuroimaging studies, we propose an alternative approach based on inverse probability weighting. …

Control-Group Feature Normalization For Multivariate Pattern Analysis Using The Support Vector Machine, Kristin A. Linn, Bilwaj Gaonkar, Jimit Doshi, Christos Davatzikos, Russell T. Shinohara

UPenn Biostatistics Working Papers

Normalization of feature vector values is a common practice in machine learning. Generally, each feature value is standardized to the unit hypercube or by normalizing to zero mean and unit variance. Classification decisions based on support vector machines (SVMs) or by other methods are sensitive to the specific normalization used on the features. In the context of multivariate pattern analysis using neuroimaging data, standardization effectively up- and down-weights features based on their individual variability. Since the standard approach uses the entire data set to guide the normalization it utilizes the total variability of these features. This total variation is inevitably …

A Simple Method To Estimate The Time-Dependent Roc Curve Under Right Censoring, Liang Li, Bo Hu, Tom Greene

COBRA Preprint Series

The time-dependent Receiver Operating Characteristic (ROC) curve is often used to study the diagnostic accuracy of a single continuous biomarker, measured at baseline, on the onset of a disease condition when the disease onset may occur at different times during the follow-up and hence may be right censored. Due to censoring, the true disease onset status prior to the pre-specified time horizon may be unknown on some patients, which causes difficulty in calculating the time-dependent sensitivity and specificity. We study a simple method that adjusts for censoring by weighting the censored data by the conditional probability of disease onset prior …

Bayesian Inference On Longitudinal Semi-Continuous Substance Abuse/Dependence Symptoms Data, Dongyuan Xing

USF Tampa Graduate Theses and Dissertations

Substance use data such as alcohol drinking often contain a high proportion of zeros. In studies examining the alcohol consumption in college students, for instance, many students may not drink in the studied period, resulting in a number of zeros. Zero-inflated continuous data, also called semi continuous data, typically consist of a mixture of a degenerate distribution at the origin (zero) and a right-skewed, continuous distribution for the positive values. Ignoring the extreme non-normality in semi-continuous data may lead to substantially biased estimates and inference. Longitudinal or repeated measures of semi-continuous data present special challenges in statistical inference because of …

On Varieties Of Doubly Robust Estimators Under Missing Not At Random With An Ancillary Variable, Wang Miao, Eric Tchetgen Tchetgen

Harvard University Biostatistics Working Paper Series

No abstract provided.