Model selection for assessing the effects of doxorubicin on triple-negative breast cancer cell lines.

Doxorubicin is a chemotherapy widely used to treat several types of cancer, including triple-negative breast cancer. In this work, we use a Bayesian framework to rigorously assess the ability of ten different mathematical models to describe the dynamics of four TNBC cell lines (SUM-149PT, MDA-MB-231, MDA-MB-453, and MDA-MB-468) in response to treatment with doxorubicin at concentrations ranging from 10 to 2500 nM. Each cell line was plated and serially imaged via fluorescence microscopy for 30 days following 6, 12, or 24 h of in vitro drug exposure. We use the resulting data sets to estimate the parameters of the ten pharmacodynamic models using a Bayesian approach, which accounts for uncertainties in the models, parameters, and observational data. The ten candidate models describe the growth patterns and degree of response to doxorubicin for each cell line by incorporating exponential or logistic tumor growth, and distinct forms of cell death. Cell line and treatment specific model parameters are then estimated from the experimental data for each model. We analyze all competing models using the Bayesian Information Criterion (BIC), and the selection of the best model is made according to the model probabilities (BIC weights). We show that the best model among the candidate set of models depends on the TNBC cell line and the treatment scenario, though, in most cases, there is great uncertainty in choosing the best model. However, we show that the probability of being the best model can be increased by combining treatment data with the same total drug exposure. Our analysis points to the importance of considering multiple models, built on different biological assumptions, to capture the observed variations in tumor growth and treatment response.

A time-resolved experimental-mathematical model for predicting the response of glioma cells to single-dose radiation therapy.

PURPOSE: To develop and validate a mechanism-based, mathematical model that characterizes 9L and C6 glioma cells' temporal response to single-dose radiation therapy in vitro by explicitly incorporating time-dependent biological interactions with radiation. METHODS: We employed time-resolved microscopy to track the confluence of 9L and C6 glioma cells receiving radiation doses of 0, 2, 4, 6, 8, 10, 12, 14 or 16 Gy. DNA repair kinetics are measured by γH2AX expression via flow cytometry. The microscopy data (814 replicates for 9L, 540 replicates for C6 at various seeding densities receiving doses above) were divided into training (75%) and validation (25%) sets. A mechanistic model was developed, and model parameters were calibrated to the training data. The model was then used to predict the temporal dynamics of the validation set given the known initial confluences and doses. The predictions were compared to the corresponding dynamic microscopy data. RESULTS: For 9L, we obtained an average (± standard deviation, SD) Pearson correlation coefficient between the predicted and measured confluence of 0.87 ± 0.16, and an average (±SD) concordance correlation coefficient of 0.72 ± 0.28. For C6, we obtained an average (±SD) Pearson correlation coefficient of 0.90 ± 0.17, and an average (±SD) concordance correlation coefficient of 0.71 ± 0.24. CONCLUSION: The proposed model can effectively predict the temporal development of 9L and C6 glioma cells in response to a range of single-fraction radiation doses. By developing a mechanism-based, mathematical model that can be populated with time-resolved data, we provide an experimental-mathematical framework that allows for quantitative investigation of cells' temporal response to radiation. Our approach provides two key advances: (i) a time-resolved, dynamic death rate with a clear biological interpretation, and (ii) accurate predictions over a wide range of cell seeding densities and radiation doses.

Validation and estimation of spleen volume via computer-assisted segmentation on clinically acquired CT scans.

Purpose: Deep learning is a promising technique for spleen segmentation. Our study aims to validate the reproducibility of deep learning-based spleen volume estimation by performing spleen segmentation on clinically acquired computed tomography (CT) scans from patients with myeloproliferative neoplasms. Approach: As approved by the institutional review board, we obtained 138 de-identified abdominal CT scans. A sum of voxel volume on an expert annotator's segmentations establishes the ground truth (estimation 1). We used our deep convolutional neural network (estimation 2) alongside traditional linear estimations (estimation 3 and 4) to estimate spleen volumes independently. Dice coefficient, Hausdorff distance, R 2 coefficient, Pearson R coefficient, the absolute difference in volume, and the relative difference in volume were calculated for 2 to 4 against the ground truth to compare and assess methods' performances. We re-labeled on scan-rescan on a subset of 40 studies to evaluate method reproducibility. Results: Calculated against the ground truth, the R 2 coefficients for our method (estimation 2) and linear method (estimation 3 and 4) are 0.998, 0.954, and 0.973, respectively. The Pearson R coefficients for the estimations against the ground truth are 0.999, 0.963, and 0.978, respectively (paired t -tests produced p < 0.05 between 2 and 3, and 2 and 4). Conclusion: The deep convolutional neural network algorithm shows excellent potential in rendering more precise spleen volume estimations. Our computer-aided segmentation exhibits reasonable improvements in splenic volume estimation accuracy.

Experimentally-driven mathematical modeling to improve combination targeted and cytotoxic therapy for HER2+ breast cancer.

The goal of this study is to experimentally and computationally investigate combination trastuzumab-paclitaxel therapies and identify potential synergistic effects due to sequencing of the therapies with in vitro imaging and mathematical modeling. Longitudinal alterations in cell confluence are reported for an in vitro model of BT474 HER2+ breast cancer cells following various dosages and timings of paclitaxel and trastuzumab combination regimens. Results of combination drug regimens are evaluated for drug interaction relationships based on order, timing, and quantity of dose of the drugs. Altering the order of treatments, with the same total therapeutic dose, provided significant changes in overall cell confluence (p < 0.001). Two mathematical models are introduced that are constrained by the in vitro data to simulate the tumor cell response to the individual therapies. A collective model merging the two individual drug response models was designed to investigate the potential mechanisms of synergy for paclitaxel-trastuzumab combinations. This collective model shows increased synergy for regimens where trastuzumab is administered prior to paclitaxel and suggests trastuzumab accelerates the cytotoxic effects of paclitaxel. The synergy derived from the model is found to be in agreement with the combination index, where both indicate a spectrum of additive and synergistic interactions between the two drugs dependent on their dose order. The combined in vitro results and development of a mathematical model of drug synergy has potential to evaluate and improve standard-of-care combination therapies in cancer.

Leveraging Mathematical Modeling to Quantify Pharmacokinetic and Pharmacodynamic Pathways: Equivalent Dose Metric.

Treatment response assays are often summarized by sigmoidal functions comparing cell survival at a single timepoint to applied drug concentration. This approach has a limited biophysical basis, thereby reducing the biological insight gained from such analysis. In particular, drug pharmacokinetic and pharmacodynamic (PK/PD) properties are overlooked in developing treatment response assays, and the accompanying summary statistics conflate these processes. Here, we utilize mathematical modeling to decouple and quantify PK/PD pathways. We experimentally modulate specific pathways with small molecule inhibitors and filter the results with mechanistic mathematical models to obtain quantitative measures of those pathways. Specifically, we investigate the response of cells to time-varying doxorubicin treatments, modulating doxorubicin pharmacology with small molecules that inhibit doxorubicin efflux from cells and DNA repair pathways. We highlight the practical utility of this approach through proposal of the "equivalent dose metric." This metric, derived from a mechanistic PK/PD model, provides a biophysically-based measure of drug effect. We define equivalent dose as the functional concentration of drug that is bound to the nucleus following therapy. This metric can be used to quantify drivers of treatment response and potentially guide dosing of combination therapies. We leverage the equivalent dose metric to quantify the specific intracellular effects of these small molecule inhibitors using population-scale measurements, and to compare treatment response in cell lines differing in expression of drug efflux pumps. More generally, this approach can be leveraged to quantify the effects of various pharmaceutical and biologic perturbations on treatment response.

Cancer mortality rates among US and foreign-born individuals: United States 2005-2014.

From 1970 to 2010 the foreign-born population in the United States has rapidly increased from 9.6 to 40.0 million individuals. Historically, differences in cancer rates have been observed between US-born and foreign-born individuals. However, comprehensive and up-to-date data on US cancer rates by birth place is lacking. To compare cancer mortality rates among foreign and US-born individuals, population-based cancer mortality data were obtained from the CDC's National Center for Health Statistics. Utilizing data recorded on death certificates, individuals were categorized as US-born or foreign-born. Annual population estimates were obtained from the American Community Survey. Age-adjusted mortality rates and rate ratios (RRs) for all cancer sites were calculated using SEER*Stat. A total of 5,670,535 deaths from malignant cancers were recorded in the US from 2005 to 2014 and 9% of deaths occurred among foreign-born individuals. Overall, foreign-born individuals had a 31% lower cancer mortality rate when compared to US-born individuals (Rate Ratio (RR): 0.69 (95% CI: 0.68-0.69)), and similar results were observed when stratifying by sex, race/ethnicity, age, and geographic region. However, foreign-born individuals did have significantly elevated cancer mortality rates for seven cancers sites, of which five were infection-related, including: nasopharynx (RR: 2.01), Kaposi Sarcoma (RR: 1.94), stomach (RR: 1.82), gallbladder (RR: 1.47), acute lymphocytic leukemia (RR: 1.27), liver and intrahepatic bile duct (RR: 1.24), and thyroid (RR: 1.22) cancers. Many of these deaths could be avoided through improved access to prevention, screening, and treatment services for immigrant populations in the US or in their country of origin.

Cervical Cancer Death Rates Among U.S.- and Foreign-Born Women: U.S., 2005-2014.

INTRODUCTION: Historically, foreign-born women in the U.S. are less likely to be screened and are more likely to die from cervical cancer when compared with their U.S.-born counterparts. In order to inform prevention efforts and reduce this health disparity, mortality data were obtained from the National Center for Health Statistics to describe cervical cancer mortality among U.S.- and foreign-born women. METHODS: Annual population estimates were obtained from the U.S. Census Bureau's American Community Survey from 2005 to 2014. From 2017 to 2018, age-adjusted mortality rates and rate ratios were calculated by nativity status, race/ethnicity, age, geographic region, and country of birth. RESULTS: From 2005 to 2014, a total of 5,924 deaths from cervical cancer were recorded among the foreign-born population, compared with 33,893 deaths among U.S.-born women. Overall, foreign-born women had a lower cervical cancer mortality rate when compared with the U.S.-born women (rate ratio=0.95, 95% CI=0.92, 0.97). However, older foreign-born women had significantly higher mortality rates compared with U.S.-born women: aged 65-79 years (rate ratio=1.15, 95% CI=1.09, 1.22) and ≥80 years (rate ratio=1.43, 95% CI=1.32, 1.55). Women born in Mexico had significantly elevated rates of cervical cancer mortality (rate ratio=1.35, 95% CI=1.27, 1.42) when compared with U.S.-born women. CONCLUSIONS: Efforts that work to increase cervical cancer screening access and guideline compliance might further reduce the cervical cancer deaths in the U.S., and the excess burden observed among older foreign-born women.

Gastric cancer mortality rates among US and foreign-born persons: United States 2005-2014.

BACKGROUND: Historically, foreign-born individuals in the US have had an elevated risk of dying from gastric cancer when compared to US-born individuals. This is primarily due to factors that occur prior to their immigration to the US, including diet and underlying risk of H. pylori infection. METHODS: National mortality data from 2005 to 2014 were obtained from the CDC's National Center for Health Statistics. Annual population estimates were obtained from the US Census Bureau's American Community Survey for foreign-born and US-born persons. Age-adjusted gastric cancer mortality rates and rate ratios (RR) were calculated stratified by birth place, age, race/ethnicity, and geographic location. RESULTS: From 2005 to 2014, 111,718 deaths from malignant gastric cancer occurred in the US, of which 24,583 (22%) occurred among foreign-born individuals. Overall, foreign-born individuals had higher mortality rates compared with US-born individuals (RR 1.82; 95% CI 1.80, 1.85) and this difference remained after stratifying by sex, age, and geographic location. However, this finding was primarily driven by the low rate of gastric cancer mortality among US-born whites, with similar mortality rates observed among all other foreign-born and US-born groups. Gastric cancer mortality rates significantly decreased during the study period overall (AAPC - 2.50; 95% CI - 3.21, - 1.79) with significant declines observed among US-born (AAPC - 2.81; 95% CI - 3.55, - 2.07) and the foreign-born (AAPC - 2.53; 95% CI - 3.20, - 1.86) population. CONCLUSIONS: Efforts directed at reducing the prevalence of gastric cancer risk factors could help reduce the elevated burden observed among foreign-born individuals and US-born minority groups.

Mechanism-Based Modeling of Tumor Growth and Treatment Response Constrained by Multiparametric Imaging Data.

Multiparametric imaging is a critical tool in the noninvasive study and assessment of cancer. Imaging methods have evolved over the past several decades to provide quantitative measures of tumor and healthy tissue characteristics related to, for example, cell number, blood volume fraction, blood flow, hypoxia, and metabolism. Mechanistic models of tumor growth also have matured to a point where the incorporation of patient-specific measures could provide clinically relevant predictions of tumor growth and response. In this review, we identify and discuss approaches that use multiparametric imaging data, including diffusion-weighted magnetic resonance imaging, dynamic contrast-enhanced magnetic resonance imaging, diffusion tensor imaging, contrast-enhanced computed tomography, [18F]fluorodeoxyglucose positron emission tomography, and [18F]fluoromisonidazole positron emission tomography to initialize and calibrate mechanistic models of tumor growth and response. We focus the discussion on brain and breast cancers; however, we also identify three emerging areas of application in kidney, pancreatic, and lung cancers. We conclude with a discussion of the future directions for incorporating multiparametric imaging data and mechanistic modeling into clinical decision making for patients with cancer.

Trends in liver cancer mortality in the United States: Dual burden among foreign- and US-born persons.

BACKGROUND: Since the mid-1980s, the burden of liver cancer in the United States has doubled, with 31,411 new cases and 24,698 deaths occurring in 2014. Foreign-born individuals may be more likely to die of liver cancer than individuals in the general US-born population because of higher rates of hepatitis B infection, a low socioeconomic position, and language barriers that limit the receipt of early cancer detection and effective treatment. METHODS: To determine whether liver cancer mortality rates were higher among foreign-born individuals versus US-born individuals in the United States, population-based cancer mortality data were obtained from the National Center for Health Statistics of the Centers for Disease Control and Prevention. Annual population estimates were obtained from the US Census Bureau's American Community Survey. Age-adjusted mortality rates and rate ratios (RRs) for liver cancer stratified by birth place were calculated, and the average annual percent change (AAPC) was used to evaluate trends. RESULTS: A total of 198,557 deaths from liver and intrahepatic bile duct cancer were recorded during 2005-2014, and 16% occurred among foreign-born individuals. Overall, foreign-born individuals had a 24% higher risk of liver cancer mortality than US-born individuals (RR, 1.24; 95% confidence interval [CI], 1.22-1.25). Foreign-born individuals did not have any significant changes in liver cancer mortality rates overall, but among US-born individuals, liver cancer mortality rates significantly increased (AAPC, 2.7; 95% CI, 2.1-3.3). CONCLUSIONS: Efforts that address the major risk factors for liver cancer are needed to help to alleviate the health disparities observed among foreign-born individuals and reverse the increasing trend observed in the US-born population.

Mathematical models of tumor cell proliferation: A review of the literature.

INTRODUCTION: A defining hallmark of cancer is aberrant cell proliferation. Efforts to understand the generative properties of cancer cells span all biological scales: from genetic deviations and alterations of metabolic pathways to physical stresses due to overcrowding, as well as the effects of therapeutics and the immune system. While these factors have long been studied in the laboratory, mathematical and computational techniques are being increasingly applied to help understand and forecast tumor growth and treatment response. Advantages of mathematical modeling of proliferation include the ability to simulate and predict the spatiotemporal development of tumors across multiple experimental scales. Central to proliferation modeling is the incorporation of available biological data and validation with experimental data. Areas covered: We present an overview of past and current mathematical strategies directed at understanding tumor cell proliferation. We identify areas for mathematical development as motivated by available experimental and clinical evidence, with a particular emphasis on emerging, non-invasive imaging technologies. Expert commentary: The data required to legitimize mathematical models are often difficult or (currently) impossible to obtain. We suggest areas for further investigation to establish mathematical models that more effectively utilize available data to make informed predictions on tumor cell proliferation.

Colorectal cancer screening for average­risk adults: 2018 guideline update from the American Cancer Society

In the United States, colorectal cancer (CRC) is the fourth most common cancer diagnosed among adults and the second leading cause of death from cancer. For this guideline update, the American Cancer Society (ACS) used an existing systematic evidence review of the CRC screening literature and microsimulation modeling analyses, including a new evaluation of the age to begin screening by race and sex and additional modeling that incorporates changes in US CRC incidence. Screening with any one of multiple options is associated with a significant reduction in CRC incidence through the detection and removal of adenomatous polyps and other precancerous lesions and with a reduction in mortality through incidence reduction and early detection of CRC. Results from modeling analyses identified efficient and model­recommendable strategies that started screening at age 45 years. The ACS Guideline Development Group applied the Grades of Recommendations, Assessment, Development, and Evaluation (GRADE) criteria in developing and rating the recommendations. The ACS recommends that adults aged 45 years and older with an average risk of CRC undergo regular screening with either a high­sensitivity stool­based test or a structural (visual) examination, depending on patient preference and test availability. As a part of the screening process, all positive results on noncolonoscopy screening tests should be followed up with timely colonoscopy. The recommendation to begin screening at age 45 years is a qualified recommendation. The recommendation for regular screening in adults aged 50 years and older is a strong recommendation. The ACS recommends (qualified recommendations) that: 1) average­risk adults in good health with a life expectancy of more than 10 years continue CRC screening through the age of 75 years; 2) clinicians individualize CRC screening decisions for individuals aged 76 through 85 years based on patient preferences, life expectancy, health status, and prior screening history; and 3) clinicians discourage individuals older than 85 years from continuing CRC screening. The options for CRC screening are: fecal immunochemical test annually; high­sensitivity, guaiac­based fecal occult blood test annually; multitarget stool DNA test every 3 years; colonoscopy every 10 years; computed tomography colonography every 5 years; and flexible sigmoidoscopy every 5 years. CA Cancer J Clin 2018;000:000­000. © 2018 American Cancer Society.

Colorectal cancer screening for average-risk adults: 2018 guideline update from the American Cancer Society.

In the United States, colorectal cancer (CRC) is the fourth most common cancer diagnosed among adults and the second leading cause of death from cancer. For this guideline update, the American Cancer Society (ACS) used an existing systematic evidence review of the CRC screening literature and microsimulation modeling analyses, including a new evaluation of the age to begin screening by race and sex and additional modeling that incorporates changes in US CRC incidence. Screening with any one of multiple options is associated with a significant reduction in CRC incidence through the detection and removal of adenomatous polyps and other precancerous lesions and with a reduction in mortality through incidence reduction and early detection of CRC. Results from modeling analyses identified efficient and model-recommendable strategies that started screening at age 45 years. The ACS Guideline Development Group applied the Grades of Recommendations, Assessment, Development, and Evaluation (GRADE) criteria in developing and rating the recommendations. The ACS recommends that adults aged 45 years and older with an average risk of CRC undergo regular screening with either a high-sensitivity stool-based test or a structural (visual) examination, depending on patient preference and test availability. As a part of the screening process, all positive results on noncolonoscopy screening tests should be followed up with timely colonoscopy. The recommendation to begin screening at age 45 years is a qualified recommendation. The recommendation for regular screening in adults aged 50 years and older is a strong recommendation. The ACS recommends (qualified recommendations) that: 1) average-risk adults in good health with a life expectancy of more than 10 years continue CRC screening through the age of 75 years; 2) clinicians individualize CRC screening decisions for individuals aged 76 through 85 years based on patient preferences, life expectancy, health status, and prior screening history; and 3) clinicians discourage individuals older than 85 years from continuing CRC screening. The options for CRC screening are: fecal immunochemical test annually; high-sensitivity, guaiac-based fecal occult blood test annually; multitarget stool DNA test every 3 years; colonoscopy every 10 years; computed tomography colonography every 5 years; and flexible sigmoidoscopy every 5 years. CA Cancer J Clin 2018;68:250-281. © 2018 American Cancer Society.

Precision Medicine with Imprecise Therapy: Computational Modeling for Chemotherapy in Breast Cancer.

Medical oncology is in need of a mathematical modeling toolkit that can leverage clinically-available measurements to optimize treatment selection and schedules for patients. Just as the therapeutic choice has been optimized to match tumor genetics, the delivery of those therapeutics should be optimized based on patient-specific pharmacokinetic/pharmacodynamic properties. Under the current approach to treatment response planning and assessment, there does not exist an efficient method to consolidate biomarker changes into a holistic understanding of treatment response. While the majority of research on chemotherapies focus on cellular and genetic mechanisms of resistance, there are numerous patient-specific and tumor-specific measures that contribute to treatment response. New approaches that consolidate multimodal information into actionable data are needed. Mathematical modeling offers a solution to this problem. In this perspective, we first focus on the particular case of breast cancer to highlight how mathematical models have shaped the current approaches to treatment. Then we compare chemotherapy to radiation therapy. Finally, we identify opportunities to improve chemotherapy treatments using the model of radiation therapy. We posit that mathematical models can improve the application of anticancer therapeutics in the era of precision medicine. By highlighting a number of historical examples of the contributions of mathematical models to cancer therapy, we hope that this contribution serves to engage investigators who may not have previously considered how mathematical modeling can provide real insights into breast cancer therapy.

Variable Cell Line Pharmacokinetics Contribute to Non-Linear Treatment Response in Heterogeneous Cell Populations.

We develop a combined experimental-mathematical framework to investigate heterogeneity in the context of breast cancer treated with doxorubicin. We engineer a cell line to over-express the multi-drug resistance 1 protein (MDR1), an ATP-dependent pump that effluxes intracellular drug. Co-culture experiments mixing the MDR1-overexpressing line with its parental line are evaluated via fluorescence microscopy. To quantify the impact of population heterogeneity on therapy response, these data are analyzed with a coupled pharmacokinetics/pharmacodynamics model. The proliferation and death rates of each line vary with co-culture condition (the relative fraction of each cell line at the time of seeding). For example, the death rate in the parental line under low-dose doxorubicin treatment is increased from 0.64 (± 0.22) × 10-2 to 1.46 (± 0.58) × 10-2 h-1 with increasing fractions of MDR1-overexpressing cells. The growth rate of the MDR1-overexpressing line increases 29% as its relative fraction is decreased. Simulations of the pharmacokinetics/pharmacodynamics model suggest increased efflux from MDR1-overexpressing cells contributes to the increased death rate in the parental cells. Experimentally, the death rate of parental cells is constant across co-culture conditions under co-treatment with an MDR1 inhibitor. These data indicate that intercellular pharmacokinetic variability should be considered in analyzing treatment response in heterogeneous populations.

A Predictive Mathematical Modeling Approach for the Study of Doxorubicin Treatment in Triple Negative Breast Cancer.

Doxorubicin forms the basis of chemotherapy regimens for several malignancies, including triple negative breast cancer (TNBC). Here, we present a coupled experimental/modeling approach to establish an in vitro pharmacokinetic/pharmacodynamic model to describe how the concentration and duration of doxorubicin therapy shape subsequent cell population dynamics. This work features a series of longitudinal fluorescence microscopy experiments that characterize (1) doxorubicin uptake dynamics in a panel of TNBC cell lines, and (2) cell population response to doxorubicin over 30 days. We propose a treatment response model, fully parameterized with experimental imaging data, to describe doxorubicin uptake and predict subsequent population dynamics. We found that a three compartment model can describe doxorubicin pharmacokinetics, and pharmacokinetic parameters vary significantly among the cell lines investigated. The proposed model effectively captures population dynamics and translates well to a predictive framework. In a representative cell line (SUM-149PT) treated for 12 hours with doxorubicin, the mean percent errors of the best-fit and predicted models were 14% (±10%) and 16% (±12%), which are notable considering these statistics represent errors over 30 days following treatment. More generally, this work provides both a template for studies quantitatively investigating treatment response and a scalable approach toward predictions of tumor response in vivo.

Strategies for improved interpretation of computer-aided detections for CT colonography utilizing distributed human intelligence.

Computer-aided detection (CAD) systems have been shown to improve the diagnostic performance of CT colonography (CTC) in the detection of premalignant colorectal polyps. Despite the improvement, the overall system is not optimal. CAD annotations on true lesions are incorrectly dismissed, and false positives are misinterpreted as true polyps. Here, we conduct an observer performance study utilizing distributed human intelligence in the form of anonymous knowledge workers (KWs) to investigate human performance in classifying polyp candidates under different presentation strategies. We evaluated 600 polyp candidates from 50 patients, each case having at least one polyp ≥6 mm, from a large database of CTC studies. Each polyp candidate was labeled independently as a true or false polyp by 20 KWs and an expert radiologist. We asked each labeler to determine whether the candidate was a true polyp after looking at a single 3D-rendered image of the candidate and after watching a video fly-around of the candidate. We found that distributed human intelligence improved significantly when presented with the additional information in the video fly-around. We noted that performance degraded with increasing interpretation time and increasing difficulty, but distributed human intelligence performed better than our CAD classifier for "easy" and "moderate" polyp candidates. Further, we observed numerous parallels between the expert radiologist and the KWs. Both showed similar improvement in classification moving from single-image to video interpretation. Additionally, difficulty estimates obtained from the KWs using an expectation maximization algorithm correlated well with the difficulty rating assigned by the expert radiologist. Our results suggest that distributed human intelligence is a powerful tool that will aid in the development of CAD for CTC.

Seeing is believing: video classification for computed tomographic colonography using multiple-instance learning.

In this paper, we present development and testing results for a novel colonic polyp classification method for use as part of a computed tomographic colonography (CTC) computer-aided detection (CAD) system. Inspired by the interpretative methodology of radiologists using 3-D fly-through mode in CTC reading, we have developed an algorithm which utilizes sequences of images (referred to here as videos) for classification of CAD marks. For each CAD mark, we created a video composed of a series of intraluminal, volume-rendered images visualizing the detection from multiple viewpoints. We then framed the video classification question as a multiple-instance learning (MIL) problem. Since a positive (negative) bag may contain negative (positive) instances, which in our case depends on the viewing angles and camera distance to the target, we developed a novel MIL paradigm to accommodate this class of problems. We solved the new MIL problem by maximizing a L2-norm soft margin using semidefinite programming, which can optimize relevant parameters automatically. We tested our method by analyzing a CTC data set obtained from 50 patients from three medical centers. Our proposed method showed significantly better performance compared with several traditional MIL methods.

Quantitative estimate of the risks and benefits of possible alternative blood donor deferral strategies for men who have had sex with men.

BACKGROUND: Implementation of sensitive screening methods for human immunodeficiency virus (HIV) and hepatitis viruses prompts the question of what quantitative risks may result from altered deferral strategies for donation of blood by men who have had sex with men (MSM). STUDY DESIGN AND METHODS: Quantitative probabilistic models were developed to assess changes in the residual risk of transfusion-transmitted HIV and hepatitis B virus (HBV) associated with blood testing and quarantine release errors (QREs) in the initial year of two hypothetical policy scenarios that would allow donations from donors who have abstained from MSM behavior for at least 5 years (MSM5) or at least 1 year (MSM1). RESULTS: The MSM5 and MSM1 models, respectively, predicted annual increases in units of HIV-infected blood of 0.5% (0.03 mean additional units; 95% confidence interval [CI], 0-1) and 3.0% (0.18 mean additional units; 95% CI, 0-1) over current estimated HIV residual risk using recent, nationwide biologic product deviation reports to estimate QRE rates. These estimates are approximately 10-fold lower than estimates based on New York State QRE data from the previous decade. The models predicted smaller increases in infectious HBV donations. CONCLUSIONS: QREs remain the most significant preventable source of risk. More accurate inputs, including the percentage of MSM in the population, the percentage of MSM who have abstained from MSM activity for 1 or 5 years, the prevalence of HIV and HBV in MSM who have abstained from MSM activity for 1 or 5 years, the rate of self-deferral, and QRE rates, are required before making more precise predictions.

U.S. burden of disease--past, present and future.

PURPOSE: To review the history and challenges of "burden of disease" studies, how these are dependent on robust epidemiologic data as well as complex conceptual constructions, and to identify the public health policy issues these studies can most usefully inform. METHODS: The emergence of the concept of the "burden of disease" in the public health literature is reviewed, with a focus on the results of an analysis of data from the United States that used the methodology presented in the Global Burden of Disease Study. RESULTS: The systematic analysis of public health mortality data to identify major health problems was conducted by Graunt in 16th-century London. He found that many of the predominant sources of mortality were not the focus of public attention. Today, despite refinements in epidemiologic measurement methods designed to capture the impact of non-fatal health conditions, there are similar incongruities between the major public health problems and expenditures on prevention interventions. CONCLUSIONS: Controversies surrounding the interpretation of "burden of disease" studies are not new. Particularly in developed countries, these studies appear more useful for setting research priorities rather than allocating resources to support prevention efforts. Such investigations are not possible without ongoing support for systematic collection and analysis of descriptive epidemiologic data.