Biomedical subjects

Frederick Koster

Publications and source records attributed to Frederick Koster.

6 recordsLinked to original sources

PubMed2007

Computer-aided diagnosis with potential application to rapid detection of disease outbreaks.

Our objectives are to quickly interpret symptoms of emergency patients to identify likely syndromes and to improve population-wide disease outbreak detection. We constructed a database of 248 syndromes, each syndrome having an estimated probability of producing any of 85 symptoms, with some two-way, three-way, and five-way probabilities reflecting correlations among symptoms. Using these multi-way probabilities in conjunction with an iterative proportional fitting algorithm allows estimation of full conditional probabilities. Combining these conditional probabilities with misdiagnosis error rates and incidence rates via Bayes theorem, the probability of each syndrome is estimated. We tested a prototype of computer-aided differential diagnosis (CADDY) on simulated data and on more than 100 real cases, including West Nile Virus, Q fever, SARS, anthrax, plague, tularaemia and toxic shock cases. We conclude that: (1) it is important to determine whether the unrecorded positive status of a symptom means that the status is negative or that the status is unknown; (2) inclusion of misdiagnosis error rates produces more realistic results; (3) the naive Bayes classifier, which assumes all symptoms behave independently, is slightly outperformed by CADDY, which includes available multi-symptom information on correlations; as more information regarding symptom correlations becomes available, the advantage of CADDY over the naive Bayes classifier should increase; (4) overlooking low-probability, high-consequence events is less likely if the standard output summary is augmented with a list of rare syndromes that are consistent with observed symptoms, and (5) accumulating patient-level probabilities across a larger population can aid in biosurveillance for disease outbreaks.

Algorithms↗

PubMed2006

Characterization of shock in a hamster model of hantavirus infection.

Human hantavirus cardiopulmonary syndrome (HCPS) due to Andes, sin nombre and other hantaviruses is characterized by severe pulmonary capillary leak and cardiogenic shock. Hamsters, the only animal manifesting HCPS-like disease, were instrumented with radiotelemeters that enabled ambulatory intracarotid blood pressure recording within an animal biosafety level-4 facility. Following infection with Andes virus, blood pressure and heart rate decreased slowly in a biphasic manner during the first 7 days of infection, followed by a rapid fall in pressure and rapid increase in heart rate during the 10-20 h preceding death on day 9 or 10. The preterminal narrowing of pulse pressure was consistent with a cardiogenic impairment. Heart rate variability analysis implicated increased sympathetic nervous system activity as seen in human HCPS. The hamster model of HCPS mimics not only the pulmonary capillary leak but also the hypotension characteristic of human HCPS.

Animals↗

PubMed2006

Modeling intercellular interactions in early Mycobacterium infection.

Infection with Mycobacterium tuberculosis (Mtb) is characterized by localized, roughly spherical lesions within which the pathogen interacts with host cells. Containment of the infection or progression of disease depends on the behavior of individual cells, which, in turn, depends on the local molecular environment and on contact with neighboring cells. Modeling can help us understand the nonlinear interactions that drive the overall dynamics in this system. Early events in infection are particularly important, as are spatial effects and inherently stochastic processes. We describe a model of early Mycobacterium infection using the CyCells simulator, which was designed to capture these effects. We relate CyCells simulations of the model to several experimental observations of individual components of the response to Mtb.

Computer Simulation↗

PubMed2006

Rodent-associated Bartonella febrile illness, Southwestern United States.

Serum specimens from 114 patients hospitalized with a febrile illness were tested with an indirect immunofluorescence assay (IFA) using Bartonella antigens prepared from 6 species of sigmodontine rodents and 3 known human Bartonella pathogens: B. henselae, B. quintana, and B. elizabethae. Acute- and convalescent-phase serum samples from 5 of these patients showed seroconversion with an IFA titer >512 to rodent-associated Bartonella antigens. The highest titer was against antigen derived from the white-throated woodrat (Neotoma albigula), although this rodent is not necessarily implicated as the source of infection. Three of the 5 who seroconverted showed no cross-reaction to the 3 Bartonella human pathogens. Common clinical characteristics were fever, chills, myalgias, leukopenia, thrombocytopenia, and transaminasemia. Although antibodies to Bartonella are cross-reactive, high-titer seroconversions to rodent-associated Bartonella antigens in adults with common clinical characteristics should stimulate the search for additional Bartonella human pathogens.

Adolescent↗

PubMed2004

High seroprevalence of hantavirus infection on the Azuero peninsula of Panama.

The first outbreak of hantavirus pulmonary syndrome (HPS) in Central America was documented on the Azuero peninsula of Panama in late 1999 and 2000. Reverse transcriptase-polymerase chain reaction evidence implicated only Choclo virus in symptomatic HPS with a mortality rate of 20%, although two rodent-borne hantaviruses (Choclo virus and Calabazo virus) were identified in the peridomestic habitat. Neighborhood serosurveys around case households found seroprevalence rates as high as 30%, the highest in the Americas except for western Paraguay. We report here population-based serosurveys for 1,346 adults and children in four communities, three on the Azuero peninsula and one in adjacent central Panama. Overall seroprevalence ranged from 33.2% in a population engaged in farming and fishing on Isla de Cañas, to 16.3% and 21.2% in two mainland agricultural communities, to 3.1% in central Panama, with a modest male predominance of 1.2:1. Nine percent of children 4-10 years old were seropositive, and seroprevalence increased with age in all communities, with highest levels of 52% in those 41-50 years old cohort on Isla de Cañas. Univariate analysis identified correlations between seroprevalence and multiple agricultural and animal husbandry activities. However, stepwise logistic regression models identified only raising animals (cows, pigs, goats, poultry) and fishing as significant independent variables. Human infection with hantavirus on the Azuero peninsula, either with Choclo virus or combined with Calabazo virus, is frequent but rarely results in hospitalization due to respiratory illnesses resembling HPS.

Adolescent↗

PubMedDate not supplied

Schistosomiasis in expatriates in the Arusha region of Tanzania.

BACKGROUND: In 1996 the seroprevalence of schistosomiasis in expatriates and travelers who had contact with Lake Malawi, a fresh water source thought to be schistosomiasis-free, was measured at 32%. Clinicians in Arusha, Tanzania, questioned the prevalence of Schistosoma infection in expatriates living in the Arusha region, and how schistosomiasis might relate to symptoms of chronic fatigue in Arusha expatriates. METHOD: We performed a cross-sectional survey of 80 expatriates living in the Arusha region of Tanzania to determine the seroprevalence of schistosome infection. Whole blood was analyzed by the Falcon assay screening test-enzyme-linked immunosorbent assay (FAST-ELISA) for the presence of species-specific Schistosoma mansoni and Schistosoma haematobium antibodies to microsomal antigens of adult Schistosoma worms, followed by confirmatory enzyme-linked immunoelectrotransfer blot (Western blot). Volunteers answered a questionnaire which included length of residence in Arusha, risk factors, symptoms, and previous diagnosis of schistosomiasis. RESULTS: Of the 80 expatriates sampled, 8 (10%) were positive for schistosomiasis (6 to S. mansoni only, 1 to S. haematobium, 1 to both species). Significant risk factors, elicited by questionnaire, included longer residence in the Arusha region (p =.020), history of fatigue (p =.010) and myalgias (p =.047), and previous diagnosis of schistosomiasis by stool or urine ova (p =.0007). CONCLUSION: The lower seroprevalence of schistosomiasis in Arusha expatriates, compared with expatriates and travelers to Lake Malawi, suggests a regional variation of rate of schistosomiasis infection. Although a history of fatigue and myalgias was related to seropositivity, there is no strong evidence that schistosomiasis infection is the cause of chronic fatigue in Arusha expatriates.

Blotting, Western↗