US 20090118596 A1
A patient monitoring system provides enhanced functional capability relative to known systems and provides a wireless communication link between a patient monitoring device, worn by a patient, and a local hub. The patient monitoring system is adapted to monitor various patient physiological characteristics, such as blood pressure, pulse rate, blood glucose, weight, pulse oximetry and others. The data from the patient monitoring device is wirelessly transmitted to a local hub, which, in turn, is configured to automatically transfer the data to a remote server, for example, over a public or private communications network. In one embodiment of the invention, the server is configured as a web portal to selectively allow access to such patient physiological data by designated third parties, such as physicians, clinicians, relatives and the patient themselves.
1. A patient monitoring system including:
a patient monitoring device determining a plurality of physiological readings of a patient and wirelessly transmitting said plurality of patient readings to a wireless hub;
a wireless hub wirelessly receiving said plurality of physiological readings; and
a remote server, wherein said remote server receives said plurality of physiological readings from said wireless hub and compares said plurality of physiological readings to a predetermined alert condition based on a plurality of physiological readings,
wherein an alert indication is provided to said patient when said plurality of patient readings violate said alert condition.
2. The system of
3. The system of
4. The system of
5. The system of
6. The system of
7. The system of
8. The system of
9. The system of
10. The system of
11. The system of
12. The system of
14. A patient monitoring system including:
a patient monitoring device determining a first physiological reading of a patient and a second physiological reading of a patient and wirelessly transmitting said first physiological reading and said second physiological reading to a wireless hub;
a wireless hub wirelessly receiving said first physiological reading and said second physiological reading; and
a remote server, wherein said remote server receives said first physiological reading and said second physiological reading from said wireless hub, wherein said remote server includes a predetermined alert condition that requires a plurality of physiological readings from a patient to determine whether said alert condition has been triggered,
wherein said remote server uses both said first physiological reading and said second physiological reading to determine whether said alert condition has been triggered,
wherein an alert indication is provided to said patient when said alert condition has been triggered.
15. The system of
wherein at least one of said first physiological reading and said second physiological reading are compared to said second predetermined alert condition to determining whether said second predetermined alert condition has been triggered,
wherein an alert indication is provided to said patient when said second alert condition has been triggered.
16. The system of
17. The system of
18. The system of
19. The system of
20. The system of
21. The system of
22. The system of
23. The system of
24. A method for monitoring a patient, said method including:
determining a plurality of physiological readings of a patient at a patient monitoring device;
wirelessly transmitting said plurality of patient readings to a wireless hub;
wirelessly receiving said plurality of physiological readings at said hub;
transmitting said plurality of physiological readings to a remote server;
receiving said plurality of physiological readings from said wireless hub at said remote server;
comparing said plurality of physiological readings to a predetermined alert condition based on a plurality of physiological readings; and
providing an alert indication to said patient when said plurality of patient readings violate said alert condition.
25. A method for monitoring a patient, said method including:
determining a first physiological reading of a patient at a patient monitoring device;
determining a second physiological reading of said patient at said patient monitoring device;
wirelessly transmitting said first physiological reading and said second physiological reading to a wireless hub;
wirelessly receiving said first physiological reading and said second physiological reading at said hub;
transmitting said first physiological reading and said second physiological reading to a remote server;
receiving said first physiological reading and said second physiological reading from said wireless hub at said remote server, wherein said remote server includes a predetermined alert condition that requires a plurality of physiological readings from a patient to determine whether said alert condition has been triggered;
using both said first physiological reading and said second physiological reading to determine whether said alert condition has been triggered; and
providing an alert indication to said patient when said alert condition has been triggered.
This is a continuation application which claims the benefit and priority of U.S. patent application Ser. No. 10/414,326, filed Apr. 15, 2003, which claimed the benefit of U.S. provisional application No. 60/372,894, filed on Apr. 16, 2002.
This application includes a Computer Listing Appendix on compact disc, hereby incorporated by reference.
The present invention relates to a patient monitoring system and more particularly to a patient monitoring system for monitoring various physiological characteristic data of a patient, such as blood pressure, pulse rate, blood glucose, weight and others, which wirelessly transmits such data to a hub, located near the patient, which, in turn transfers the data automatically to a remote server, for example, over a public or private communications network, which, in one embodiment, the remote server is configured as a web portal which selectively allows access to patient data by selected third party users, such as physicians, clinicians, patients, and/or relatives, and, in addition provides increased functionality relative to known systems by enabling trends of the patient data to be developed as well as automatically generate communications with the patient or other third parties by way of e-mail or pager when predetermined thresholds for selected physiological characteristics are exceeded and/or to remind the patient to take physiological measurements.
Healthcare costs have been increasing at a tremendous rate for the past decade, far exceeding the rate of inflation. The compound average growth rate for healthcare spending over the past decade was 6%, amounting to nearly $1.3 trillion in the year 2001. Chronic disease patients, whose numbers have doubled during the same decade, account for nearly $700 billion of this spending. Managed care organizations have begun to seek help from disease management companies to contain the spending on chronic diseases. Disease management companies thus have developed systems to monitor the chronically ill patients and help lower healthcare spending by improving patient compliance to medication schedules thereby lowering the hospitalization rates.
Various types of patient monitoring systems are known. For example, U.S. Pat. Nos. 5,810,747 and 5,694,940 disclose patient monitoring systems in which the patient monitoring devices are hard wired to a local hub, which, in turn, is connected to a remote station, for example, over a communication network. The fact that the patient monitors are hard wired to the local station significantly reduces the utility of such systems. For example, such systems are obviously not suitable for ambulatory patients and many applications where it may be desired to remotely monitor the physiological characteristics of a patient outside of a non-clinical environment.
Accordingly, various systems have been developed in which the patient monitoring device, normally worn by a patient, is connected by way of a wireless link to a local hub, which, in turn, is connected to a remote station or server by way of a public communication network, such as the Plain Old Telephone System (POTS). Examples of such systems are disclosed in U.S. Pat. Nos. 3,882,277; 5,522,396; and 6,093,146. Each of these systems include a patient monitoring device, connected to a local hub by way of a wireless link, which, in turn, is connected to a remote server by way of a public communication link, such as POTS. Such systems depend on the use of telephone modems which require patients to place a phone receiver into a modem cradle and dial up the remote server. Such systems have been found to be far too complicated and difficult for elderly and critically ill patients.
In order to resolve this problem, patient monitoring systems, for example, as disclosed in U.S. Pat. No. 6,336,900, have been developed, which not only provide a wireless link between the patient monitor and the local hub, but also automatically dial and connect to the remote server to facilitate the transfer of the physiological data to the remote server. Unfortunately, the functional capability of such systems is relatively limited. For example, such systems only provide limited access to the patient data. In addition, such systems can not be used to provide reminders to patients to take readings or provide messages to the patients or third parties when the physiological characteristics of a patient exceed predetermined thresholds. Thus, there is a need for a patient monitoring system for monitoring the physiological characteristics of a patient that provides enhanced functionality and expanded access to the healthcare data while not tethering the patient to a local hub, thereby improving patient compliance, affording the healthcare provider the ability to capture adverse events sooner and avoid costly emergency care and reduce costly home health visits.
Briefly, the present invention relates to a patient monitoring system which provides enhanced functional capability relative to known systems and provides a wireless communication link between a patient monitoring device, worn by a patient, and a local hub. The patient monitoring system in accordance with the present invention is adapted to monitor various patient physiological characteristics, such as blood pressure, pulse rate, blood glucose, weight, pulse oximetry and others. The data from the patient monitoring device is wirelessly transmitted to a local hub, which, in turn, is configured to automatically transfer the data to a remote server, for example, over a public or private communications network. In one embodiment of the invention, the server is configured as a web portal to selectively allow access to such patient physiological data by designated third parties, such as physicians, clinicians, relatives and the patient themselves. In accordance with another important aspect of the invention, the system provides for enhanced functionality relative to known systems and allows trends of the physiological data to be selectively generated. In addition, any third party can set thresholds to automatically notify the patient or other third parties by various communication methods including e-mail and/or pager when a particular physiological characteristic exceeds a predetermined threshold. The system is also configured to provide reminders to patients to take readings.
These and other advantages of the present invention will be readily understood from the following specification and attached drawing wherein:
The present invention relates to a patient monitoring system for monitoring various physiological characteristics of a patient, such as blood pressure, pulse rate and others, as discussed below. The system includes a portable physiological transducer or patient monitoring device that can be worn by a patient in both single user and multi-user applications. The patient monitoring device is adapted to monitor physiological characteristic data of the patient and wirelessly transmit the data to a local hub or base station. In order to reduce the complexity of the system, the hub or base station is configured to automatically transfer the data to a remote server by way of a public or private communication network. In accordance with an important aspect of the invention, the system is configured to selectively enable access to a patients' physiological data by various selectable third party users, as well as the patient. In accordance with another aspect of the present invention, the patient monitoring system provides enhanced functionality relative to known systems. For example, the system in accordance with the present invention enables selectable third party users, as well as the patient, to monitor trends in the data for the physiological characteristics, as well as set alarms, which can automatically generate communications with the patient and/or third parties when a physiological characteristic exceeds a predetermined value. These communications can be, for example, by e-mail or pager. In one embodiment of the invention, a bi-directional communication link is provided between the patient monitoring device and the local hub. This bi-directional communication link allows communications to be sent to the patient monitoring device, for example, to remind a patient to take a physiological reading. The bi-directional communication link also allows handshaking between the patient monitoring device and the local hub to insure the transfer of physiological data to the local hub.
As will be discussed in more detail below, the system in accordance with the present invention can be utilized in a wide variety of applications. For example,
In one exemplary embodiment of the invention, the system is configured as a web portal. In this embodiment, the server is configured as a web server and selectively allows third party access as well as access by the patient to the physiological data transmitted by the portable patient monitoring device. In addition, the system can be used to provide aural and/or visual signals to the patient monitoring device to remind a patient to take readings. As discussed above and as will be discussed in more detail below, various third parties, such as the physician and/or relatives, as well as the patient, can not only access the physiological data of the patient but also view trends of such data and also set alarms which automatically generate messages to either the patient and/or other third parties when the physiological characteristics of the patient exceed a preset value. In accordance with this embodiment, various screen shots as illustrated in
Referring first to
In this embodiment of the invention, the system includes a local hub or receiving station 34 which wirelessly receives data from the patient monitoring device 32 and automatically communicates it to a middleware device 36, which may be a personal computer, an Internet gateway, a home gateway, a phone, a video phone, or a phone modem. Alternatively, the middleware device 36 may not be required if the hub 34 includes its own processing equipment, as will be discussed below. In particular, the middleware device 36 may form part of the hub 34.
The hub or receiving station 34 may be a receiver or transceiver for receiving data from the patient monitoring device 32. Communication, if required, between the hub 34 and the middleware device 36, may be over various communication links, such as a direct connection, such a serial connection, USB connection, Firewire connection or may be optically based, such as infrared or wireless based, for example, home RF, IEEE standard 802.11a/b, Bluetooth or the like.
The middleware device 36 transfers the data received by the hub or receiving station 34 to a remote server 38. The communication link between the middleware device 36 and the remote server may be by direct connection, such as a serial connection, USB, Firewire or optically based such as infrared or wireless based such as home RF, IEEE standard 802.11b, Bluetooth or others. For longer distances, a communication link between the middleware device 36 and a remote server 38 may be by DSL, T-1 connection over a private communication network or a public information network, such as the Internet.
In accordance with one aspect of the invention, the remote server 38 may be configured to provide third party access to the patient physiological data forming a portal. The portal, for example, a web portal, allows the patient and/or third party users, such as relatives and physicians, to interact with the patient data in various ways as discussed below.
The hardware for the patient monitoring device, as well as the hub, is illustrated in
Exemplary Patient Monitoring Devices
Various embodiments of the patient monitoring devices 84 and 85 are contemplated. Both embodiments may optionally include a visual or audio indicating device to allow visual and/or audio communications to be sent from the system to the patient monitoring device. The patient monitoring device 85 (
The MCU 92 may be, for example, Motorola model number 68HC908GP32. The radio 96 may be, for example, a Xemics model number XE 1201. The memory 94 may be, for example, an Atmelmodel number CAT 24AA65. The real time clock may be a Dallas Semiconductor model number DS 1675.
More sophisticated patient monitoring devices 84, as illustrated in
Various physiological transducers can be integrated into the system. These transducers are similar to various commercially available transducers such as: a blood pressure transducer, such as A&D blood pressure cuff, model number UA-767PC; a weight scale, such as an A&D scale, model number UC-321; a blood glucose meter, such as a LifeScan ONE TOUCH.RTM. Ultra; a pulse oximeter, such as a PalmSat Model 2500; a spirometer, such as a Micro Direct model MICRO DL; a prothromben time test for Coumadin.RTM. therapy, such as a PT/NR Protime Microcoagulation System; and a cholesterol monitor, such as a LSP 3101 Personal Cholesterol Monitor Kit. Virtually any physiological transducer which generates an electrical signal can be implemented in the system.
Exemplary block diagrams for the hub for use with the present invention is illustrated in
Referring first to
Once a hub is located, the hub responds with a handshake signal indicating that it is available to receive data as indicated in step 150. After the handshake, the hub may communicate various information to the patient monitoring device in step 152. For example, the hub may send a message to the patient monitoring device indicating that previous readings have been received. The hub may also send a message to the patient monitoring device with the current date and time to update that device. In addition, the hub may send reminders at specified dates and times to remind a patient when to take readings for those patient physiological devices which do not constantly take readings. Finally, the hub may also send a message to the patient monitoring device to change the frequency or baud rate of the communication. Subsequently, in step 154, the patient monitoring device synchronizes its date and time stamp and deletes all previous readings to the date and time in order to initialize the patient monitoring device. After the patient monitoring device is initialized in step 154, the patient monitoring device then sends the data to the hub in step 156 with its associated ID and date and time. The patient monitoring device may also send data to the hub indicating that its battery is low. In step 158, the hub acknowledges receipt of each message received.
The message exchange between the patient monitoring device and the hub is illustrated diagrammatically in
The above message exchange may be identified as a pre-cycle message exchange. Such a pre-cycle message exchange is identified as a time period when the patient monitoring transducer is not actively recording data such as during a time period when a blood pressure cuff has not been inflated.
After the data has been transmitted from the patient monitoring device to the hub, post cycle messages include repeating the messages 162 and 166 after each data cycle has been completed. In addition, a message 172 may be sent from the patient monitoring device to the hub for the first piece of data, indicated by the box 174. After the hub receives the data 174, it sends a message back to the patient monitoring device indicating that the data 174 has been received. Once the patient monitoring device acknowledges from the hub that the first bit of data has been received, additional data is sent by way of a message 178 to the hub, as indicated by the data block 180. The hub responds that the data 180 was received by way of a message 182 back to the patient monitoring device. After all of the data has been transmitted from the patient monitoring device to the hub, the patient monitoring device sends a message 184 indicating that the next bit of data is the last bit of data in the memory to be sent by the patient monitoring device by way of the message 184. The hub then acknowledges receiving the last bit of data 186 by acknowledging by way of message 188 back to the patient monitoring device 188 that it has received that data 186.
As indicated in step 226, when connection, such as Internet connection, is established between the hub and the server, the server sends the hub various messages. These messages may include the current device ID list for the user to the hub. The messages may also include date and time stamps for reminders and the date/time stamp of the last updated reading.
In a exemplary embodiment, as discussed in detail below, the remote server may be configured as a web portal. Exemplary web pages for a patient monitoring system are illustrated in
Server Application Architecture
The web tier 230 makes application functionality available on the worldwide web. This tier 230 accesses data and business functionalities, manages screen flow and often encapsulates some user interaction. It may be used to decouple the client from the business logic tier 232 to provide a uniform service model of client request. The web tier interacts with other tiers using standard protocols. Other applications may take place of the client programs accessing the underlying application through its web tie, such as Apache/Tomcat.
The business logic tier 232 encapsulates the business logic and comprises the software application functionality. The logic is organized as a set of interacting objects allowing for additions and enhancement of the logic at any point while still preserving the interface of other tiers. This business logic tier 232 may be implemented using J2EE, Microsoft.NET or a custom architecture, such as Tomcat.
The data access tier 234 may be used to integrate the application with other enterprise information systems. This tier 234 provides data storage and other information services to the application. It may consist of a database, enterprise resource planning systems, mainframe transaction processors, legacy systems and enterprise integration technologies. Other tiers access these databases using industry standard or custom protocols such as Oracle.
Web Portal Details
As mentioned above, in one exemplary embodiment of the invention, a web portal (hereafter referred to as the portal) provides meaningful access to a patient's personal health data, as collected by the patient monitoring device(s) and stored in a server-based data repository (hereafter referred to as the repository). Some key mechanisms of the portal are listed below and are explained in further detail in subsequent sections. These key mechanisms include:
Device Registration Through Portal
Device registration is the process by which a patient monitoring device is associated with one or more users of the system. This mechanism is also used when provisioning devices for a user by a third party, such as a clinician (or their respective delegate). The mechanism is as follows:
1. The user (or delegate) identifies himself/herself by logging into the portal. The user must have previously registered an account created either by themselves or by delegates.
2. User (or delegate) registers a device by:
3. The device(s) are then associated with this user's account and a Device Control File (discussed below) comprised of device identification information is synchronized between the server and the Receiving Station on initialization (see the section Data Synchronization between Receiving Station and Repository). This enables the Receiving Station to respond to and accept input from the devices that have their identification information in the control file.
4. On each interaction between the hub and the server, the control file entries for the current user (or hub) are synchronized.
5. Similarly, whenever the user registers a device, its entry is added to the control file on the hub if not already there on the next synchronization interaction with the server.
6. No association information between the actual user (person) and the device is kept on the user's hub. This determination is made after the data is uploaded to the server.
7. The hub may be shared between users and data for all users resident on the hub is uploaded to the server on synchronization.
Device Control File
This file contains records in the format
This file is internally maintained in XML format for ease of interfacing but is either kept encrypted or in a non-readable format on the hub for security reasons.
User Data File
This file contains records in the format
Data Synchronization Between Receiving Station and Repository
The hub and the Repository frequently synchronize data. The hub may use one of various transportation methods to connect to the repository e.g. using a PC as conduit or via a connection established using an embedded modem (connected to a phone line) or via another network-connected device (such as, but not limited to, a web-phone, video-phone, embedded computer, PDA or handheld computer). The mechanism is as follows:
1. When a user logs into the portal, the User Data File, resident on the hub, is uploaded to server. The server filters duplicate readings records.
2. Alternatively, the Data Synchronization Client may upload the data from the hub to the Server without user intervention automatic data synchronization without user intervention also happens in the case of a stand-alone hub.
3. The data may be decrypted on the server, assigned to each user (by determining which user the device is registered with) and put into a staging area. This staging area is so that the user may optionally preview the data before accepting the values. The server also filters for duplicate reading values.
4. The device control file is uploaded to the server.
5. The server downloads updated device control file with last updated timestamps for each device in list. If device has been removed from the system, then its entry in the file is deleted. The old device list file is backed up.
6. The data file is cleared/deleted.
Data Synchronization Client
There is often the need for the reading data to be uploaded to the server without explicit user intervention. The Data Synchronization Client runs on the hub to upload the user's data to the server while synchronizing the information contained in the control file.
Users may set up alerts that are triggered when one or more reading meet a certain set of conditions, depending on parameters defined by the user. The mechanism is as follows:
1. The user sets up an alert by choosing the condition that they would like to be alerted to and by providing the parameters (e.g. threshold value for the reading) for alert generation.
2. Each alert thus set up has an interval associated with it. This interval may be either the number of data points or a time duration in units such as hours, days, weeks or months, depending on the type of reading for which this alert is being set up.
3. The values of readings lying within the interval specified by the user must all either positively or negatively exceed the specified threshold for the alert to be generated i.e. there is an implied “and” across all the readings within the interval.
4. The user chooses the destination where the alert may be sent. This destination may include the user's portal, e-mail, pager, voice-mail or any combination of the above.
5. Specific, preset alerts are provided based on medically established conditions in the case of common reading types e.g. heart rate, weight, blood sugar etc. The user may modify the parameter values for these alerts if they so desire.
The system computes trends in the user's data values over an interval. The mechanism is as follows:
1. Trends are determined by applying mathematical and statistical rules (e.g. moving average and deviation) over a set of reading values. Each rule is configurable by parameters that are either automatically calculated or are set by the user.
2. Currently the system tracks deviations from baseline values (see section Baselines) either automatically calculated using intelligent defaults or provided by the user.
3. The user also supplies a range over which the trend is calculated. This is specified either as the number of data points or a numerical unit of time (e.g. hours, days, weeks etc, depending on reading type).
4. This range is always applied backwards from the most current reading i.e. if the user specifies that the trend needs to be calculated over the past 5 days, then the system will calculate the trend of the past 5 days from current date.
5. The system displays the deviation between the trend values computed from received data and the baseline value. This deviation is indicated both as a percentage as well as a numerical difference, along with icons (“+”,“−”,“=”) to visually indicate the trend direction.
6. The system may be used to calculate a baseline value for each reading type using algorithms specific to each reading type. The user can override the default baseline average by changing the default parameters used to calculate the baseline e.g. the start date and the number of readings. Additionally, the user may directly provide baseline value(s) into the system, overriding the computed baseline values. If the user provides parameters such that there is insufficient data for calculating the baseline, then the system alerts the use to the fact. Previously set baseline values are not altered.
During early use of the system, there may be insufficient data for automatically calculating baselines for the user's reading types. In this case too, the system alerts the user to the fact.
Strength of Trend Correlation
The system may contain algorithms to compute the strength of the trend and correlate it to the visual perception of the trend as would be observed by a human observer.
Example Algorithm: In order for a weight trend to be visually perceived as increasing or decreasing, a certain positive or negative standard deviation from a moving average baseline must be observed over a pre-defined time period to be able to inform the user that their weight is either increasing or decreasing.
The user may set up reminders on the device(s) via the portal. The device then draws the user's attention at the appointed time e.g. by sounding a buzzer. The mechanism is as follows:
1. Each device associated with the user's account has its own independent set of reminders.
2. Each reminder can be made repeatable. The user can set the reminder to repeat every n<units of time> (where <units of time> may be hours, days etc.) starting from a preset date and time.
3. The reminders are associated with a user's account and are synchronized with the device(s) via the hub. The reminder information is maintained in the control file (see section Device Control File).
4. The user can specify their time zone so as to account for differences between time on the server vs. the user's local time.
Buddy or Role Based Access
The user may give permission to others as needed to read or edit their personal data. This is useful, for example, for allowing a well-wisher to view ones data/charts, or receive alerts. The clinician could be allowed to edit data for example to annotate it, while the patient would have read-only privileges and that too for certain pages. An authorized person could set the reminders and alerts parameters with limited access to others.
The user or clinician could have a list of people that they want to monitor and have it show on their “My Account” page, which serves as the central monitoring station. The central monitoring concept could be particularly useful in an environment where on person wishes or needs to monitor multiple people.
A clinician administrator (also referred to as Administrator in this context) may monitor the data for and otherwise administer a number of users of the system (also referred to as Patients in this context). The mechanism is as follows:
1. A summary “dashboard” of readings from all Patients assigned to the Administrator is displayed upon log in to the Portal by the Administrator. At a minimum, the Patient's last reading received from all devices is shown (color coded to visually distinguish normal vs. readings that have generated an alert), along with description of the alert generated.
2. The Administrator may drill down into the details for each Patient to further examine the readings data, view charts etc. in a manner similar to the Patient's own use of the system.
3. The Administrator may also view a summary of all the devices registered to all assigned Patients, including but not limited to all device identification information.
4. An Administrator has access only to information about Patients that have been assigned to the Administrator by a Super Administrator. This allows for segmenting the entire population of monitored Patients amongst multiple Administrators.
The Super Administrator may assign, remove and/or reassign Patients amongst a number of Administrators.
The Portal provides a mechanism to enhance the healthcare interaction and education of the users of the system (also referred to as Patients in this context). The mechanism is as follows:
1. A questionnaire may be made available to the Patient containing questions pertinent to their health condition.
2. The Patient's answers to the questions are stored and tracked by the time of each “session” (user interaction). Statistical summaries of these answers for various intervals of time are available for display.
3. The Administrator (or delegate) defines the entire set of questions that may be made available to all Patients. Of these, pertinent questions are selected based on rules applied to the Patient's recent data in the Repository.
Data Export and Reporting
Data may be extracted, exported and reported from the Repository into a variety of formats, both application specific and industry standard. The mechanism is as follows:
1. The data in the Repository is internally extracted into a neutral format represented in XML.
2. This data is further processed to either aggregate information or to remove user identification information, as may be the specific requirement.
3. The internal, neutral format may be transformed into custom, application specific formats or industry standard formats depending on specific application requirements.
The architecture for the web application meets the requirements below. In particular, the architecture must be relatively independent of the user's client environment. At a minimum, the most common browsers (Netscape 4.0 and higher, IE 4.0 and higher) are supported on Windows (95, 98, NT, ME, 2000 now, XP in the near future). In addition, the architecture must be continuously adaptable to support newly emerging and evolving business requirements. This includes changes in the presentation, the business logic as well as the data tier.
In addition, the architecture should meet the following characteristics:
By adapting a standard reference architecture, it allows for the time required for developers to gain familiarity with the architecture to be compressed provided they are well versed in J2EE technology and architecture. This allows for the development effort to be outsourced at a later point in the development cycle. Moreover, the standard architecture, supports rapid and frequent changes to the look of the application. Such an architecture also supports the division of labor amongst the development team e.g. the application is partitioned such that the front-end designer requires minimal programming knowledge and vice versa.
The architecture should also partition the application to allow for the business logic to be accessed via a variety of interfaces e.g. the web (HTML) for now, wireless (WML) and/or voice response (VRU) in the future. It should also accommodate internationalization and localization of the application in the near future without significant changes. In addition, the architecture should leverage emerging technologies to provide a superior user experience e.g. it does not require the user to download, install and configure complicated software by utilizing emerging server-side technology. Moreover, the architecture should have a growth path to allow use of new technology as it becomes available over the application's lifespan (5 years). It should be designed to be resilient for at least the next 1-2 years while accommodating change. Finally, the architecture should not be tied to a particular set of products i.e. any J2EE compatible server can serve as application host. It also is relatively independent of database. This allows for flexibility in choosing and switching between hosting providers.
The Java 2 Enterprise Edition (J2EE) may be used as the standard technology platform for this application. Of the available J2EE components, the initial development phase may use Java Server Pages (JSPs), servlets, Java Database Connectivity (JDBC), Java API for XML Processing (JAXP). Other J2EE components, such as Enterprise Java Beans (EJBs), transactions APIs etc. may be used as required for subsequent development phases. Using J2EE components has the following advantages:
J2EE provides mechanisms that support simplified scaling of distributed applications, without requiring significant effort on the part of the application development team. Because application services such as transaction support, database connections, life cycle management etc. are provided by the platform, the application can leverage these capabilities by adhering to the rules requiring it to be compatible with these services. For example, by using the appropriate JDBC APIs, connection pooling is automatically available for increased performance. Similarly, J2EE servers now provide many scaling mechanisms that were previously designed by the application developer.
Security (access restriction, authentication) is configured declaratively at deployment time instead of being coded at development time. This allows for continuous refinement of the security mechanism (matching users with roles, with groups of users having specific access permissions) without requiring changes to the code followed by redeployment. For comparison, commercial products providing these kinds of capabilities for web applications (e.g. getAccess from Entrust Technologies) are typically priced at $15-$25 per user for quantities less than 5000 users.
As the functionality provided by the J2EE expands, these capabilities are immediately available for use by the application. For example, the Connector architecture in the upcoming release of the J2EE will provide built-in capabilities of communicating with legacy systems (once commercial implementations are available).
Supported by Hosting Services
There are several commercial and open source implementations of J2EE compatible servers with the result that most hosting services offer reasonably priced hosting packages for J2EE applications.
The pages should be as lightweight as possible while being aesthetically pleasing. This leads to using Flash as the rendering technology for items such as buttons and image maps etc. since this leads to compact graphics. However, there are no large animations.
The pages should not be hard-coded for a particular user screen size i.e. pages may be designed using Cascading Style Sheets (CSS).
Popchart (http://www.popchart.com) from Corda may be used to provide charting functions. It runs as an image server and services requests for charts from the web server over a pre-configured TCP port. It supports multiple output formats (including Flash, WML and SVG) and good chart interaction such as pop-up text and drill-down. The charts can be visually designed by a graphic designer and later rendered programmatically.
The reporting technology/product should support the following requirements:
Dreamweaver software may be used as a tool for developing the front-end HTML. Flash and (optionally) Fireworks software can be used along with Dreamweaver. MS Frontpage software, although popular, is not preferred since it introduces irrelevant HTML into each document.
Popchart software (see section Charting above) includes a graphics chart builder that allows charts to be interactively designed. The designer may use this builder to create aesthetically pleasing charts.
All HTML should pass validation for acceptance. Tidy is one utility that can check HTML for validity.
An Apache web server may be used for development. The deployment web server is provided by the hosting service.
Tomcat software may be used for development due to it providing integrated debugging with many IDEs (Integrated Development Environments) and its smaller memory footprint when compared with other similar commercial products.
J2EE Blueprints (JPS)
The current application architecture is modeled after the J2EE Blueprints reference application (the Java Pet Store version 1.1.1). Although EJBs are not used at this stage, the tier separation is maintained to allow them to be introduced at a later stage.
This is a collection of utility classes available at www.servlets.com that may be useful in servlet development. Developed and maintained by Jason Hunter, author of the book Java Servlet Programming, 2nd Ed. (must read).
Integrated Development Environment
Forte for Java, Community Edition may be used because it provides a rich feature set for free. JRun Studio has also been evaluated and may be used if we switch to JRun in a later development phase.
JSP Tag Libraries
Tag libraries from the Apache-Jakarta (and the Struts framework) project may be used if needed. Proprietary taglibs will not be used. Taglibs should be used if their use results in significant effort saving.
Any industry standard build tool (such as Ant or make) can be used as the build tool.
The JUnit unit-testing framework (wwwjunit.org) may be used to unit test all developed code. Extensions to JUnit for server-side testing and J2EE testing (e.g. HttpUnit, Cactus) may also be used.
Either Oracle or MS SQL Server 2000 may be used as the database. It is important to be careful to not have dependencies on the database to allow for switching if required by hosting providers.
Visual SourceSafe or CVS may be used. CVS is accessible over the Internet.
EbugStomp software may be used for web-based bug tracking.
This section describes the architecture of the application: exploring the partitioning of functionality into modules, the assignment of functionality to tiers, and object decomposition within the tiers.
This is a list of the functional modules of the system. Some of the modules are relatively simple but have been broken out to provide an idea of the broad functional areas of the application.
This module provides the basic framework for the other functional modules. It provides key features such as navigation, security etc.
User Account Module
This module maintains information about a person's account (contact information, payment and credit card information) as well as the devices associated with a particular account.
User Profile Module
This module maintains information about a person's profile such as demographic information, medical conditions, medications and allergies.
Device Information Module
This module maintains information about the various types of devices supported by the system.
Data Synchronization Module
This module provides the functionality to upload the newly acquired readings data from a user's PC to the server. It flags potentially erroneous readings and provides a preview of the data so that the user may optionally reject some readings. Please refer to the section Data Synchronization between User's PC and Server for more details.
Data Management Module
This module manages the data obtained for each user. It also provides functionality to associate annotations with data points.
Data View: Charts Module
This module provides the functionality for the user to view his/her readings data as charts. The user may choose to view either preset ranges of data or choose an arbitrary range.
Data View: Tables Module
This module provides the user with a tabular view of their data. There are no complex table-editing functions.
This module contains the logic to provide the user indication of certain trends in their data.
This module provides the functionality where the user can select from preset patterns to look for in the data to be alerted about. The user may choose certain parameters of the triggering events e.g. if their blood pressure is above X for Y days, then trigger an alert, where X and Y are chosen by the user.
The reminders module generates preset reminders to a user about certain events that were supposed to have occurred but have not.
This module will provide functionality in which the user can choose to invite another registered user to be a part of his/her group. The invited user can then have access to the inviting user's data.
This module provides the functionality to send e-mail arising from a number of places in the application: when an alert is generated, for sending a user report etc.
The report module will provide functionality to generate simple reports for the user incorporating charts.
This module will provide the functionality to print on the client PC.
This module provides the capability to export the user's data in Excel format.
The web application will have the following interfaces with external systems. These interfaces may be either synchronous (i.e. online, in real time) or asynchronous (i.e. as a batch operation, deferred in time).
Interface with Credit Card System
The application will transmit the user's credit card number and obtain online verification. It will also make periodic charges to the credit card as the user's account is automatically renewed on expiry.
Interface with E-Store
An outsourced e-store will provide the capability for a user to buy a device online. The application should be able to transfer information such as credit card information etc. to this system. Additionally, the front-end for such a system needs to be created and integrated with the rest of the application.
Interface with Refund Generation System
This system will generate pro-rated refund checks for users canceling their accounts. The application must transfer the required information to this system.
Interface with Customer Support System
A customer support system providing (at least) e-mail based support needs to be integrated with the application/site.
Interface with Device Control Software on the User's PC
At this time, the interface with the device control software running on the user's PC is asynchronous and is handled by uploading and downloading files to a known location on the user's disk. This will eventually be enhanced to obtain the file location from the Windows registry.
The Business Object Framework (BOF) is designed as a lightweight stand-in for Enterprise Java Beans (EJBs). The idea is to create structural parallels with the EJB component structure (such as the home and remote interfaces, session and entity beans etc.) in a simple object framework that allows for easy migration to EJBs at a later time. The key differences between EJBs and BOs are:
There is no container that the BOF objects (and their derived objects/components) run in. Because of that, BOF-derived objects must manage their own lifecycle such as creation and destruction. This is however minimal and is abstracted behind utility (factory) methods.
BOF-derived objects are generically referred to as BOs (for business objects) and are packaged in the bo package within the application structure. The scope of BOs is the same as for regular Java objects. This is different from EJBs, whose scope is controlled by the container and managed by methods on the home interface. The BO implementation objects must directly implement the BOF equivalent of the remote and home interfaces i.e. the BOInterface and BOHome. This is a key distinction from EJBs, in which the implementation object does not directly implement its home and remote interfaces.
The methods in the BOHome that are implemented by the BO have the same names as defined in the interfaces. This is different from EJBs which have a naming convention of prefixing ejb to the method name e.g. ejbCreate for a method named create in the home interface. BOs extend either the SessionBO or EntityBO base class. This is different from EJBs, which instead implement either the SessionBean or EntityBean interfaces. BOs are not remotely referenced and execute in the local JVM instead.
The BOF has the following main types of components as set forth below:
Business Object Interface
This is the equivalent to the EJB remote interface. All business process methods are defined in this interface. Interfaces of this type extend com.carematix.bof.BOInterface and are named per the EJB convention i.e. using the domain class name. The BO must implement this interface directly.
Business Object Home (Interface)
This interface is the equivalent of the EJB home interface. At the very least, it must define a create( . . . ) method to return a concrete BO object implementing the BOInterface. Interfaces of this type extend com.carematix.bof.BOHome and are named per the EJB convention i.e. by appending “Home” to the domain class name. The BO must implement this interface directly.
This is the abstract base class (com.carematix.bof.bo) for Session Business Objects (SBOs) and Entity Business Objects (EBOs) described below.
Session Business Object (SBO)
An SBO is the equivalent of a session EJB. One key distinction is that SBOs can only be stateless, not stateful beyond the scope of the calling method. It is best to assume them as stateless. Typically, SBOs define most of their methods in the BOInterface and have just a simple create( ) method in the BOHome interface. A session BO extends the base class com.carematix.bof.SessionBO.
Entity Business Object (EBO)
An EBO is the equivalent of an entity EJB and encapsulates some persistent data representation. Typically, EBOs have a simple BOInterface and have most of their methods in the BOHome interface such as create ( . . . ), load ( . . . ), store ( . . . ), remove ( . . . ), findByPrimaryKey ( . . . ) etc. Not all of these methods are required in every EBO and only those needed by the application must be defined. An entity BO extends the base class com.carematix.bof.EntityBO.
Source code for the placeholder classes comprising the BOF is available. Accompanying that are two example packages: boexample1 and boexample2. These provide example implementations for a session BO and entity BO respectively. They illustrate the patterns and conventions required to create BOs and should be thoroughly studied.
In order to access a BO from calling code, the following example of a handler method for handling account events is illustrative:
Entries need to be created in the web.xml file for each BO in order to enable JNDI lookup. Entries for the BO examples are as follows:
The XML-based templating solution provides many benefits. XML is a useful, consistent way of representing application parameters. The XML format is easy to use for both presentation specialists and third-party tools. The fact that XML is standardized means we can use off-the-shelf software to handle all the details of parsing and writing our screen definitions, configuration files, and other forms of structured data. Finally, XML can be validated against a DTD (Document Type Definition), so the parser can handle errors relating to ill-formed input, instead of in the application code.
Three files control the templating system:
The template file (template.jsp), which defines the basic layout of all of the pages generated from the template. There is one template file per supported language.
The request mappings file (requestmappings.xml), which maps virtual URLs to screen names (or request handler classes). There is one request mappings file in the application.
The screen definitions file (screendefinitions.xml), which defines the parameters for each screen. There is one screen definitions file per supported language.
The main controller servlet MainServlet receives incoming HTTP requests. MainServlet matches the incoming URLs to the URLs defined in the request mappings file, obtaining a screen name. The template parameters corresponding to the screen name are retrieved from the screen definitions file, and the template is then served using those parameters. The MainServlet does not do all of the work. Most of the work is delegated to other classes that specialize in screen flow management, template instantiation, and the like.
The XML files in the templating system are parsed by a standard XML parser, created using the JAXP (Java API for XML Parsers) optional package. The template is served as a JSP page, and the inclusion of the dynamic content occurs because the JSP engine calls the custom tag <j2ee:insert> each time it occurs.
In a Web application, each screen presented to the user can be considered as a different view. However, unlike the classic MVC architecture, all these views share the same controller. There needs to be a mechanism that allows the controller to choose a particular view to render in response to a user request. In the sample application, the controller makes this selection by specifying the screen ID of the screen to present as the response. This screen ID is mapped to a screen definition, then the template is instantiated. Recall that the file template.jsp defines the template for the sample application. This file includes another file, ScreenDefinitions.jsp, which defines all the screens of the sample application. When the controller invokes the template file at request time, it sets the appropriate screen definition in the request scope. The template file passes this information to the screen definitions file which then returns the appropriate screen definition for the request. One goal in structuring template and screen definition files is to facilitate internationalization (discussed in Section 4.5). This is achieved by separating text content from Java code. Since screen definitions that contain direct and indirect parameters are candidates for internationalization, we want to keep ScreenDefinitions.jsp devoid of Java technology code. We achieve this through the use of JSP custom tags.
Data Access Mechanism
The application uses a bimodal data access mechanism to increase efficiency of data access. Since it is anticipated that the majority of data access will be read-only (50% or more, as a conservative estimate), the data access mechanism will not utilize a transactional component for reading, relying instead on direct JDBC access to reduce overhead. Conversely, data updates will still continue to rely on a transaction component. When a transactional update does occur, the model will be refreshed from the data store and a model update event propagated to all registered views to update themselves, ensuring data display consistency.
Data Synchronization Between User's PC and Server
The functional specifications for the data synchronization mechanism are listed below, followed by details on each of the components that implement the mechanism. In cases where the implementation details of a component is not yet decided (such as whether to use an applet or not on the client PC), it is referred to generically (e.g. data sync client).
When a user logs on to the home page (
For this phase of implementation, the locations of the control file and the data file on the user's PC is fixed. Note the following:
Device Control File
This file contains records in the format
User Data File
This file contains records in the format
Data Synchronization Client
This client runs on the user's PC to handle the file operations to and from the server such as uploading and downloading, checking for file existence, creating backup files etc. Note that these are all file operations and do not require any business logic in this client. All of the business logic decisions are made at the server and the appropriate files are downloaded. This allows for the client to be implemented as a signed applet.
A standalone Data Synchronization Client that does not require a PC and is directly able to connect to the server via a network connection is developed by embedding the synchronization logic in the firmware of a microcontroller and using it to connect to the server through a connection established either via a phone line (which requires an embedded modem) or via a connection established by another network terminal such as a web-phone or video-phone to which the Data Synchronization Client is connected.
Data Synchronization Manager
The Data Synchronization Manager component runs on the server and is the primary interface between the Data Synchronization Client and the rest of the application. Its primary operations include: receiving upload of the data file from Data Synchronization Client; decrypting the data records; determining which user each record belongs to (from RegisteredDevice mapping); putting each user's respective records in staging area (table) for user preview if so specified. This function is actually delegated to the ReadingsManager; putting each user's respective records in data tables if user does not want preview. This function is actually delegated to the ReadingsManager; signaling Data Synchronization Client to delete data file; receiving upload of control file; if control file does not exist, then creating it initially; updating control file with last uploaded timestamp for each registered device; downloading updated control file to Data Synchronization Client.
The Data Synchronization Manager has the following invocation points:
The Data Synchronization Manager is implemented as a Session BO as well as a standalone server application known as the iModem server.
IModem Server and Protocol Specification
The key attributes and mechanisms of the iModem server (i.e. Data Synchronization Manager for Standalone Data Synchronization Clients) are:
Accept multiple connections over the same server port at the same time: This will happen (in production) since more than one iModem could be connecting to the server and will require that the server be multi-threaded, using an independent thread for handling each connection. This is somewhat analogous to how web servers handle HTTP, which is a stateless protocol wherein several clients connect over the same port (80) at the same time. This also implies that the port number to which the iModem connects will remain constant.
Build in a versioning scheme into the protocol: This allows for upgrades over time without breaking backward compatibility.
Provide a status value in each message response. Zero indicates OK, non-zero values indicate various error conditions.
Error handling: There must be enough error handling and recoverability that a user's data is never lost and that each end (iModem or server) can signal to the other that data transfer either occurred or not. There must also be timeouts to handle conditions when no response is achieved. The value of the timeout (in seconds) must be a parameter read from a configuration XML file.
Break down current messages into multi-step messages: This is to handle two basic limitations: 1) each message to and from the iModem must be less than 256 bytes and 2) the iModem embedded code is more efficient in parsing fixed-length responses rather than variable length responses.
Encryption and keys: A basic mechanism (using placeholders as necessary) needs to be put in place such that in case the data stream is intercepted by a third party, all information that is private to the patient (such as reading values) and all information that can be correlated to deduce the message format (such as device identification information) is scrambled or encrypted.
This servlet is a “listener” for an embedded internet-modem (iModem) that has established a dial-up connection with a PPP server and is then communicating over TCP/IP to a specific server and port. The primary purpose of this servlet is to accept TCP connections over the specified port, parse the data coming over the connection and return responses. It also does basic connection management of connections and sockets as necessary. The incoming data stream comprises a series of messages in a simple protocol detailed below. This servlet parses the data stream, decodes the protocol message and responds based on logic detailed below. In summary, this servlet serves as a protocol handler for the iModem hub data transmission, which is done over TCP/IP sockets.
The data protocol is a variant of the data exchange (XML) protocol that is used by the main web application for data interchange (defined by the files devicecontrol.xml and readingsdata.xml, hereafter referred to as the Datasync protocol). It differs from the Datasync protocol in the following manner to allow for the limited processing requirements of the iModem hub:
Numeric values are specified as hexadecimal text with each byte value comprised of 2 ASCII characters i.e. a decimal value 1085, which is 43D hex (or 043D hex, more accurately), will be specified as the characters ‘0’, ‘4’, ‘3’ and ‘D’. The most significant byte (MSB) is first when decoding multi-byte values.
Descriptive tags in the Datasync protocol are replaced with 4-digit hexadecimal numbers referred to as tokens. E.g. a <control-entry> tag may be denoted as <04AD>.sup.1. The mapping of the tags in the Datasync protocol to the iModem protocol must be configurable via another XML file (tagmappings.xml) so that it may be changed as the protocols evolve. End tags are denoted by a 4-digit hex number that is calculated by setting the most-significant-bit (MSB) of the number denoting the start tag. This defines the start tag and end tag relationship of the iModem protocol..sup.1 The angle braces < > will still be used to surround tags
Readings are sent serially as a series of Readings Data messages (see below). This is given the current limitation of the iModem restricting the message length to be less than 256 bytes. The server response is either a success or error indicator.
All timestamps are in seconds elapsed since Jan. 1, 1970, denoted as a 4-byte number and encoded as the ASCII representation of the hex digits, MSB first. There is an additional command for getting the current server time that is not part of the Datasync protocol. The details of this message are defined below, with a descriptive tag being used in place of a token for illustration purposes.
The HubID is an identifier that is encoded in the iModem hub and is transmitted as a means of identifying itself. It is a long i.e. 4 byte number. This ID will be populated in the database through the device registration UI (which will be enhanced to support this later) and is used to associate a user (and his registered devices) with a specific hub.
On receiving this message, the server should set up a “session” with this connection and Hub ID and return a status code of 0000 in the response. All subsequent messages coming over this connection are considered in the context of the established session. If an unrecognized Hub ID is received, a status code of 0001 is returned with other message data being same. The <protocol-version> value is two byes, with the MSB being the major version and the LSB being the minor version. So currently, this value would be 0101 (current version is 1.1=>1<<8+1=257 decimal=0101 hex, formatted as 2 characters for each byte).
The response returned is the current server time, in seconds elapsed since Jan. 1, 1970. It is a 4-byte number and is encoded as the ASCII representation of the hex digits, MSB first.
There are two additional fields added for version 1.1. These are the update-flag and key. <update-flag> is a byte value that is used to signal to the iModem that there are parameter updates (see message Parameter Updates) that it needs to query for. The intent is to provide a way for parameters such as dial out numbers, dialing prefixes etc. to be provisioned via a web interface and indicated to the iModem so that it can update its own memory-based parameters. For version 1.1, the value returned is always 00 (zero).
The <key> is a placeholder value for now. It is a 16 bit number and its value is always 0000 (zero).
This is a placeholder message for now and will be expanded in future protocol revisions. The status code is always 0000.
This message is a fixed-format equivalent to the readingsdata.xml file containing a single reading. The appropriate token values (as set up in tagmappings.xml file) are used for <reading-type> and <units> tag values.
The <status> value is a 16-bit integer. The value 0000 (zero) indicates no error whereas non-zero values indicate error conditions. Currently, only the value 0001 is assigned as a general error indicator. This list of values will be expanded over time and the code should be written to allow for it without significant change.
Because the iModem requires a fixed-length, fixed-format data command, the server must do some conversion/transformation to derive the application-specific XML <readings-data> command. The iModem will always send all data (BP, weight or blood sugar) in the format described above (for BP). The only difference for weight and blood sugar will be that the <reading-type> and <units> will correspond to the type of reading, so that for weight, <reading-type> value is WEIGHT and <units> value is POUNDS and for blood sugar, <reading-type> is BLOODSUGAR and <units> value is mg/dL. The value of the reading is derived from the byte values contained in the <bp-systolic>, <bp-diastolic> and <pulse-rate> tags. Because of this, the <value> tag is no longer required.
The server must use the following algorithm/logic to convert these fixed format iModem <readings-data> commands to the Datasync protocol's <readings-data> messages:
Number of Devices
This command is used to obtain the number of devices currently registered by the user. The status is generally expected to be 0000, non-zero indicates an unspecified (as of yet) error.
The iModem sequentially requests device details for each device registered by the user, passing in a device number (actually, device index) as part of the command. The list is zero-indexed i.e. the first device is 0 (zero) and the index increments sequentially from there. The information in the response is the same as a single-entry device control file. Currently, nothing is returned for reminders (even if there are some set) since the iModem is not set up to handle them. These will be handled in a subsequent protocol revision.
This command does not return a response.
This command is for the iModem to indicate to the server that it is done with the session prior to terminating the connection so that the server may release resources and perform other cleanup. The server does not return a response to this command since the client's state is undefined after this. It releases resources and terminates the server's socket connection.
Reminders are in a slightly different format from the Datasync protocol. The structure is as follows:
Chart Creation and Display
The chart creation and display mechanism is comprised of collaborating objects and components across all four application tiers (view, web, business object and data persistence tiers). As previously mentioned, Popchart is the charting server being used to render the data for display in the Flash format. This requires the Flash plug-in to be installed in the client browser, which the majority of the browsers (over 90%) already do.
The approach to chart creation and display is to isolate dependency on Popchart in the view tier only. This is to allow change over to another charting product if required later. Doing this requires encapsulating all code to create a Popchart chart in presentation components in the view tier. All the information (data) required to produce the chart is kept as generic objects or components and supplied to the view components via the standard application framework. This mechanism is elaborated upon below.
The View Tier
This tier is comprised of the JSP page containing the chart as well as an embedded JavaBean that converts the chart model objects into Popchart specific instructions (there may be the need to do specific JavaBeans for each chart type but the approach will remain the same for each). This JavaBean encapsulates the logic to create Popchart instructions from the chart model objects and isolates the dependency on Popchart.
The Popchart JSPExample demonstrates such a JSP/JavaBean combination. The key points of difference are:
The Web Tier
This tier contains the standard web handler for processing user input to the chart (such as when a user displays the chart, changes the date range, clicks on a point for annotation or drill-down etc.). The handler creates a chart event object (class ChartEvent) encapsulating the user request and forwards it to the Business Object tier for handling.
Note that there may be requests that can be handled in the web tier itself instead of having to go to the BO tier. For example, the user may choose to view a data range that is a subset of the range already displayed. In this case, there is no need to go to the BO tier to get data for display since the data has already been sent to the web tier in the chart model object(s). The web tier handlers need to be intelligent enough to check for these cases in order to improve application performance.
The Business Object Tier
This tier contains the components required to handle chart events generated from the web tier. The main facade component is a session BO named ChartManagerSBO. This BO interprets the chart event and creates/retrieves the chart model objects that are required to generate the requested chart. It invokes the ReadingsManager to get the actual readings model objects. This BO also provides the defaults for chart presentation.
The chart model objects also belong in this tier. These objects encapsulate the complete information required to render the chart. This includes information about the various chart display parameters as well references to the model objects that hold the actual data for display. The data used to populate the chart model is obtained from the chartconfig.xml file using a DAO implementing the ChartDAO interface.
The chart model object is mutable i.e. it can be changed when a user adds or changes annotations on data points. The annotation actually changes a particular reading referenced by the range, which is in turn referenced by the chart model object.
Popchart servlet redirector may be used.
The implementation DAO object actually is ChartXmlDAOlmpl, meaning that it reads the XML configuration file.
The elements in the chart model object that are specific to the charting product are kept as properties so as to not make the structure of the model object dependent on product-specific attributes.
Following are the design patterns utilized in the architecture. Further details can be found in publicly available J2EE Blueprints documents as well as in a book, entitled “Design Patterns”, by Vlissides, Gamma, et al and are thus elaborated upon further here.
Data Access Object (DAO)
Bimodal Data Access
Device Control File
This file is currently in plain text XML format. It may later be stored encrypted if necessary. It does not currently have a DTD and therefore requires a non-validating parser. For now, the URI for this file is file://c:.backslash.carematix.backslash.config.backslash.devicecontrol.xml.
User Data File
This file is currently in plain text XML format. It may later be converted into a format in which each XML <reading-record> entry is encrypted and stored as a fixed length record, with new records being appended to the end. It does not currently have a DTD and therefore requires a non-validating parser. For now, the URI for this file is file://c:.backslash.carematix.backslash.data.backslash.readingsdata.xml.
Note: This file may contain duplicate <reading-record> entries. That is allowed.
Chart Configuration File
This file contains the necessary information for formatting and displaying data charts. This information is kept in XML format neutral of the specific charting product so as to allow changeover to another charting application if necessary later. It does not currently have a DTD and therefore requires a non-validating parser. This file is stored on the server, not the client and is called chartconfig.xml
Please note the following: this format is an initial version and will likely be enhanced over time. The method for altering and enhancing chart display is to add tags and/or properties in the sections for each chart and use the values assigned to each to drive the behavior of the code. This is preferred over hard-coding the way a chart is displayed in the code itself.
In subsequent phases, this file may have an XSL transform applied to generate the product specific charting instructions e.g. if using Popchart, an XSLT would be provided to generate PCScript from this XML file.
Obviously, many modifications and variations of the present invention are possible in light of the above teachings. Thus, it is to be understood that, within the scope of the appended claims, the invention may be practiced otherwise than as specifically described above.