WO2001073616A2 - Method and system for bio-surveillance detection and alerting - Google Patents
Method and system for bio-surveillance detection and alerting Download PDFInfo
- Publication number
- WO2001073616A2 WO2001073616A2 PCT/US2001/009244 US0109244W WO0173616A2 WO 2001073616 A2 WO2001073616 A2 WO 2001073616A2 US 0109244 W US0109244 W US 0109244W WO 0173616 A2 WO0173616 A2 WO 0173616A2
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- bio
- alerting
- recited
- surveillance detection
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/80—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Definitions
- FIGS. 2A and 2B illustrate the potential impact of earlier warning on the survivability of a hypothetical bio-terrorist attack. As shown in Figure 2B, even seemingly small advances in early warning timing could save a tremendous number of lives.
- Such a system should be capable of operating continuously with minimal human intervention, and should exploit the data collection and analysis capabilities of modern information technology and advanced telecommunications.
- the present invention presents a solution to the aforementioned problems associated with the prior art.
- An object of the invention is early detection of health events, such as bio-terrorist attacks, in populations to enable timely responses that save lives.
- Another object is to monitor multiple relevant data sets to detect signals from health events in populations that are undetectable from any single data set.
- a further object of the invention is to automate monitoring of relevant data sets related to the health of populations.
- the present invention is therefore an automated method and system for detecting health events in populations, such as bio-terrorist attacks, that can operate continuously and with minimal human intervention.
- An embodiment of the invention includes a method for bio-surveillance detection and alerting that subtracts background noise from relevant data sets using a background estimation algorithm to create residual data.
- the method also includes modeling the effects of a hypothetical anomalous event on the relevant data sets to create replica data.
- the residual data is matched with the replica data using a detector to detect a real anomalous event similar to the hypothetical anomalous event.
- An alert is triggered if a real anomalous event similar to the hypothetical anomalous event is detected.
- FIGURE 1 is a table listing characteristics of several biological agents.
- FIGURES 2A and 2B are graphs illustrating the potential impact of improved technological surveillance on the survivability of a hypothetical bio-terrorist attack.
- FIGURE 3 is a block diagram illustrating general information flow according to one embodiment of the invention.
- FIGURE 4 is a waterfall graph showing high school absentee data used in one example of the present invention.
- FIGURES 5A and 5B show county sales totals for two pharmacy chains in a test region according to one example of the invention.
- FIGURE 6 is a map showing the relative locations of high schools in a test county according to one example of the invention.
- FIGURE 7 is a graph showing predicted and observed absentee rates at a single high school according to one example of the invention.
- FIGURES 8 A and 8B show examples of matched-filter output for filter lengths of three and four days, respectively, for a 16% infection rate, according to one example of the invention.
- FIGURES 9A through 9C contain plots of Receiver Operating Characteristic (ROC) curves that show detector performance on the third, fourth, and seventh day after incubation, respectively, according to one example of the invention.
- ROC Receiver Operating Characteristic
- FIGURE 10 is a graph showing the matched-filter output according to one example of the invention.
- FIGURE 1 1 A is a graph showing the averaged matched-filter output curves computed using five different data types and ER data only, according to one example of the invention.
- FIGURE 1 IB is a graph showing ROC curves computed for the same two cases from the same 1000 runs depicted in Figure 11A.
- FIGURE 12 is a graph showing the number of victims required for P D ⁇ 0.95 with PFA ⁇
- FIGURE 13 shows a screen shot illustrating the first page of an on-line bio-surveillance system according to one example of the invention.
- FIGURE 14 shows screen shots illustrating various options of a navigation bar of an on-line bio-surveillance system according to one example of the invention.
- FIGURES 15A through 15C show screen shots illustrating various maps displayed in an online bio-surveillance system according to one example of the invention.
- FIGURE 16 shows a screen shot illustrating the detector output of an on-line bio-surveillance system according to one example of the invention.
- FIGURE 17 shows a screen shot illustrating a slide show of an on-line bio-surveillance system according to one example of the invention.
- FIGURE 18 is a data flow diagram illustrating the path of data through an on-line bio- surveillance system according to one example of the invention.
- Modern information technology can be used for data collection and analysis to provide an early alert within a surveillance system.
- the present invention is therefore an automated system for detecting health events in populations, such as bio-terrorist attacks, and is designed to operate continuously and with minimal human intervention.
- Figure 3 is a block diagram illustrating general information flow according to one embodiment of the invention.
- Bio-warfare agents have the potential for infecting not only humans, but also plants and animals. If a bio-warfare agent sensor is at or near the site of a release, detection could JHU/APL DOCKET NO. 1560-0002
- the next indication of a release may be the behaviors exhibited by humans and animals during the early symptoms of disease.
- One behavior may be to stay home and be absent from normal daily activities, while another may be to self medicate with over-the-counter pharmaceuticals.
- the sick individuals may visit their family physicians who may confuse the symptoms with the latest cold or flu virus in the community. Because the physician may not be aware of other cases outside their practice, the disease could go unnoticed for several days.
- Biological agents are generally delivered as aerosols. In such cases the agents will rapidly disperse until they reach concentrations insufficient to cause disease. Alternatively, they can be delivered as water-borne or food-borne agents. Regardless of the delivery method, the initial attack is an event that is local in space (relative to the size, e.g., of a county) and in time. An infected population would likely remain within a spatially local region.
- the onset of disease from a bio-terrorist attack is thus characterized typically by a rapid increase in diseased individuals in a local region.
- a transient signal related to such a rapid increase is of such a short duration it is inevitably non-specific.
- the magnitude of the transient is variable due to the uncertainty in the number of infected individuals. This implies that very simple, relatively non-specific models will likely suffice for early detection.
- Practice of the present invention therefore exploits specific human behaviors exhibited during the onset of disease caused by bio-warfare agents — e.g., purchasing over-the-counter influenza medications — to provide the early alerting needed to reduce mortality.
- the invention involves selection of and access to relevant data sets containing information that is likely to be impacted by an event such as a bio-terrorist attack.
- the invention exploits non- traditional data sources like school and work absenteeism, over-the-counter pharmaceutical sales, electronically filed HMO claims as well as traditional emergency room and nursing home reports. These indicators are grouped into syndromes, weighted and correlated to obtain a view of the health status of the population at resolutions down to the zip code level.
- OTC Over-the-Counter
- FIG. 5A and 5B show county sales totals for two pharmacy chains in the test county. Data are plotted on a weekly scale, although daily data are available. The totals in Figure 5A are higher than the totals in Figure 5B because only four stores of chain B were located in the test county during 1998-2000. Influenza outbreaks of mid February 1999 and early January 2000 are evident on both plots. JHU/APL DOCKET NO. 1560-0002
- a next step in the practice of the invention is to subtract background noise from relevant data sets using a background estimation algorithm to create residual data. Removal of the systematic features of the background results in a residual time series that can be described by a stationary random noise model.
- a Gaussian process describes the nonsystematic fluctuations in order to characterize the residual noise statistics by a covariance matrix that is estimated directly from the residuals.
- the onset of such an event is reasonably assumed to be characterized by the appearance of transient flu-like symptoms in the population.
- the initial transient is assumed to be geographically constrained relative to some spatial domain. With an onset event that is local in time (a transient) and local in space (confined to a spatial neighborhood), relatively small anomalous events can be isolated from the large and highly systematic fluctuations that characterize the day-to-day behavior of the various data sources. Extremely sensitive architectures for detecting spatial-temporal transients that are buried in highly complex and systematic noise are well known in nature.
- Retinas are an important class of such detectors.
- a retina consists of a membrane of cells that function as light transducers and filters. The output of a retina is not simply a function of the intensity of the visual image falling on it. Instead, the retina acts as a spatial-temporal bandpass filter. Slow, global changes in intensity are filtered out. Retinas are sensitive only to transient changes that occur over small spatial scales in the visual field.
- an embodiment of the invention implements an algorithm that is analogous to a retina, and results in a detector that is sensitive to the onset transient of a localized bio-terrorist event.
- Practice of the invention may apply retinal-like algorithms over multiple spatial scales (e.g., county and s nationwide scales). It may also apply retinal-like algorithms to different classes of data sources (e.g., school absentee records or OTC sales data).
- each data stream is filtered by subtracting from it an JHU/APL DOCKET NO. 1560-0002
- each data source defines a "center,” and its neighbors define the corresponding "surround.”
- Figure 6 shows a comparison of predicted and observed absentee rates at a single high school.
- a more refined retinal model is obtained by using a weighted average instead of a simple average.
- this approach is particularly effective.
- ARMA or TDNN models can account for systematic socio-economic effects that are manifested over time and/or space.
- Center-surround techniques can be applied simultaneously over multiple time scales. In particular, depending on the resolution of the data, they can be applied at the neighborhood, district, county, state, or national levels.
- the center-surround approach can be illustrated using the daily absentee rates of a set of neighboring schools as the spatially distributed sensors.
- N be the number of schools reporting daily absenteeism in a region of interest (i.e., the test county in this case).
- the absentee rate at school / ' , / ' 1, ...,J, on day t is represented byf t (t), the number of absentees divided by the total school enrollment.
- Our estimate of the absentee rate at school is then: JHU/APL DOCKET NO. 1560-0002
- the Cj are adaptive coefficients fitted to a window of historical data. (Note that a precise notation would add a subscript denoting the school being modeled, i.e., the coefficients used to estimate the absentee rate for school / ' are more correctly written as c,y, ..., cu- The additional subscript in the c, and in the vector c and matrix F is omitted below for simplicity.) For a chosen set of days t — 1, .... T, a separate set of these coefficients is computed for each school by minimizing the sum E, of residuals:
- the independent variable is the own-school absentee rate
- the dependent variables are the other or "surround" school absentee rates
- the constant term is zero.
- a Kalman filter requires the specification of a number of state variables.
- the state variables would represent the fundamental health state of a population. For example, the number of people with various natural diseases and also the potential number of people with initial disease caused by a terrorist event.
- a Kalman filter incorporates known input variables such as the day of the week, season of the year, dates of holidays, price-reduction sales events, weather, etc. From the state and input variables, models describe the predicted effect of an event on the measured data streams. In the present invention the models would be stochastic in nature.
- a Kalman filter optimally estimates the state to minimize the difference between the prediction and the current measurements. The efficacy of a Kalman filter depends on the ability to develop accurate models for the effects influencing the indicator data streams.
- a next step in the practice of the invention models the effects of a hypothetical anomalous event, such as a bio-terrorist attack, on the relevant data sets to create replica data.
- Replica data is defined as data that simulates the effects of a health event on a relevant data set that is monitored by the methods of the present invention.
- the replica data may be based, for example, on input from epidemiologists and various scenario templates including information on disease manifestation, pathogen release models, dosage estimation, and other intelligence. Models also may exploit historical data from, for example, influenza epidemics.
- One way to perform this step is to use adaptive matched filters.
- the adaptive matched filter was developed in the radar community as an optimal detector in the presence of Gaussian noise and has been used widely in a variety of noise environments.
- Thresholds are then applied to these successive products to make detection decisions.
- the ramping and peaking of public health data sources at the onset of an infection outbreak indicate a time-varying signal, and this time variation may be quantified using models for the outbreak behavior.
- models must be based on known characteristics of the infection, on estimates of the populations involved, and on how behaviors of those populations are exhibited in the public health data being observed, which can be determined from observed population behavior during influenza season.
- the data vary on a time scale of days instead of fractions of seconds.
- a second advantage of an adaptive matched filter is its ability to handle disparate noise characteristics from different data sources.
- An optimal detector must consider the noise background as well as the signal model. It should suppress data streams that have significant noise fluctuations that may imitate the desired signal and cause false alarms. However, data channels with low noise should be emphasized for increased detector sensitivity.
- the adaptive matched filter makes an optimal tradeoff between signal and noise in each data source.
- the adaptive matched filter estimates the noise in each channel with covariance matrices computed from data residuals. The residuals are obtained from the data by subtracting adaptive background values. Methods of estimating the background depend on the type of data being processed.
- the filter extends over N days of data and that , is the vector of residual data at day /.
- the first J elements of X are residuals derived from absentee rates of schools 17-8J for that day, the next K elements are from OTC sales at stores 1,...,K, etc.
- C be the estimated covariance matrix of X die and let r be a replica vector of modeled effects of the outbreak on the data.
- M r ⁇ /(r C, r) ⁇ n
- the adaptive matched-filter statistic is:
- matched filter detector was subjected to a plausible test for a preliminary evaluation of the approach.
- the input data required for this test were the simulated effects on the available data sources of an outbreak of infection triggered by the airborne release of a toxic agent in a crowded public area.
- the hypothetical threat chosen for the data simulation was the infectious disease tularemia.
- This disease is caused by the bacterium Francisella tularensis, found worldwide in wild animals, birds, and insects. Humans contract tularemia most frequently by physical contact with animals carrying the organism or from tick bites, but the less common pneumonic form of the disease may be contracted by inhalation. Tularemia was weaponized by the former Soviet Union, hence its choice here as a hypothetical airborne threat. After a 3- to 5-day incubation period, victims become acutely ill, with a 5 to 15% mortality rate.
- a shopping mall in the test county mentioned above was chosen as the site of the hypothetical bio-terrorist event. Demographic data were obtained from the mall management to allow estimates of the size and likely age distribution of the exposed population. These estimates were combined with plausible infection rates and with the knowledge of the effects of widespread upper respiratory illness, seen during influenza outbreaks, to model the effects of an outbreak of pneumonic tularemia on the data sources.
- An artificial signal was formed by adding to each data stream the modeled effect of a hypothetical bio-terrorist event.
- the magnitude of this signal reflected in the number of additional OTC sales, insurance claims, etc., was proportional to the assumed number of people infected because of the event.
- Modeled infection rates ranged from 16% ( ⁇ 1140 infected) to 0.3% of the people exposed to the toxic agent at the mall site. A week during mid- winter was chosen for this event so that the effects of the flu season would provide authentic masking of the signal.
- FIGS. 8A and 8B show examples of matched-filter output for filter lengths of three and four days, respectively, for a 16% infection rate.
- the "*" and “o” symbols indicate matched-filter output with and without, respectively, the added signal on the third day after incubation of the released agent, a couple of days before measures would otherwise be taken to deal with the JHU/APL DOCKET NO. 1560-0002
- the replica perfectly matched the simulated signal that was injected into the data. It is unlikely that one would be able to model a real anomalous event (e.g., an actual bio-terrorist attack) that precisely.
- a Monte Carlo simulation was performed with 1000 random trials. In each trial, a random signal was drawn from a Poisson distribution whose mean matched the replica. Thus, there was generally some degree of mismatch between the signal and the replica. Results of the simulation are summarized by receiver operating characteristic (ROC) curves. These curves plot the detection probability and the false alarm rate as the threshold is varied.
- ROC receiver operating characteristic
- Figures 9A through 9C contain plots of ROC curves computed from this set of simulations and show detector performance on the third, fourth, and seventh day after incubation, respectively. Individual curves represent different infection rates as labeled and thus different signal-to-noise ratios in the data. For the larger rates and stronger signals, the detector yields high probabilities of detection (P D 'S) relative to probabilities of false alarm (PFA's). For example, on the fourth day, a 95% P D is achieved with a PFA of only 5% for the case of a 4% infection rate. By the seventh day, outbreaks resulting from much smaller infection rates are detected.
- the second example utilized the absentee data discussed above. Data streams were the daily absentee rates often high schools. For each school, the center-surround predictions from the other 9 schools were used as the background data estimate.
- the signal-generating event was again the hypothetical toxic aerosol release at the shopping mall location.
- the signal was constructed to simulate the relative effects at each school according to the distance of the school from the release site.
- the release date was set JHU/APL DOCKET NO. 1560-0002
- P[S(j)_M] probability of enrollment at schooly given presence at the mall.
- the vector of values of N was computed from the demographic data and from the assumed infection rate.
- Values of P[S ⁇ )_M] were computed using the Bayes Theorem for conditional probability.
- P[M_S(j)] probability of presence at the mall given enrollment at schooly
- P[M& S ⁇ )] joint probability that a student is at the mall and attends school j
- PfS J probability that a county public high school student attends schooly Then, the desired probability is
- P[S(j)] is simply the local school enrollment divided by the total enrollment.
- Components of vector P[M_S(j)] were estimated according to the distance of schooly from the mall, and used with Eq. (6) to compute the number of infected students in each school. The vector of these infection counts for all schools was used as the replica for the matched-filter processing. These counts were also added to the absentee data on the days chosen for the hypothetical event to add a signal to the noise.
- the matched-filter implementation of Eq. (4) requires the matrix X, of residuals obtained by subtracting a background process from the signal-plus-noise data for all 10 schools on each day in question. (Thus, the matrix X, represented 10 data streams.)
- the background in this case was the set of center-surround absenteeism estimates from the neighboring schools, as described above.
- the covariance matrices C were formed and updated by averaging the outer products (X, * X, ⁇ ) for a full school year preceding the day addressed. The availability of the complete set of absentee data for this length of time yielded stable matched-filter behavior.
- a third example of the present invention includes a demonstration of the detector with multiple data sources versus only Emergency Room (ER) data.
- data sources from high school absenteeism, OTC sales, insurance claims, nursing home illness records, and emergency room visits were all included.
- the February 2000 mall infection event simulation described above was repeated using the requested data sets and an assumed infection rate of 16%.
- the center-surround methodology was generalized and applied to those data sets for which data could be separated geographically.
- the OTC sales for pharmacy chain B were separated by the store of purchase, while insurance claim and ER admission records were sorted by patient zip code.
- the four county stores were treated as center-surround neighbors so that background estimates for each store could be formulated as JHU/APL DOCKET NO. 1560-0002
- Figure 11A shows the averaged matched-filter output curves computed using all five data types and JHU/APL DOCKET NO. 1560-0002
- Figure 11B shows ROC curves computed for the two cases from the same 1000 runs. From the standpoint of ROC analysis, the advantage of using the extra sources is considerable.
- the horizontal axis gives false alarm probabilities, with arrows indicating the probability of a single false alarm per week, per month, and per year. At a level of one false alarm per month, the P D for the five-source case is about 95%, but it falls below 10% if only ER data are used.
- the set of 1000 trials was repeated for each of a set of lower infection rates ranging from 8% to 0.3%. The purpose of these runs was to determine how small an outbreak — i.e., how weak a signal — could be detected.
- Figure 12 summarizes the results obtained by plotting the number of victims required for P D ⁇ 0.95 with PFA ⁇ 0.05 as a function of days after the earliest incubation of the disease.
- the number of victims was obtained from the ROC curve for each infection rate that satisfied the probability requirements.
- the summary is shown for the detector using the same two sets of data: the solid curve represents the full set of data sources with all 31 data streams, while the dashed curve represents only ER data. According to this comparison, for the crucial days following the earliest incubation of the disease, the number of victims required for an alert when all five data sources are used is half of that resulting from the use of ER data only.
- a matched-filter detector was devised to enable the early alert desired. This detector was exercised by simulation of the data effects of widespread infections caused by the hypothetical release of a toxic biological agent in a public area. Current demographic data and the expertise of an epidemiologist were used to estimate the effects on the various data streams and to create the replica data. In the simulated scenarios, the detector results were clear enough to permit notification of public health authorities two to three days before a likely conventional alert based on emergency room admissions. Other embodiments of the detector of the present invention include the use of
- Neyman-Pearson detectors change detectors and Bayesian Inference Networks.
- General Neyman-Pearson detectors could be used to improve receiver-operator-characteristics. This includes the use of nonlinear filters, e.g., neural-network-based density-estimation of non- Gaussian statistic filters.
- Change detectors as used with the present invention, are based on the theory that data are drawn from a random distribution. Then at some point in time the distribution changes. Detection of the time of the change is accomplished through various combinations of samples of a log-likelihood ratio. There are similarities between the use of a change detector and an adaptive matched filter.
- Change detectors take the general form of an integrator applied to a series of log-likelihood ratio samples. The integration is useful in reducing the background noise variance, so that the detector does not trigger on every noise fluctuation.
- One implementation of change detection theory is the application of an integrator to the output of a matched filter. The duration of the integrator must be carefully chosen to ensure short duration events are not missed. Nevertheless, integration over just a few samples (e.g., days) of data could produce a significant integration gain. JHU/APL DOCKET NO. 1560-0002
- Bayesian Inference Networks could be employed as detectors.
- Bayesian networks have been shown to be an efficient and general way of representing complex distribution functions incorporating both discrete and continuous variables.
- Still another embodiment of the present invention includes an automated expert that can maintain the effectiveness of the detector as a function of changing demographics, consumer behavior, and other input data characteristics. This maintenance function may be served by a test/evaluation capability for the automated agent, including the generation, execution, and analysis of a set of benchmark scenarios, and by a capability to modify the detectors. Yet another embodiment of the invention includes a methodology to draw inferences from detections, such as the location and scale of a suspected outbreak, the portion of the population at most risk, the type of agent responsible, etc. Such features further enhance the alerting capability of the invention.
- the present embodiment (referred to here as the "On-Line Bio-surveillance System”) is a graphical web-based system that allows users to plot many data sources in many combinations onto a single map. It gives users tools to manipulate data and is both an alarm based and information based system. It has features to help those in need of seeing alerts and potential disease hot spots, as well as those who just need to see the details of incoming data.
- the On-Line Bio-surveillance System described in detail below is an example of one embodiment of the present invention that is a graphical web-based system that allows users to view Geographic Information System (GIS) based images of bio-surveillance data.
- GIS Geographic Information System
- on-line user is authorized to view are a function of the access level of the user. For example, county public health officials would only be able to access data for their county, while state public health officials would have access to s nationwide data.
- ESRI Arc View ® GIS is the map- generating component of the system and Microsoft Access ® is the database for the system.
- the system connects Arc View ® , Access ® , and the user via a web browser and employs static pictures in its displays. It should be noted that the system was built using ArcView and Access ® but that it is not limited to these applications. Any software that provides the basic features of these applications can be used. Due to data size constraints, and the nature of putting information on the internet, a system designer should ensure that on-line images are both useful and give the most information per pixel as possible but without giving too much information to the user. This may be accomplished by several methods.
- a first method is to generalize some of the data sources from specific named sources to more general terms.
- a second method is to display the data normalized against itself, and then display the levels, instead of the actual numbers. Not only does the latter method solve privacy problems associated with putting specific numbers on a website from which users may gather more information than intended, but it also can result in a more valuable representation of data.
- the technique of normalizing is done using running averages and standard deviations. For each category, a running average is stored. A running standard deviation is also stored. After subtracting the average from a current value, one divides that result by the standard deviation to determine the number of standard deviations away from the average. By normalizing the data against itself in this way, one can then display the level of the data instead of the original values.
- This technique provides a common alerting threshold across all data. With just one color-scheme and one legend, every type of data can be displayed on the same scale. This also facilitates the comparing of multiple data sources.
- Another issue includes operations and maintenance of the system.
- the process of dealing with the everyday tasks of maintaining a system must be as automated as possible.
- Avenue scripts for the Arc View ® tasks, and JavaScripts for the web page designs.
- Avenue scripts are created to set up a map, update the data to the correct value for each day/data type, and export that map with the correct file name to the appropriate folder. So by executing one script, every image on the web site can be updated.
- the design of the web site using client-side JavaScript is broken up into a few pieces that work together to make a complete package of tools for the user.
- the design of one embodiment of the On-Line Bio-surveillance system is described in detail below.
- the user Upon visiting the web site, the user first sees the image shown in Figure 13.
- the upper-left comer is a calendar used for changing the date of the image a user wants to view.
- the navigation bar used for choosing which data type to view.
- the map area that starts initially with a splash screen of information, but is ultimately the area for viewing all maps and charts.
- the title bar displays logos and/or the title of a specific project. This frame-style design is used so the user always has simultaneous access to a calendar, navigation bar, title bar, and an image of interest.
- the Calendar allows the user to quickly select a day for which the user wants to view data. This is done by clicking on a calendar date, or by using the "Prev Day”/ "Next Day” buttons.
- the calendar modifies whatever map is currently open in the map area, to reflect the new date selected. If the calendar moves the date outside of a given range, it will automatically show an error page in the map area, telling the user they have selected an unavailable date.
- the Calendar modifies the map page by using JavaScripting, and a variable passing technique that adds parameters to the end of the location URL for the map area.
- the "location URL" is the URL that a particular frame is pointed to at a particular time.
- the Calendar determines the location URL for the map area frame, and deciphers what type of page it is currently pointed to by parsing the URL. For example: https://secwww.jhuapl.edu/ncabiosurv/restrict/2001 -01 -0 l/maps.html?FLU_region.
- the Calendar would read that the current map is pointed to a region-based map of over-the- counter flu remedy sales for January 1, 2001. To change the day, but keep the same type of JHU/APL DOCKET NO. 1560-0002
- the Calendar simply changes the 2001-01-01 part of the URL to the correct day.
- the Navigation Bar has several sections, which are displayed in Figure 14.
- One section is the Region Maps section. It allows the user to view the Region Status and Region Legend maps. Once clicked on, the map area automatically changes to display the map of choice. This is done in this embodiment by using JavaScript.
- the radio buttons call JavaScript functions that parse the map area URL and change the end of the URL string to point to the correct type of map.
- a next section of the navigation bar is the Detailed Maps and Charts section.
- the user can select from region or zip code based maps of each data type in the system.
- the data types include over-the-counter anti-diarrhea medication sales, over-the-counter flu remedy sales, emergency room data, and syndrome-based data.
- the emergency room and syndrome data are then broken up into subcategories for a more detailed picture.
- JavaScript is used to automatically change the map area to show the selected map.
- Detector Outputs is another section of the navigation bar. This allows the user to select the regions of interest, and a disease of interest, and view the detector outputs in the map area. Buttons for "check all” and “clear all” are used to allow the user to select or deselect each checkbox easily. After the user has selected appropriate checkboxes, the "get detector output” button is pressed, and a resulting JavaScript is then activated to change the map area to show each combination of region and disease selected.
- the images are coded so that each region and disease correspond to a particular number code. For example, 01_01_mf.jpg is the image used to show the detector output for Tularemia in a particular region. This allows for the JavaScript to associate numbers with each checkbox, and simply combine the values of the selected checkboxes to get a filename for the image that needs to be displayed.
- the Slide Show Control is also displayed on the navigation bar. This feature allows the user to select a data type and view all the images between two dates.
- the slide show control like all other navigation pieces, operates by using form inputs and JavaScript. It uses the inputs from these form pieces to create a new URL for the map area to use. This URL passes parameters to the slideshow.html page, which then decodes the parameters and generates the slide show for a selected data type.
- Still another section of the navigation bar informs the user of when the site was last modified, and gives contact information if the user has questions.
- This section of the navigation bar also uses JavaScript, as the date produced is written using a JavaScript to write out a data variable used throughout the system that contains the last modified date.
- the map view area is where the maps, charts, and graphs are all displayed. This area is illustrated in Figures 15A-15C. Though it may display many different pictures, this area is actually controlled by only four html pages.
- Maps.html takes inputs that produce region status maps ( Figure 15A), detailed zip code based maps ( Figure 15B), and detailed region-based maps ( Figure 15C).
- the URL that is given to the map area contains information about which type of map to display. For example, https://secwww.jhuapl.edu/ncabiosurv/restrict/2001 -01 -0 l/maps.html?FLU_region. From this URL a user can see that maps.html is called, and it is passed the variable "FLU region". This instructs maps.html to display the region-based over-the-counter flu remedy sales map. The system also must label the map correctly with the date of the map, which is also decoded from the URL. Since all maps use the same level system for color- coding, only one static legend image is needed to describe the legend.
- a special ability of the zip code based maps allows the user to determine the zip code pointed to on the map.
- An image map is added that places invisible circles onto the image. JHU/APL DOCKET NO. 1560-0002
- This image map information is created from an Avenue script that gets information from the zip code map in Arc View ® .
- Triggering an alert includes any standard method of notifying a system user that a detector threshold has been crossed. This includes simply issuing a warning when a threshold is crossed or displaying detector output images. Auto alerting could also be performed via page, e-mail, fax, or phone messages sent to disease control personnel for the jurisdictions where an abnormal condition exists.
- Detectors.html is the html page that deals with detector output images. Detector output images are different enough that they warrant a different html page from maps.html. Detectors.html, however, acts very similar to maps.html in that it reads parameters from its URL and uses those parameters to display the correct image. https://secwww.jhuapl.edu/ncabiosurv/restrict/2001-01-0 l/detectors.html?09_01 &. This URL shows that detectors.html is called, and it is passed the parameter "09_01&". The "&" is there to separate multiple parameters, and the final "&" in the URL is ignored.
- the "09” tells the html page that it needs to display an image from a particular region.
- the "01” tells the html page that it needs to display a Tularemia image.
- the html page then puts the two together to display the "09_01_mf.jpg” image for the day to which the map area is currently pointed. If multiple parameters are passed, they are separated by the "&” and each image is displayed separated by a " ⁇ BR>” tag to stack the images on the page.
- the slideshow page is somewhat different from the two previous pages.
- An example of the slideshow page is shown in Figure 17.
- the slideshow.html page is not located inside of a date folder, but in the main folder of the system. Because the slideshow is passed dates as parameters, there needs to only be one slideshow.html page, not one in each date folder.
- the slideshow.html page takes in parameters similar to the previous html pages. https://secwww.jhuapl.edu/ncabiosurv/restrict/slideshow.html?FLU_zip «&2001 -01 -01&2001 - 01-20.
- the "FLU_zip” tells the html page which data type to display.
- the "2001-01-01” tells the page when to start the slideshow, and the "2001-01-20” tells the page when to end the JHU/APL DOCKET NO. 1560-0002
- the page After it receives the parameters, the page will begin to display the first image, and will begin downloading the rest of the images in the background. It also writes onto the screen a hyperlink for each date in the range. However, these hyperlinks are there not for the user to click on, but are there for the user to run their mouse over.
- Each hyperlink has an "onMouseOver” function attached to it. When the user runs their mouse over the link, it calls the “onMouseOver” function, which in turn calls other JavaScript functions that change the image the user can see on the page. It will change that image to the day the user's mouse has passed over. This allows the user to run their mouse up and down the date range, and watch how a data source changes throughout the range quickly and easily.
- the final html page that is used in the map region area is the region_legend.html page.
- This is a static page that shows the users what regions are used in the system.
- the term "region” is used instead of county, because the system maps all information to the zip code level. Zip codes do not map into counties cleanly, so the system considers the center point (as given by Arc View ® ) of each zip code, and uses that center point to determine what county each zip code is in. From there the system uses each of the zip codes in each county to outline a "region". This region is similar to the county, but follows the shape of the outer zip codes, and not the county.
- the Data Flow diagram shown in Figure 18 illustrates the path of data through the system. It begins with raw data being added to the Access ® Database. Raw data is initially placed into a set of tables and then is modified through a collection of macros and queries that populate a different set of tables. Arc View ® then uses an SQL connection to query these new tables. It places a copy of the tables in Arc View ® . Avenue scripts then generate a map for each data type and date based on the values in the tables. After generating each map, an Avenue script is called to export a JPEG image of the map into its correct place in the web site directory structure. Once all the images have been created, FTP is used to transfer the web site information to the secure web server. From the secure web site, users use different JHU/APL DOCKET NO. 1560-0002
- map slide shows methods to view the maps and images including region and zip code based maps, detector output strip charts, and map slide shows.
- the present invention is an automated system for detecting health events in populations that is designed to operate continuously and with minimal human intervention. It exploits modem information technology and advanced telecommunications to rapidly detect JHU/APL DOCKET NO. 1560-0002
- the invention enables the detection of health events significantly in advance of other methods and systems that monitor a smaller number of data sources.
- the invention also enables a user to make additional inferences about the severity of a disease outbreak as well as its location, the nature of the disease, who is at risk, etc.
Abstract
Description
Claims
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP01954574A EP1311992A2 (en) | 2000-03-23 | 2001-03-23 | Method and system for bio-surveillance detection and alerting |
US10/130,404 US7249006B2 (en) | 2000-03-23 | 2001-03-23 | Method and system for bio-surveillance detection and alerting |
JP2001571264A JP2003529832A (en) | 2000-03-23 | 2001-03-23 | Biomonitoring detection and warning methods and systems |
CA002403797A CA2403797A1 (en) | 2000-03-23 | 2001-03-23 | Method and system for bio-surveillance detection and alerting |
AU2001276815A AU2001276815A1 (en) | 2000-03-23 | 2001-03-23 | Method and system for bio-surveillance detection and alerting |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US19157600P | 2000-03-23 | 2000-03-23 | |
US19156300P | 2000-03-23 | 2000-03-23 | |
US60/191,576 | 2000-03-23 | ||
US60/191,563 | 2000-03-23 |
Publications (2)
Publication Number | Publication Date |
---|---|
WO2001073616A2 true WO2001073616A2 (en) | 2001-10-04 |
WO2001073616A3 WO2001073616A3 (en) | 2003-03-20 |
Family
ID=26887172
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US2001/009244 WO2001073616A2 (en) | 2000-03-23 | 2001-03-23 | Method and system for bio-surveillance detection and alerting |
Country Status (6)
Country | Link |
---|---|
US (1) | US7249006B2 (en) |
EP (1) | EP1311992A2 (en) |
JP (1) | JP2003529832A (en) |
AU (1) | AU2001276815A1 (en) |
CA (1) | CA2403797A1 (en) |
WO (1) | WO2001073616A2 (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2004030512A2 (en) * | 2002-08-19 | 2004-04-15 | Stout Solutions, Llc. | Bio-surveillance system |
EP1488355A1 (en) * | 2002-03-26 | 2004-12-22 | Ibex Healthdata Systems Inc. | Detection of health-related events system and method |
EP1635282A1 (en) * | 2003-06-18 | 2006-03-15 | Matsushita Electric Industrial Co., Ltd. | Biological information utilization system, biological information utilization method, program, and recording medium |
WO2009073604A2 (en) * | 2007-12-03 | 2009-06-11 | Smiths Detection Inc. | Mixed statistical and numerical model for sensor array detection and classification |
Families Citing this family (70)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7194395B2 (en) * | 2000-02-23 | 2007-03-20 | The United States Of America As Represented By The Secretary Of The Army | System and method for hazardous incident decision support and training |
US7058637B2 (en) * | 2001-05-15 | 2006-06-06 | Metatomix, Inc. | Methods and apparatus for enterprise application integration |
US8572059B2 (en) * | 2001-05-15 | 2013-10-29 | Colin P. Britton | Surveillance, monitoring and real-time events platform |
US6925457B2 (en) * | 2001-07-27 | 2005-08-02 | Metatomix, Inc. | Methods and apparatus for querying a relational data store using schema-less queries |
US6856992B2 (en) * | 2001-05-15 | 2005-02-15 | Metatomix, Inc. | Methods and apparatus for real-time business visibility using persistent schema-less data storage |
US20030208499A1 (en) * | 2002-05-03 | 2003-11-06 | David Bigwood | Methods and apparatus for visualizing relationships among triples of resource description framework (RDF) data sets |
US7822621B1 (en) | 2001-05-16 | 2010-10-26 | Perot Systems Corporation | Method of and system for populating knowledge bases using rule based systems and object-oriented software |
US7831442B1 (en) | 2001-05-16 | 2010-11-09 | Perot Systems Corporation | System and method for minimizing edits for medical insurance claims processing |
EP1440409A2 (en) | 2001-11-02 | 2004-07-28 | Siemens Medical Solutions USA, Inc. | Patient data mining, presentation, exploration, and verification |
US7313531B2 (en) * | 2001-11-29 | 2007-12-25 | Perot Systems Corporation | Method and system for quantitatively assessing project risk and effectiveness |
US7457731B2 (en) * | 2001-12-14 | 2008-11-25 | Siemens Medical Solutions Usa, Inc. | Early detection of disease outbreak using electronic patient data to reduce public health threat from bio-terrorism |
US7096125B2 (en) * | 2001-12-17 | 2006-08-22 | Honeywell International Inc. | Architectures of sensor networks for biological and chemical agent detection and identification |
DE10222203B4 (en) * | 2002-05-18 | 2004-03-25 | Robert Bosch Gmbh | Image monitoring system |
JP4463196B2 (en) * | 2002-10-08 | 2010-05-12 | フード セキュリティ システムズ, エル.エル.シー. | A consumer risk diversification model for identifying and tracking food contamination incidents |
US7069233B2 (en) * | 2002-12-31 | 2006-06-27 | Bracken Foster & Associates, Llc | System and method for identification and notification of elevated over-the-counter medication sales with response coordination |
US7951409B2 (en) | 2003-01-15 | 2011-05-31 | Newmarket Impressions, Llc | Method and apparatus for marking an egg with an advertisement, a freshness date and a traceability code |
US9412141B2 (en) * | 2003-02-04 | 2016-08-09 | Lexisnexis Risk Solutions Fl Inc | Systems and methods for identifying entities using geographical and social mapping |
US7399276B1 (en) * | 2003-05-08 | 2008-07-15 | Health Hero Network, Inc. | Remote health monitoring system |
US20050060189A1 (en) * | 2003-09-15 | 2005-03-17 | Xiaohui Zhang | Methods and system for bio-intelligence from over-the-counter pharmaceutical sales |
US7665063B1 (en) | 2004-05-26 | 2010-02-16 | Pegasystems, Inc. | Integration of declarative rule-based processing with procedural programming |
WO2006044689A2 (en) * | 2004-10-15 | 2006-04-27 | Food Security Systems, Inc. | Food product contamination event management system and method |
US7770224B2 (en) * | 2004-10-18 | 2010-08-03 | Lockheed Martin Corporation | CBRN attack detection system and method I |
US8335704B2 (en) | 2005-01-28 | 2012-12-18 | Pegasystems Inc. | Methods and apparatus for work management and routing |
US7627545B2 (en) * | 2005-02-02 | 2009-12-01 | At&T Intellectual Property Ii, Lp. | Geocoding method using multidimensional vector spaces |
WO2006107879A2 (en) * | 2005-04-01 | 2006-10-12 | Leader Technologies, Inc. | Command and control architecture |
US20060265253A1 (en) * | 2005-05-18 | 2006-11-23 | Rao R B | Patient data mining improvements |
US10354760B1 (en) * | 2005-10-18 | 2019-07-16 | At&T Intellectual Property Ii, L.P. | Tool for visual exploration of medical data |
US8706514B1 (en) * | 2005-10-18 | 2014-04-22 | At&T Intellectual Property Ii, L.P. | Case management system and method for mediating anomaly notifications in health data to health alerts |
US8924335B1 (en) | 2006-03-30 | 2014-12-30 | Pegasystems Inc. | Rule-based user interface conformance methods |
US20090132232A1 (en) * | 2006-03-30 | 2009-05-21 | Pegasystems Inc. | Methods and apparatus for implementing multilingual software applications |
US20090319295A1 (en) * | 2006-07-25 | 2009-12-24 | Kass-Hout Taha A | Global disease surveillance platform, and corresponding system and method |
US20080177614A1 (en) * | 2007-01-24 | 2008-07-24 | Lianjun An | Method and system for estimating dynamics of workforce absenteeism using information on pandemic spread and mitigation actions |
US8250525B2 (en) | 2007-03-02 | 2012-08-21 | Pegasystems Inc. | Proactive performance management for multi-user enterprise software systems |
US20080253611A1 (en) * | 2007-04-11 | 2008-10-16 | Levi Kennedy | Analyst cueing in guided data extraction |
JP4926891B2 (en) * | 2007-08-30 | 2012-05-09 | 株式会社日立製作所 | Geographic information system and geographic information management method |
US8515160B1 (en) | 2007-10-04 | 2013-08-20 | Hrl Laboratories, Llc | Bio-inspired actionable intelligence method and system |
US8090592B1 (en) * | 2007-10-31 | 2012-01-03 | At&T Intellectual Property I, L.P. | Method and apparatus for multi-domain anomaly pattern definition and detection |
US9746985B1 (en) | 2008-02-25 | 2017-08-29 | Georgetown University | System and method for detecting, collecting, analyzing, and communicating event-related information |
US8775960B1 (en) | 2008-03-10 | 2014-07-08 | United Services Automobile Association (Usaa) | Systems and methods for geographic mapping and review |
CA2722417A1 (en) * | 2008-04-25 | 2009-10-29 | Btsafety Llc | System and method of providing product quality and safety |
US20090319297A1 (en) * | 2008-06-18 | 2009-12-24 | Upmc | Workplace Absenteeism Risk Model |
US10481878B2 (en) * | 2008-10-09 | 2019-11-19 | Objectstore, Inc. | User interface apparatus and methods |
WO2010066033A1 (en) * | 2008-12-08 | 2010-06-17 | Niall Wallace | Disease mapping and infection control system and method |
US8843435B1 (en) | 2009-03-12 | 2014-09-23 | Pegasystems Inc. | Techniques for dynamic data processing |
US8468492B1 (en) | 2009-03-30 | 2013-06-18 | Pegasystems, Inc. | System and method for creation and modification of software applications |
GB0908184D0 (en) * | 2009-05-13 | 2009-06-24 | Univ Exeter | Bayesian-based online burst detection in water distribution systems |
US20110040660A1 (en) * | 2009-08-10 | 2011-02-17 | Allison Damon R | Monitoring And Management Of Lost Product |
US8862448B2 (en) | 2009-10-19 | 2014-10-14 | Theranos, Inc. | Integrated health data capture and analysis system |
US10943676B2 (en) | 2010-06-08 | 2021-03-09 | Cerner Innovation, Inc. | Healthcare information technology system for predicting or preventing readmissions |
US8880487B1 (en) | 2011-02-18 | 2014-11-04 | Pegasystems Inc. | Systems and methods for distributed rules processing |
US20130024205A1 (en) * | 2011-07-18 | 2013-01-24 | International Business Machines Corporation | Dynamically updating electronic medical records and leveraging a person's path in a facility to mitgate risks |
US9129039B2 (en) * | 2011-10-18 | 2015-09-08 | Ut-Battelle, Llc | Scenario driven data modelling: a method for integrating diverse sources of data and data streams |
US9195936B1 (en) | 2011-12-30 | 2015-11-24 | Pegasystems Inc. | System and method for updating or modifying an application without manual coding |
US20140095417A1 (en) * | 2012-10-01 | 2014-04-03 | Frederick S.M. Herz | Sdi (sdi for epi-demics) |
US10366791B1 (en) * | 2014-03-31 | 2019-07-30 | EMC IP Holding Company LLC | Method and system for global epidemic disease outbreak prediction |
US10469396B2 (en) | 2014-10-10 | 2019-11-05 | Pegasystems, Inc. | Event processing with enhanced throughput |
US10185807B2 (en) * | 2014-11-18 | 2019-01-22 | Mastercard International Incorporated | System and method for conducting real time active surveillance of disease outbreak |
CN107624249B (en) * | 2015-03-30 | 2020-07-03 | 意大利电信股份公司 | Method and system for real-time counting of aggregated persons by aggregated data of telecommunication network |
BR112017020831B1 (en) * | 2015-03-30 | 2023-12-05 | Telecom Italia S.P.A. | METHOD AND SYSTEM FOR ESTIMATING A NUMBER OF PEOPLE GATHERING IN AN AREA OF INTEREST |
US10418133B2 (en) | 2015-12-07 | 2019-09-17 | Mastercard International Incorporated | System and method for monitoring containment of an epidemic |
US10698599B2 (en) | 2016-06-03 | 2020-06-30 | Pegasystems, Inc. | Connecting graphical shapes using gestures |
US10698647B2 (en) | 2016-07-11 | 2020-06-30 | Pegasystems Inc. | Selective sharing for collaborative application usage |
US11048488B2 (en) | 2018-08-14 | 2021-06-29 | Pegasystems, Inc. | Software code optimizer and method |
JP2020123354A (en) * | 2020-01-28 | 2020-08-13 | Advanced Medical InfoTec株式会社 | Medical system |
US11567945B1 (en) | 2020-08-27 | 2023-01-31 | Pegasystems Inc. | Customized digital content generation systems and methods |
US20220122700A1 (en) * | 2020-10-20 | 2022-04-21 | LiveBeyond | Predictive Electronic Healthcare Record Systems and Methods for the Developing World |
CN112734977B (en) * | 2020-12-25 | 2022-07-05 | 安徽省安泰科技股份有限公司 | Equipment risk early warning system and algorithm based on Internet of things |
US11948694B2 (en) | 2021-05-12 | 2024-04-02 | Merative Us L.P. | Controlling compartmental flows in epidemiological modeling based on mobility data |
US20220384056A1 (en) * | 2021-05-27 | 2022-12-01 | International Business Machines Corporation | Hypothetical Scenario Evaluation in Infectious Disease Dynamics Based on Similar Regions |
CN117174332A (en) * | 2023-05-25 | 2023-12-05 | 江苏瀚云医疗信息技术有限公司 | Infectious disease monitoring and early warning system and method based on multi-source data |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5911132A (en) * | 1995-04-26 | 1999-06-08 | Lucent Technologies Inc. | Method using central epidemiological database |
-
2001
- 2001-03-23 JP JP2001571264A patent/JP2003529832A/en not_active Withdrawn
- 2001-03-23 AU AU2001276815A patent/AU2001276815A1/en not_active Abandoned
- 2001-03-23 EP EP01954574A patent/EP1311992A2/en not_active Withdrawn
- 2001-03-23 CA CA002403797A patent/CA2403797A1/en not_active Abandoned
- 2001-03-23 WO PCT/US2001/009244 patent/WO2001073616A2/en not_active Application Discontinuation
- 2001-03-23 US US10/130,404 patent/US7249006B2/en not_active Expired - Lifetime
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5911132A (en) * | 1995-04-26 | 1999-06-08 | Lucent Technologies Inc. | Method using central epidemiological database |
Non-Patent Citations (4)
Title |
---|
BAILEY N T J: "Macro-modelling and prediction of epidemic spread at community level" MATHEMATICAL MODELLING, 1986, USA, vol. 7, no. 5-8, pages 689-717, XP008007959 ISSN: 0270-0255 * |
INFANTOSI A F C ET AL: "Phase spectral analysis of measles epidemic outbreaks" MEDINFO 89. PROCEEDINGS OF THE SIXTH CONFERENCE ON MEDICAL INFORMATICS, BEIJING, CHINA AND SINGAPORE, 16-20 OCT. 1989 AND 11-15 DEC. 1989, pages 497-501, XP008007972 1989, Amsterdam, Netherlands, North-Holland, Netherlands ISBN: 0-444-88138-7 * |
NOBRE F F: "Detecting abnormal patterns in public health surveillance data with a probability index function" MEDINFO 92. PROCEEDINGS OF THE SEVENTH WORLD CONGRESS ON MEDICAL INFORMATICS, GENEVA, SWITZERLAND, 6-10 SEPT. 1992, pages 904-909 vol.2, XP008007963 1992, Amsterdam, Netherlands, North-Holland, Netherlands ISBN: 0-444-89668-6 * |
SNACKEN R ET AL: "The CARE Telematics network for the surveillance of influenza in Europe" METHODS OF INFORMATION IN MEDICINE, DEC. 1995, F.K. SCHATTAUER VERLAGSGESELLSCHAFT, GERMANY, vol. 34, no. 5, pages 518-522, XP008007965 ISSN: 0026-1270 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1488355A1 (en) * | 2002-03-26 | 2004-12-22 | Ibex Healthdata Systems Inc. | Detection of health-related events system and method |
EP1488355A4 (en) * | 2002-03-26 | 2007-10-10 | Ibex Healthdata Systems Inc | Detection of health-related events system and method |
WO2004030512A2 (en) * | 2002-08-19 | 2004-04-15 | Stout Solutions, Llc. | Bio-surveillance system |
WO2004030512A3 (en) * | 2002-08-19 | 2004-12-09 | Stout Solutions Llc | Bio-surveillance system |
EP1635282A1 (en) * | 2003-06-18 | 2006-03-15 | Matsushita Electric Industrial Co., Ltd. | Biological information utilization system, biological information utilization method, program, and recording medium |
EP1635282A4 (en) * | 2003-06-18 | 2010-11-24 | Panasonic Corp | Biological information utilization system, biological information utilization method, program, and recording medium |
US8147406B2 (en) | 2003-06-18 | 2012-04-03 | Panasonic Corporation | Biological information utilization system, biological information utilization method, program, and recording medium |
WO2009073604A2 (en) * | 2007-12-03 | 2009-06-11 | Smiths Detection Inc. | Mixed statistical and numerical model for sensor array detection and classification |
WO2009073604A3 (en) * | 2007-12-03 | 2009-07-23 | Smiths Detection Inc | Mixed statistical and numerical model for sensor array detection and classification |
US7672813B2 (en) | 2007-12-03 | 2010-03-02 | Smiths Detection Inc. | Mixed statistical and numerical model for sensor array detection and classification |
Also Published As
Publication number | Publication date |
---|---|
JP2003529832A (en) | 2003-10-07 |
US7249006B2 (en) | 2007-07-24 |
EP1311992A2 (en) | 2003-05-21 |
CA2403797A1 (en) | 2001-10-04 |
AU2001276815A1 (en) | 2001-10-08 |
US20030009239A1 (en) | 2003-01-09 |
WO2001073616A3 (en) | 2003-03-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7249006B2 (en) | Method and system for bio-surveillance detection and alerting | |
Lombardo et al. | Disease surveillance: a public health informatics approach | |
US10055502B2 (en) | System and method for detecting, collecting, analyzing, and communicating event related information | |
Lewis et al. | Disease outbreak detection system using syndromic data in the greater Washington DC area | |
US8881040B2 (en) | System and method for detecting, collecting, analyzing, and communicating event-related information | |
Lombardo et al. | ESSENCE II and the framework for evaluating syndromic surveillance systems. | |
US7266484B2 (en) | Techniques for early detection of localized exposure to an agent active on a biological population | |
Bravata et al. | Systematic review: surveillance systems for early detection of bioterrorism-related diseases | |
Goldenberg et al. | Early statistical detection of anthrax outbreaks by tracking over-the-counter medication sales | |
Lombardo et al. | A systems overview of the electronic surveillance system for the early notification of community-based epidemics (ESSENCE II) | |
US20120197896A1 (en) | System and method for detecting, collecting, analyzing, and communicating event-related information | |
Tsui et al. | A review of healthcare, public health, and syndromic surveillance | |
Yan et al. | Syndromic surveillance systems: Public health and biodefense | |
Liang et al. | Investigating public health emergency response information system initiatives in China | |
Robertson et al. | Avian influenza risk surveillance in North America with online media | |
Martins et al. | Household disaster preparedness in New York City before Superstorm Sandy: findings and recommendations | |
Khasha et al. | Mobile GIS-based monitoring asthma attacks based on environmental factors | |
Schmeltz et al. | Examination of human health impacts due to adverse climate events through the use of vulnerability mapping: A scoping review | |
Cantor et al. | Femicide and Attempted Femicide before and during the COVID-19 Pandemic in Chile | |
Kim | Ecological system levels of risk factors for intimate partner homicide perpetration and victimization: A three-level meta-analysis | |
Chen et al. | Spatial and temporal aberration detection methods for disease outbreaks in syndromic surveillance systems | |
Texier et al. | From outbreak detection to anticipation | |
Coxen | A risk analysis and data driven approach to combating sex trafficking | |
Odoi et al. | Disparities in temporal and geographic patterns of myocardial infarction hospitalization risks in Florida | |
Cakici | Disease surveillance systems |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A2 Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A2 Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG |
|
DFPE | Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101) | ||
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 10130404 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref country code: JP Ref document number: 2001 571264 Kind code of ref document: A Format of ref document f/p: F |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2403797 Country of ref document: CA |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2001276815 Country of ref document: AU |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2001954574 Country of ref document: EP |
|
WWP | Wipo information: published in national office |
Ref document number: 2001954574 Country of ref document: EP |
|
WWW | Wipo information: withdrawn in national office |
Ref document number: 2001954574 Country of ref document: EP |