Herbert Davis

Multiscale AM-FM Methods for Diabetic Retinopathy Lesion Detection

In this paper, we propose the use of multiscale amplitude-modulation-frequency-modulation (AM-FM) methods for discriminating between normal and pathological retinal images. The method presented in this paper is tested using standard images from the early treatment diabetic retinopathy study. We use 120 regions of 40 × 40 pixels containing four types of lesions commonly associated with diabetic retinopathy (DR) and two types of normal retinal regions that were manually selected by a trained analyst. The region types included microaneurysms, exudates, neovascularization on the retina, hemorrhages, normal retinal background, and normal vessels patterns. The cumulative distribution functions of the instantaneous amplitude, the instantaneous frequency magnitude, and the relative instantaneous frequency angle from multiple scales are used as texture feature vectors. We use distance metrics between the extracted feature vectors to measure interstructure similarity. Our results demonstrate a statistical differentiation of normal retinal structures and pathological lesions based on AM-FM features. We further demonstrate our AM-FM methodology by applying it to classification of retinal images from the MESSIDOR database. Overall, the proposed methodology shows significant capability for use in automatic DR screening.

Comparative effectiveness of patient education methods for type 2 diabetes: a randomized controlled trial.

BACKGROUND Group education for patients with suboptimally controlled diabetes has not been rigorously studied. METHODS A total of 623 adults from Minnesota and New Mexico with type 2 diabetes and glycosylated hemoglobin (HbA(1c)) concentrations of 7% or higher were randomized to (1) group education (using the US Diabetes Conversation Map program), (2) individual education, or (3) usual care (UC; ie, no assigned education). Both education methods covered content as needed to meet national standards for diabetes self-management education and were delivered through accredited programs from 2008 to 2009. General linear mixed-model methods assessed patient-level changes between treatment groups in mean HbA(1c) levels from baseline to follow-up at 6.8 months. Secondary outcomes included mean change in general health status (Medical Outcomes Study 12-Item Short Form Health Survey [SF-12]), Problem Areas in Diabetes (PAID), Diabetes Self-Efficacy (DES-SF), Recommended Food Score (RFS), and Physical Activity (PA, min/wk). RESULTS Mean HbA(1c) concentration decreased in all groups but significantly more with individual (-0.51%) than group education (-0.27%) (P = .01) and UC (-0.24%) (P = .01). The proportion of subjects with follow-up HbA(1c) concentration lower than 7% was greater for individual education (21.2%) than for group (13.9%) and UC (12.8%) (P = .03). Compared with UC, individual education (but not group) improved SF-12 physical component score (+1.88) (P = .04), PA (+42.95 min/wk) (P = .03), and RFS (+0.63) (P = .05). Compared with group education, individual education reduced PAID (-3.62) (P = .02) and increased self-efficacy (+0.1) (P = .04). CONCLUSIONS Individual education for patients with established suboptimally controlled diabetes resulted in better glucose control outcomes than did group education using Conversation Maps. There was also a trend toward better psychosocial and behavioral outcomes with individual education. TRIAL REGISTRATION clinicaltrials.gov Identifier: NCT00652509.

Some Implications for Hunter-Gatherer Ecology Derived from the Spatial Structure of Resources

Abstract Frequency‐domain methods for analyzing data on the spatial structure of resources are discussed. The effects of spatial variation on reliability and diversity are high when the variance of resource abundance is greater than the squared mean abundance. Mobility of foragers should be high in poor environments with few resources, all maximally out of phase with each other. Mobility should be low when resources show either little spatial variation or when they vary in phase.

Vision-based, real-time retinal image quality assessment

Real-time medical image quality is a critical requirement in a number of healthcare environments, including ophthalmology where studies suffer loss of data due to unusable (ungradeable) retinal images. Several published reports indicate that from 10% to 15% of images are rejected from studies due to image quality. With the transition of retinal photography to lesser trained individuals in clinics, image quality will suffer unless there is a means to assess the quality of an image in real-time and give the photographer recommendations for correcting technical errors in the acquisition of the photograph. The purpose of this research was to develop and test a methodology for evaluating a digital image from a fundus camera in real-time and giving the operator feedback as to the quality of the image. By providing real-time feedback to the photographer, corrective actions can be taken and loss of data or inconvenience to the patient eliminated. The methodology was tested against image quality as perceived by the ophthalmologist. We successfully applied our methodology on over 2,000 images from four different cameras acquired through dilated and undilated imaging conditions. We showed that the technique was equally effective on uncompressed and compressed (JPEG) images. We achieved a 100 percent sensitivity and 96 percent specificity in identifying “rejected” images.

Multi-scale AM-FM for lesion phenotyping on age-related macular degeneration

Age-related macular degeneration (AMD) is the most common cause of visual loss in the United States and is a growing public health problem. The presence and severity of AMD in current epidemiological studies is detected by the grading of color stereoscopic fundus photographs. The purpose of this study was to show that a mathematical technique, amplitude-modulation frequency modulation (AM-FM) can be used to generate multi-scale features for classifying pathological structures, such as drusen, on a retinal image. AM-FM features were calculated for N=120 40×40 regions from 5 retinal images presenting with age-related macular degeneration. The results show that with this technique, drusen can be differenced from normal retinal structures by more than three standard deviations using the AM-FM histograms. In addition, by using different color spaces highly accurate classification of structures of the retina is achieved. These results are the first step in the development of an automated AMD grading system.

A comparative analysis of recruitment methods used in a randomized trial of diabetes education interventions.

Recruitment methods heavily impact budget and outcomes in clinical trials. We conducted a post-hoc examination of the efficiency and cost of three different recruitment methods used in Journey for Control of Diabetes: the IDEA Study, a randomized controlled trial evaluating outcomes of group and individual diabetes education in New Mexico and Minnesota. Electronic databases were used to identify health plan members with diabetes and then one of the following three methods was used to recruit study participants: 1. Minnesota Method 1--Mail only (first half of recruitment period). Mailed invitations with return-response forms. 2. Minnesota Method 2--Mail and selective phone calls (second half of recruitment period). Mailed invitations with return-response forms and subsequent phone calls to nonresponders. 3. New Mexico Method 3--Mail and non-selective phone calls (full recruitment period): Mailed invitations with subsequent phone calls to all. The combined methods succeeded in meeting the recruitment goal of 623 subjects. There were 147 subjects recruited using Minnesotas Method 1, 190 using Minnesotas Method 2, and 286 using New Mexicos Method 3. Efficiency rates (percentage of invited patients who enrolled) were 4.2% for Method 1, 8.4% for Method 2, and 7.9% for Method 3. Calculated costs per enrolled subject were

In-Hospital Mortality Among Patients With Type 2 Diabetes Mellitus and Acute Myocardial Infarction: Results From the National Inpatient Sample, 2000–2010

71.58 (Method 1),

Initial Comparative Response of Experimental Tumors to Peak Pions and X Rays

85.47 (Method 2), and

Factors Influencing Patient Completion of Diabetes Self-Management Education

92.09 (Method 3). A mail-only method to assess study interest was relatively inexpensive but not efficient enough to sustain recruitment targets. Phone call follow-up after mailed invitations added to recruitment efficiency. Use of return-response forms with selective phone follow-up to non-responders was cost effective.

Mortality Among Patients Hospitalized With Heart Failure and Diabetes Mellitus Results From the National Inpatient Sample 2000 to 2010

Background Case‐fatality rates in acute myocardial infarction (AMI) have significantly decreased; however, the prevalence of diabetes mellitus (DM), a risk factor for AMI, has increased. The purposes of the present study were to assess the prevalence and clinical impact of DM among patients hospitalized with AMI and to estimate the impact of important clinical characteristics associated with in‐hospital mortality in patients with AMI and DM. Methods and Results We used the National Inpatient Sample to estimate trends in DM prevalence and in‐hospital mortality among 1.5 million patients with AMI from 2000 to 2010, using survey data‐analysis methods. Clinical characteristics associated with in‐hospital mortality were identified using multivariable logistic regression. There was a significant increase in DM prevalence among AMI patients (year 2000, 22.2%; year 2010, 29.6%, Ptrend<0.0001). AMI patients with DM tended to be older and female and to have more cardiovascular risk factors. However, age‐standardized mortality decreased significantly from 2000 (8.48%) to 2010 (4.95%) (Ptrend<0.0001). DM remained independently associated with mortality (adjusted odds ratio 1.069, 95% CI 1.051 to 1.087; P<0.0001). The adverse impact of DM on in‐hospital mortality was unchanged over time. Decreased death risk over time was greatest among women and elderly patients. Among younger patients of both sexes, there was a leveling off of this decrease in more recent years. Conclusions Despite increasing DM prevalence and disease burden among AMI patients, in‐hospital mortality declined significantly from 2000 to 2010. The adverse impact of DM on mortality remained unchanged overall over time but was age and sex dependent.