method and apparatus for creating a report

[0001] The present invention generally relates to document generation, and in particular to creating a structured report and creating an expanded report from that structured report.

[0002] In many industries, particularly in the healthcare industry, technicians or specialists evaluate data and generate written reports to be utilized by others in providing a service or product. Other individuals refer to the written reports to provide the appropriate service or product to the consumer. The process of creating the written reports, however, is a time-consuming and expensive process.

[0003] For example, radiologists diagnose diseases by analyzing radiological images and generate written reports describing their findings. Generally, the radiologists dictate their findings via a Dictaphone. The tapes are picked up by transcriptionists and taken off-site for transcription. Approximately 24-48 hours later, the transcribed reports are returned to the radiologist for corrections. The marked reports are picked up by the transcriptionists and taken off-site for correction. The radiologists and the transcriptionists continue in this manner until all corrections have been incorporated into the transcribed report. Finally, the radiologist signs the corrected reports, and the corrected reports are made available for distribution via runners, who typically hand-carry the corrected reports to the requesting physicians.

[0004] Technologies such as the fax and e-mail have eliminated the step of physically picking up the reports, thus, decreasing the amount of time required to produce the corrected reports. These technologies, however, failed to significantly impact the dictation and transcription time cycle.

[0005] Tele-transcription systems reduce the dictation and transcription time because the systems allow radiologists to pick up an ordinary phone and dial into a dictation system provider and report their cases. These systems eliminate the runners and cut transcription turnaround time significantly, to about 12-24 hours. Corrections to the reports, however, are still time-consuming.

[0006] Some systems utilize voice recognition technology in an attempt to reduce the dictation/transcription time. These voice recognition systems, however, have to be trained to each speaker's specific voice recognition patterns, and the accuracy rates are too low for many applications, including radiology in which an accuracy rate of 95% is not acceptable. Corrections in many of these systems are even more cumbersome, and the systems are expensive. Although the voice recognition system provides many advantages, such as the electronic distribution of reports, the disadvantages discussed above limit its usefulness.

[0007] Other systems utilize menu-based systems that provide a template for a report that contains a number of fields. Each field is filled in by the radiologist. Some fields provide a cascading set of menus that enable the radiologist to create a structured set of sentences. In these fields, the sentence is derived from the radiologist's selections in the menus. After each selection, the output sentence is updated and the radiologist continues to make selections until the output accurately states the radiologist's findings. This approach, however, has a number of shortcomings. First, each report field has its own input tree and selections in that tree create output in a single report field. This forces the reporting radiologist to create the report in the precise order in which the data appears in the output fields. Second, once a menu item has been selected, it is used to construct a sentence in the corresponding field of the report. Generally, the same menu item is not allowed to be utilized to construct multiple sentences. Third, when a particular subtree is used in many locations, it must be copied into each location of the template. Therefore, when it requires changes later, each copy must be updated. Fourth, text output is not context sensitive, which makes it difficult to generate a natural sounding report for certain phrases, such as when numerical adjectives correspond in number with the head noun in a noun phrase. Fifth, the output ordering is determined solely by the input tree structure. The order in which items from a single menu are selected does not have any affect on the output ordering.

[0008] Furthermore, the systems described generally do not provide the reports in an electronic form that enables a user to search across reports and create reports based on the content of a plurality of previously completed reports. Rather, the systems generally only provide the ability to search for specific strings of text contained within a report.

[0009] Therefore, the present invention advances the art by providing an improved technique that allows users to generate written reports.

[0010] To overcome the limitations in the prior art described above and to overcome other limitations that will become apparent upon reading and understanding the present specification, the specification discloses a method, apparatus, and article of manufacture for creating a structured report on a computer.

[0011] In accordance with an embodiment of the present invention, one or more nodes of a decision tree are displayed to a user. Information about the displayed nodes are then received from a user. The information includes at least one selected node. The set of selected nodes constitutes a structured report, because the nodes are arranged hierarchically and there is an explicit structure imposed on the data entered by the radiologist. The text report is then dynamically created by converting the selected nodes into one or more sentences.

[0012] In another embodiment of the present invention, the structured report is stored in a database or file that is connected to the computer.

[0013] In another embodiment, the structured report is electronically distributed via the Internet.

[0014] The invention has utility for dynamically creating a structured report that can be instantaneously accessed.

[0015] Referring now to the drawings in which like reference numbers represent corresponding parts throughout:

[0016] FIG. 1 is a diagram of a network environment that embodies features of the present invention;

[0017] FIG. 2 is a flow chart illustrating the relationship between components of an embodiment of the present invention, including the way in which data is passed between components;

[0018] FIGS. 3a and 3b illustrate a structured report in accordance with an embodiment of the present invention;

[0019] FIG. 4 is a flow chart illustrating the process of creating an expanded report from a structured report in accordance with an embodiment of the present invention;

[0020] FIG. 5 is a flow chart illustrating one embodiment of the present invention in which nodes of a decision tree are sequenced in accordance with the properties of the nodes;

[0021] FIGS. 6a-6b illustrates one embodiment of the present invention in which a contained set is utilized to determine the sequence of nodes in an expanded report;

[0022] FIG. 7 is a flow chart illustrating one embodiment of the present invention in which a decision tree is converted into a structured report; and

[0023] FIG. 8 illustrates one embodiment of the present invention in which a findings decision tree is displayed and an expanded report is dynamically displayed.

[0024] In the following description, numerous specific details are set forth to provide a thorough understanding of the present invention. However, it will be obvious to those skilled in the art that the present invention may be practiced without such specific details. In other instances, well-known elements have been illustrated in schematic or block diagram form in order not to obscure the present invention in unnecessary detail. Additionally, details concerning programming, communications, and database structures, and the like have been omitted inasmuch as such details are not considered necessary to obtain a complete understanding of the present invention, and are considered to be within the skills of persons of ordinary skill in the relevant art.

[0025] It is further noted that, unless otherwise indicated, all functions described herein may be performed by either hardware or software, or some combination thereof. In a preferred embodiment, however, the functions are performed by a processor such as a computer or an electronic data processor in accordance with code such as computer program code, software, and/or integrated circuits that are coded to perform such functions, unless otherwise indicated.

[0026] The present invention provides a method and an apparatus for generating a structured report and, from that structured report, generating an expanded report. As one of ordinary skill in the art will appreciate, the present invention applies to any area in which a report is to be generated for use by other individuals or groups. For example, an embodiment of the present invention may be utilized in radiological, toxicological, pharmaceutical, other healthcare specialties, automotive, electronics, construction industries, or the like. For illustrative purposes only, however, the following description provides examples specific to radiology and reports generated by radiologists. A person of ordinary skill in the art, however, will appreciate that the examples are equally applicable in any industry.

[0027] As will be discussed in greater detail below, an embodiment of the invention preferably allows a user to develop report templates that define possible observations that a person may observe. After the report template is created, another user, such as a technician or specialist, observes an event and selects the appropriate observations in the report templates, creating a structured report. Grammar rules are applied to the structured report to create an expanded report. Preferably, the structured report and/or the expanded report may be electronically transmitted to other individuals or groups that require the structured or expanded reports to provide a service or product.

[0028] For example, an embodiment of the present invention provides a method and an apparatus that allow radiologists to develop report templates that define possible observations that a radiologist may observe for a particular type of radiological study, such as a chest x-ray or a CT scan of the brain. Furthermore, an embodiment of the invention preferably allows radiologists to utilize the report templates to record observations and diagnoses for a particular study, such as a chest x-ray for patient 1234 on Jul. 15, 2002, at 9:34 A.M. These observations and diagnoses are captured in a structured report. Optionally, the structured report may be converted to an expanded report by applying grammar rules. Preferably, the expanded report contains information in an easy to read format necessary for a referring physician to determine the results of the radiological study and treat the patient accordingly.

[0029] Referring to FIG. 1 of the drawings, the reference numeral 100 generally designates a portion of a reporting system, which embodies features of the present invention. Generally, the reporting system 100 comprises one or more interface devices 110, such as appliances, personal digital assistants (PDAs), wrist watches, television systems, mainframe computers, minicomputers, personal computers, workstations, handheld computers, cell phones, or the like. The interface device 110 preferably includes a processor, one or more data communications devices (e.g., modems, network interfaces, etc.), a monitor (e.g., CRT, LCD display, etc.), and an input device (e.g., a mouse and/or a keyboard). It is envisioned that attached to the client computer may be other devices such as random-access memory (RAM), read-only memory (ROM), a video card, bus interface, printers, and the like. The interface device 110 is configured to allow a user (not shown) to enter data regarding observations that the user made regarding a subject (not shown).

[0030] Optionally, the interface device 110 is coupled to a server 120 via a network 130 to enable sharing of templates, generated data and reports. Accordingly, the server 120 is preferably a server accessible via an external network, such as the Internet, to which the interface device 110 accesses via a communications link, such as a TCP/IP communications link, wireless communications link, or the like. The interface device 110 preferably accesses the server 120 through a standard web browser, such as Microsoft Corporation's Internet Explorer, utilizing standard communications protocols and document standards, such as hyper-text mark-up language (HTML), extended mark-up language (XML), or the like. Other configurations, such as local area networks (LANs), wide area networks (WANs), public networks, private networks, and the like, may be utilized depending upon the application and the desired accessibility. Other components, such as a document storage 140 are preferably accessible via an internal network 150. For security purposes, server components other than the web server 120, such as the document storage 140, are typically placed on an internal network 150 behind an internal firewall (not shown).

[0031] FIG. 2 illustrates the process of generating a structured report and an expanded report in accordance with one embodiment of the present invention. The processes and the data may be implemented on the interface device 110, the web server 120, the document storage 140, or some combination thereof. One of ordinary skill in the art will appreciate that an embodiment of the present invention may be implemented either in a stand-alone environment or a distributed environment.

[0032] Accordingly, a Template Designer 210 preferably provides a user interface for a designer to create a Report Template 212, i.e., a definition of possible observations that may be made regarding a particular study of a subject. Preferably, the Report Template 212 is presented by the Template Designer 210 as a hierarchical collection of nodes referred to as a decision tree, wherein each node of the Report Template 212 represents a group/classification of related items or specific alternatives regarding observations of the subject of the report. Accordingly, the hierarchical tree structure of the Report Template 212 comprises parent nodes and child nodes. Parent nodes are nodes that contain other nodes, and child nodes are nodes that have one or more parent nodes. Generally, nodes may be both a parent node and a child node. Furthermore, nodes of the Report Template 212 are preferably expandable and collapsible to provide an easy-to-use interface.

[0033] Additionally, the Template Designer 12 preferably provides editing functions, such as Add Node, Delete Node, Copy Node, Paste Node, and the like, with which the user creates a tree structure and specifies the contents of the Report Template 212. The Template Designer 210 displays a property sheet for setting the values for one or more properties of a node. Generally, a property sheet provides a mechanism that enables the user to specify the display and selection behavior properties of a node. The preferred Property Name, Data Type, and Description of the properties of each node of an embodiment of the present invention are described below in Tables 1-5, wherein Table 1 contains General Properties, Table 2 contains Relationship Properties, Table 3 contains Input Behavior Properties, Table 4 contains Output Behavior Properties, and Table 5 contains Traversal Properties. As one of ordinary skill in the art will appreciate, however, the preferred properties and their usage may differ depending upon, among other things, the intended use, the intended audience, the industry, level of expertise, and the like.

[0034] The following Tables 1-5 illustrate the preferred Property, Data Type, and Description associated with the nodes discussed above. One of ordinary skill in the art will realize that different node properties may be required dependent upon the implementation and the intended application. For example, differing data types, values, additional properties, fewer properties, or the like may be required in different implementations.

TABLE 1
General Properties
	Data
Property	Type	Description
Display	String	The name of the node to be displayed in the
Name		Findings Editor. Default is NewNode.
UID	String	The unique identifier for this node.
Creation	Date	The date and time when the node was created.
Date/Time
Modified	Date	The date and time when the node was last modified.
Date/Time
Effective	Date	The beginning date and time when the node may be
Date/Time		used. Prior to this date, the Findings Editor should
		not display this node. Default is the creation date.
Expiration	Date	The last date and time when the node may be used.
Date/Time		After this date, the Findings Editor should not
		display this node. If no date is specified, the node
		will never expire. Default is no date.
Derived	String	The unique identifier of the node from which this
from		node was derived. A node may be derived from
template		another node at design time, either by copying or by
UID		creating an alias. Default is the empty string.
URL	String	A universal resource locator that specifies additional
		reference information related to this node that may
		be displayed in the Findings Editor. Default is the
		empty string.
Tip	String	A short string that may be displayed in the Findings
		Editor to help the user understand the purpose of
		this node. This is help text that is displayed
		when the cursor hovers over the node. Default is
		the empty string.

[0035]

TABLE 2
Relationship Properties
Property	Data Type	Description
Parent	Node	The UID of this node's parent node. If this is the root
		node, the Parent will display as the empty string.
Children	List of Node	The labels of this node's children, if any.
Only one child	Boolean	If true, the Findings Editor will allow the user to
selected		select no more than one of this node's children. This
		property is mutually exclusive with All children
		selected. Default is false, in which case the user will
		be allowed to select more than one child.
At least one child	Boolean	If true, the Findings Editor will require the user to
selected		select at least one of this node's children. Default is
		false.
Children's order	Boolean	If true, the Value derived from this node's children is
fixed		determined by the order in which they appear in the
		decision tree. Otherwise, their values are aggregated
		in the order in which they are selected in the Findings
		Editor. Default is true.
Delimiter	Enumeration	Space, Comma, And. Specifies the way in which
		children's values are to be combined. This delimiter is
		inserted between the values of the children. Comma
		indicates that the child values are separated by a
		comma; And indicates that the conjunction and is
		inserted before the last of a series of child values.
		Default value is Space, in which case child values are
		separated by a space.
Co-existence	List of Node	If this node is selected in the Findings Editor, then the
Nodes		co-existence nodes must also be selected.
Exclusion Nodes	List of Node	If this node is selected in the Findings Editor, then the
		exclusion nodes may not be selected.
Alias Template	String	Reference to another template whose root node is to
		inserted into the current template at this location in
		the tree.

[0036]

TABLE 3
Input Behavior Properties
Property	Data Type	Description
Data Source	Enumeration	Prompt the user, Get from node, Get from database.
		Specifies the source of data to be placed into the Value
		property at runtime in the Findings Editor. Prompt
		the user indicates that the user supplies the value.
		Get from node references the value of another node.
		Get from database gets the value from a database.
		Default is Prompt the user.
Prompt	String	The string to be displayed by the Findings Editor as a
		prompt to request the user to enter a value.
Parameter Name	String	The name of the parameter to be displayed in the
		design tool. Default is the empty string.
Label	String	The name of the parameter to be displayed in the
		Findings Editor. Default is the empty string.
Data Type	Enumeration	String, Date, Number. The type of data to be
		entered by the user. The Findings Editor should
		validate user input against this type. Default is
		String.
Format	String	The string may refer to a parameter, e.g. $size in
		cm cm, where size in cm is the value entered by
		the user, but something like 4 cm will be the value of
		this node. Default is the empty string.
Required	Boolean	If true, the user must enter a non-null value in the
		Findings Editor.
Maximum	String	The maximum value that may be assigned to this
		parameter. If the user tries to enter a value that is
		greater than this value, the Findings Editor will not
		accept the value. Default is the empty string.
Minimum	String	The minimum value that may be assigned to this
		parameter. If the user tries to enter a value that is less
		than this value, the Findings Editor will not accept the
		value. Default is the empty string.
Usual Maximum	String	The maximum value that may be assigned to this
		parameter. If the user tries to enter a value that is
		greater than this value, the Findings Editor will warn
		the user. However, if the user confirms the value, then
		it will be accepted. Default is the empty string.
Usual Minimum	String	The minimum value that may be assigned to this
		parameter. If the user tries to enter a value that is less
		than this value, the Findings Editor will warn the user.
		However, if the user confirms the value, then it will be
		accepted. Default is the empty string.
Default Value	String	The default value to be supplied by the Findings
		Editor when prompting the user for a value. Default is
		the empty string.
Value	String	The value entered by the user through the Findings
		Editor. Default is the empty string.
Pre-functoids	String	The name of a pre-functoid, which is a function that is
		performed before the user enters a value in the
		Findings Editor.
Post-functoids	String	The name of a post-functoid, which is a function that
		is performed after the user enters a value in the
		Findings Editor.

[0037]

TABLE 4
Output Behavior Properties
Property	Data Type	Description
Is Required	Boolean	If true, the report tool will require the user to enter a
		value. Default is false.
Selection	Integer	The relative order in which this node was selected with
sequence		respect to its siblings. This number is needed in the
		case where child values are listed in the order in which
		the children were selected in the Findings Editor.
Value	String	The value to be displayed for this node in the report.
Is Sentence	Boolean	If true, the Value of this node is to be treated as a full
		sentence in the report. Otherwise, the Value will be
		treated as a phrase that modifies its parent. Default is
		false.
Is Paragraph	Boolean	If true, insert a paragraph boundary before printing
		the value of this node. Default is false.
Triggers number	Enumeration	None, Singular, Plural. If Singular, the singular
agreement		Value will be used from siblings. If Plural, the Plural
		Object Value will be used from siblings. If None, this
		node has no influence on the choice of values from
		sibling nodes. Default value is None.
Plural value	String	If the Object Value is a noun, then the plural form of
		the word should be specified. For example,
		mass/masses, tibia/tibiae, congestive heart
		failure/congestive heart failures.
Object placement	Enumeration	Prepend to parent's object, Replace parent's object,
		Append to parent's object. Default value is Append
		to parent's object.
Pre-functoids	String	The name of a pre-functoid, which is a function that is
		performed before the user enters a value in the
		Findings Editor.
Post-functoids	String	The name of a post-functoid, which is a function that
		is performed after the user enters a value in the
		Findings Editor.

[0038]

TABLE 5
Traversal Behavior Properties
Property	Data Type	Description
Is Selected	Boolean	If true, the node has been selected either explicitly by
		the user or implicitly by inclusion (see the Co-
		Existence Nodes property in the Relationship
		properties). Otherwise, the node has not been
		selected, so it should not appear in the findings report.
Is Repeatable	Boolean	If true, the node may be selected more than once in
		the Findings Editor, in order to generate a sequence of
		values. Otherwise, the node may only be selected once
		to generate a single value. For repeating nodes, only
		the Value is included in subsequent occurrences of the
		node. The Prefix Value and Suffix Value are not
		repeated for each occurrence of the repeating node.
		Only the Prefix Value from the first occurrence is used
		and only the Suffix Value from the last occurrence.
		Default is false.
Number of times	Integer	The maximum number of times that a node may be
		repeated in the Findings Editor. Default is o.
Repeat Traversal	Integer	A sequence counter that indicates which instance this
Count		note is in a sequence of repetitions.
Get from Node	Node	This node gets the number of times to repeat from a
		number value specified in the given node.
Children Before	Enumeration	None, Auto expand, Auto collapse. Specifies the
Traversal		display of child nodes in the Findings Editor before
		this node has been selected. Auto expand causes the
		child nodes to be displayed even before this node is
		selected. Auto collapse causes the child nodes to not
		be displayed until the user explicitly selects this node.
		Default is None.
Children During	Enumeration	None, Auto expand, Auto collapse. Specifies the
Traversal		display of child nodes in the Findings Editor when this
		node is selected. Auto expand causes the child nodes
		to be displayed when this node is selected. Auto
		collapse causes the child nodes to not be displayed
		even when the user explicitly selects this node.
		Default is Auto expand.
Children After	Enumeration	None, Auto expand, Auto collapse. Specifies the
Traversal		display of child nodes in the Findings Editor when
		some other node is selected that is not a descendent of
		this node. Auto expand causes the child nodes to
		remain displayed when the other node is selected.
		Auto collapse causes the child nodes to not be
		displayed when the user explicitly selects outside of
		this node's descendents. Default is None.
Hide Node	Boolean	If true, this node is not displayed by the Findings
		Editor. This node's value will still appear in the
		output, although the node itself will not be rendered
		through the Findings Editor. Default is false.
Hide Children	Boolean	If true, this node's children will not be displayed by
		the Findings Editor. The children nodes' values will
		still appear in the output, although the nodes
		themselves will not be rendered through the Findings
		Editor. Default is false.
Hide all	Boolean	If true, none of this node's descendents will be
descendants		displayed by the Findings Editor. The descendent
		nodes' values will still appear in the output, although
		the nodes themselves will not be rendered through the
		Findings Editor. Default is false.
Expansion Style	Enumeration	Tree, System, None, CSS, Horizontal Radio Group,
		Vertical Radio Group, Horizontal Checkbox Group,
		Vertical Checkbox Group, Pull-down List, List, Table,
		Input Box, Text Box. Specifies the type of control
		used by the Findings Editor to display this node's
		children. Default is Tree.
Selectable by user	Boolean	If true, then the user can select the node at runtime.
		Otherwise, the user cannot manually select the node.
		However, regardless of the value, the system itself can
		select this node, e.g. via co-existence or selection by
		inference.

[0039] Additionally, the Template Designer 210 may embody procedure codes and diagnostic codes in the Report Template 212. Taxonomies embody knowledge on procedures, clinical information, techniques, limitations and comparisons. As users navigate these taxonomies and make selections to describe the actions that were performed during a procedure, their selections in the template triggers the system to automatically generate the corresponding procedure codes (CPT codes) and diagnostic codes (ICD codes) for the report.

[0040] This approach for automatically generating procedure codes and diagnostic codes via the embodiment of knowledge on procedures, clinical information, techniques, limitations, and comparisons is applicable to many applications. For example, procedure codes and diagnostic codes are applicable in most fields of medicine, such as radiology, cardiology, pathology, surgery, and the like.

[0041] As indicated in the node attributes for Relationship Properties in Table 2, a node may contain a reference to another template. This enables a report template to contain another, so that a single source template may be used multiple times by other templates. Utilizing this technique, the Template Designer 210 provides a way for the user to insert an alias reference wherever a node may be placed. The alias reference is a reference to another template. The Template Designer 210 and the Findings Editor 214 (discussed below) interpret that reference, load the referenced template, and treat the alias as though it were a sub-tree of nodes that had been defined directly in the current template. A source Report Template 212 may be referenced any number of times by another template and it may be referenced any number of times by any number of templates.

[0042] Preferably, the Report Template 212 is represented as a hierarchical model of decision tree nodes that may be selected by the application user. Preferably, the user selects one or more nodes of the decision tree via a mouse click, keyboard event, voice command, or the like. The preferred embodiment is to represent Report Template 212 as an XML document. However, there are other embodiments that do not depart from the scope of the present invention. Such alternative embodiments of the Report Template may include, but are not limited to, marked-up text as in an XML document, objects in an object-oriented database, objects serialized into a binary stream and saved in a file or database, rows and columns of data in a relational database, classes for node data and behavior implemented in an object-oriented programming language, data structures and functions for node data and tree operations in a procedural programming language, a standalone component for representing and managing a tree, such as an XML parser that adheres to the Document Object Model (DOM) Application Programming Interface (API), or the like. Furthermore, the Report Template 212 is preferably displayed graphically as nodes of a hierarchical decision tree. Other embodiments, such as a static hierarchical tree display, cascading menus, columns of data where each column represents the nodes at a particular level of the hierarchy, an interactive hierarchical tree display that the user may expand or collapse nodes, a list of items, or the like, may also be utilized.

[0043] For illustrative purposes only, FIGS. 3a and 3b illustrate portions of a Report Template 212 that may be created by the Template Designer 210 for use in radiological studies in accordance with an embodiment of the present invention. A , such as that displayed adjacent to group 310 labeled Demographics, indicates that the node is collapsed, and a , such as that displayed adjacent to group 312 labeled Findings, indicates that the corresponding node is fully expanded, showing all child nodes of the parent node. This collapsing and expanding of nodes preferably may occur at any level and is a common mechanism employed by programs such as Microsoft Corporation's Windows Explorer program.

[0044] Groupings, such as groups 314-330, provide the template designer a mechanism to create a hierarchical structure, allowing a user, i.e., one who is generating a report, to readily locate the appropriate area of concern. FIG. 3b illustrates one of the hierarchical structures for the group 340 labeled chest. As one of ordinary skill in the art will appreciate, the user is able to create a template with varying groups and identifications, down to the lowest level, such as the groups 340 and 342 labeled right and left, respectively, as the lowest members of the group 344 labeled shift, which is a member of the group 346 labeled mediastinum. This process is repeated to define the relevant alternatives for the various groupings defined by the user.

[0045] A selection box, such as selection box 340, provides a mechanism to allow the user to indicate the observations relevant to a study of a subject. The purpose and function of the selection box is discussed in greater detail below with reference FIG. 2 and the Findings Editor 214.

[0046] Referring back to FIG. 2, a Findings Editor 214 preferably displays the nodes of the decision tree in the form of multiple layers similar to that illustrated in FIGS. 3a and 3b, allowing the user to select and unselect nodes. Selected nodes represent the findings, observations, or selections made by a user regarding a subject. In particular, each node represents one piece of data, such as the type of pathology observed or the size of a lesion. Preferably, if a node is selected by a user, then the Is Selected property of Table 1 is set to true. Preferably, the Is Selected property of Table 1 may also be set to true implicitly by inclusion of a related node. For example, the Co-existence Nodes property of Table 2 provides a user the ability when creating a Report Template 212 to specify other nodes that are automatically included upon the selection of the specified node.

[0047] Preferably, the Findings Editor 214 provides the user an option of entering data regarding the selected nodes that represent specific findings or observations regarding the subject, storing the value entered by the user in the Value property of Table 2. As one skilled in the art will appreciate, the Selected and the Value properties are modified by the user. The remaining properties indicate how the nodes behave and interact. As discussed above, however, other values may be added dependent upon the type of application and the specific needs of the users.

[0048] Alternatively, the user may copy a previously completed structured report and modify that structured report to accurately reflect the specific findings or observations made with respect to the current study. In this manner, the user can further reduce the amount of time required to complete the report.

[0049] The Findings Editor 214 is described in greater detail below with reference to FIG. 7.

[0050] Based upon the nodes selected by the user, the Findings Editor 214 creates a Structured Report 216, which preferably is represented as a hierarchical model of decision tree nodes that have been selected by the application user and implemented as an XML document. Other implementations or embodiments, such as marked-up text in an XML document, objects in an object-oriented database, objects serialized into a binary stream and saved in a file or database, rows and columns of data in a relational database, or the like, do not depart from the scope of the present invention. Alternatively, the Structured Report 216 may be stored in one or more multiple formats, such as an HTML document, a text file, a PDF file, or the like.

[0051] Generally, only selected nodes will appear in a Structured Report 216. However, in some instances it will be desirable to affirmatively state that a condition or observation was not present, in which case unselected nodes may appear in a Structured Report 216.

[0052] Preferably, a Report Formatter 218 (FIG. 2) utilizes the Structured Report 216, which contains the specific findings or observations made by a user, to create an Expanded Report 220. In the preferred embodiment, the Expanded Report 220 is dynamically created and displayed to the user as the user selects and deselects nodes in the decision tree. The dynamic creation and display of the Expanded Report 220 is discussed in greater detail below with reference to FIG. 8.

[0053] The Expanded Report 220 preferably presents the findings or observations made by a user in a concise, easy-to-read format in accordance with the structural position of the node with the Structured Report 216 and grammatical rules contained in the property values of the nodes and/or container set rules. A container set is a rule that specifies the sequencing of the word values associated with the nodes in the Expanded Report 220. Preferably, the container sets are implemented as either code (such as C statements, XSL statements, or the like) or data (such as an XML document or the like).

[0054] Alternatively, the Report Formatter 218 creates the Expanded Report 220 directly from the nodes selected by the user without first creating the Structured Report 218. The Structured Report 216, however, is preferred because it allows the findings of the user to be efficiently stored for later use, such as searching and reporting.

[0055] Preferably, the Report Formatter 218 is implemented as an XSLT document, a computer program written in a language such as C, C, Java, or the like, object oriented design and program, store procedures of a relational database, or the like.

[0056] In the preferred embodiment, the Expanded Report 220 comprises as an XHTML document in a chosen language. One of ordinary skill in the art will appreciate that the Value property may be written in any language such as English, Spanish, Russian, or the like, which results in an Expanded Report in that chosen language. Additionally, the ordering of nodes as specified by the Object Placement property (Table 4, Row 8), container sets, and Report Formatter rules may be altered to accommodate grammatical differences in the language of choice. The Report Formatter rules refer to the sequencing and grammatical rules that define the sequencing of the word values associated with the nodes, grammar (conjunction plurality, punctuation, and the like), sentence structure (capitalization, commas, periods, and the like), paragraph structures, and the like.

[0057] Preferably, the Expanded Report 220 comprises an HTML document, so that it can be rendered via a web browser. However, other embodiments, such as plain text, rich text format (RTF), portable data format (PDF), any other marked-up format that contains text with formatting tags, or the like, do not depart from the scope of the present invention.

[0058] The processing preferably performed by the Report Formatter 218 is discussed in greater detail below with reference to FIGS. 4-6.

[0059] FIG. 4 is a flow chart that represents the steps performed by the Report Formatter to combine the nodes of a decision tree into an Expanded Report 220 (FIG. 2), in accordance with an embodiment of the present invention. Processing begins in step 410, wherein a determination is made whether or not a container set rule applies to the node. Preferably, this determination is made by pattern matching, i.e., determining whether the node that is currently being evaluated matches a node to which one or more container sets apply. If, in step 410, a determination is made that the container set rule does not apply, then processing proceeds to step 412, wherein the parent node and the children nodes, if any, are sequenced in accordance with the property values associated with the respective nodes. The process associated with step 412 is discussed in greater detail below with reference to FIG. 5.

[0060] If, in step 410, a determination is made that a container set rule does apply, then processing proceeds to step 414, wherein the children nodes are sorted in accordance with the container set. The process associated with step 414 is discussed in greater detail below with reference to FIG. 6.

[0061] FIG. 5 is a data flow diagram that represents the steps performed by the Report Formatter 218 to sequence the parent and children nodes according to each node's properties as specified in step 412 (FIG. 4), in accordance with an embodiment of the present invention. Specifically, if a determination is made in step 410 (FIG. 4) that the sequence of the parent and child nodes is performed in accordance with the property values of the node, then processing proceeds to steps 510-530 (FIG. 5), after which processing returns to FIG. 4. Preferably, the nodes are processed in accordance with an in-order (i.e., left-parent-right) binary tree traversal method.

[0062] Preferably, when the node's properties are used to establish the ordering of node values, the ordering is determined by the value of the Children's Object Placement property contained in Table 4. As described in Table 4, the Children's Object Placement property includes a descriptor that specifies how the node's Value property value is to be treated. Preferably, the Children's Object Placement property includes alternatives such as prepend to parent, replace parent, append to parent, or the like. Optionally, a default value, such as append to parent, may be assigned.

[0063] Additionally, the output generated from the node may be further formatted dependent upon the values assigned to other fields, such as the Is Paragraph and Is Sentence properties of Table 4 and the Delimiter property of Table 2. Preferably, if the Is Paragraph property is set to true, a new paragraph is generated in the Expanded Report 220 prior to outputting the value of the node. If the Is Paragraph property is set to false, a value of the node is a phrase that modifies the parent node. If the Is Sentence property is set to true, a new sentence is generated in the Expanded Report prior to outputting the value of the node. Other formatting commands may also be used, some of which are specified in Table 4, Output Behavior Properties.

[0064] Accordingly, the processing begins in step 510, wherein a determination is made whether or not the node begins a paragraph. Preferably, the Is Paragraph property of Table 4 is a Boolean data type that has a value of either true or false. A value of true indicates that a new paragraph is to be generated in the Expanded Report prior to outputting the value of the node. Accordingly, if in step 510 it is determined that the Is Paragraph property contains a value of true, then processing proceeds to step 512, wherein the Report Formatter adds a paragraph tag, such as <P>, to the output stream prior to outputting the value of the node. Preferably, the default value for the Is Paragraph property is false.

[0065] After step 512 or step 510, if in step 510 it is determined that the Is Paragraph property contains a value of false, processing proceeds to step 514, wherein the Report Formatter adds the value of child nodes whose value of the Object Placement property indicates that the value of the child node is to be prepended to the value of the parent node, using the value of the Delimiter property as a separator. As described in the sixth row of Table 2, the Delimiter property value specifies the way in which the values of the child nodes are to be combined. The value of the Delimiter property preferably is an enumeration data type that has a value of either Comma, And or Space. This delimiter is inserted between the values of the children. Comma indicates that the values of the child nodes are separated by a comma; And indicates that the conjunction and is inserted before the last of a series of child values; and Space indicates that the values of the child nodes are separated by a space character. Preferably, the default value for the Delimiter property is Space.

[0066] Thereafter, processing proceeds to step 516, wherein a determination is made whether or not the value of the child node should replace the value of the parent node. If a determination is made that the value of the child node should not replace the value of the parent node, processing proceeds to step 518, wherein a determination is made whether or not the sibling nodes (nodes at the same level) of the node triggers a number agreement. Preferably, the node properties contains a Triggers Number Agreement property of an enumeration data type having values of None, Singular or Plural. If the value of the Triggers Number Agreement property of sibling nodes is None or Singular, the singular form of the node is utilized. If, however, the value of the Triggers Number Agreement property is Plural, then the plural form, such as mass/masses, tibia/tibiae, fracture/fractures, and the like, are utilized. Preferably, the plural form of a node is stored in the Plural Object Value property, and the Value property stores the value of the singular form. Preferably, the default value of the Triggers Number Agreement property is None.

[0067] Accordingly, if in step 518 it is determined that the sibling nodes of the node triggers a plural value, then processing proceeds to step 520, wherein the value contained in the Plural Value property of the node is outputted. If it is determined, in step 518, that the sibling nodes of the node do not trigger a plural value, then processing proceeds to step 522, wherein the value contained in the Value property of the node is outputted, i.e., the singular value of the node is outputted.

[0068] After step 520, step 522, or step 516 (if a determination is made in step 516 that the child node replaces the parent node) processing proceeds to step 524, wherein the Report Formatter 218 ignores the value of the parent node and adds child values that append to the parent node to the buffer, using the value of the Delimiter property as a separator.

[0069] Processing then proceeds to step 526, wherein a determination is made whether or not the node completes a sentence. As shown in the fourth row of Table 4, the value of the Is Sentence property is a Boolean data type that has a value of either true or false. When the Is Sentence property value is true, the Report Formatter 218 treats the value of this node as a full sentence in the report. Accordingly, if a determination is made that the node is a sentence, then processing proceeds to step 528, wherein the Report Formatter 218 modifies the buffer to add punctuation and capitalization. If, however, in step 528, a determination is made that the node is not a sentence, i.e., the value of the Is Sentence property is false, the Report Formatter 218 treats the value as a phrase that modifies a parent node. The default for the value of the Is Sentence property is false. When all of the values of the parent and children have been accumulated in the buffer, the buffer contents are added to the output stream as shown in step 530.

[0070] Both step 514 and step 524 represent the Report Formatter 218 adding the values of the child nodes to the buffer. Preferably, when child nodes are added to the buffer, the value of the Children's Order Fixed property determines the order in which the children are listed. The Children's Order Fixed property is a Boolean data type that has a value of either true or false. When the value is true, the nodes are formatted in the order in which they appear in the decision tree. Otherwise, when the value is false, the nodes are aggregated in the order in which the user selected them. The default value is true. Thereafter, processing returns to FIG. 4.

[0071] FIG. 6a shows an example of a container set rule. Rule 1101 for Primary Modifiers (labeled Adjectives in FIG. 6) shows the sequence of nodes that are output for a description of a pathology. Trend (Pre) is the first category, Number is the second, and Morphology is the last category. Nodes may be presented in a decision tree in any order (for example, Pathology, Morphology, Pre-Anatomic Site, Pre-Dimension, Number, Chronicity, Pre-Trend, Interactions, Associations), but are preferably outputted in a report in the order specified by the Container Set rule. For illustrative purposes only, FIG. 6b illustrates an example of output that may be obtained utilizing the container set 1101. The value of the pre-dimension node (i.e., extensive) is positioned prior to the value associated with the pre-anatomic site node (i.e., RUL). The value associated with the pathology node (i.e., partial consolidation) is positioned last. Thus, the sentence becomes, Extensive RUL partial consolidation.

[0072] Similarly, Rule 1102 for Secondary Modifiers (labeled Adjectives in FIG. 6) shows the sequence of nodes that are also output for a description of a pathology. For illustrative purposes only, FIG. 6c illustrates an example of output that may be obtained utilizing the container set 1102. The value of the pre-dimension node (i.e., extensive) is positioned prior to the value associated with the pathology node (i.e., partial consolidation). The value associated with the post-anatomic site node (i.e., RUL) is positioned last. Thus, the sentence becomes, Extensive partial consolidation involving the RUL.

[0073] Preferably, the embodiment of a container set depends on the embodiment chosen for the Report Formatter. In one embodiment, where the Report Formatter is specified using the Extensible Stylesheet Language (XSL), a container set may be represented as an XSL template. Other embodiments may include, but are not limited to, conditional if-then-else statements in a procedural programming language, rule objects with conditions and actions in an object-oriented language, declarative rules in an expert system or a business rules engine, or the like.

[0074] FIG. 7 is a flow chart that represents the steps performed by the Findings Editor 214 (shown in FIG. 4) to convert a decision tree into a Structured Report 216, in accordance with an embodiment of the present invention. Preferably, the steps involve performing a pre-order tree traversal, wherein the properties of selected parent nodes are added to the Structured Report 216, and then the properties of each of the selected child nodes of the parent node are added to the Structured Report 216. Preferably, the hierarchical structure of a decision tree is represented as an XML document, where the tree root node is represented as the root XML element. Each node in the tree is transformed as an element that contains elements which represent the node's properties.

[0075] Accordingly, processing begins in step 710, wherein the Findings Editor 214 outputs opening tags for a parent node. Tags define how a Web page is to be displayed and enable user interactions. For example, the opening node tag appears as <NODE> wherein the tag or instruction is surrounded by a less-than and a greater-than sign. Tags generally appear in pairsthe opening tag and the closing tag. The closing tags are similar to the opening tags except for a forward slash in the closing tag. For example, a node will be surrounded by the NODE tags: <NODE>child elements</NODE>. The closing </NODE> tag is discussed below in reference to step 722.

[0076] Processing proceeds to step 712, wherein the Findings Editor 214 outputs the desired properties contained in the Tables 1-5 as tags and values. Preferably, all of the properties contained in Tables 1-5 are added to the Structured Report 216. However, one of ordinary skill in the art will realize that some embodiments of the present invention may not require retention of all of the properties contained in Tables 1-5, such as, for example, Table 3, Input Behavior Properties. In step 714, the Findings Editor 214 outputs opening list tags for the list of child nodes.

[0077] In step 716, a determination is made whether or not the parent node contains more child nodes. If a determination is made that the parent node contains more child nodes, processing proceeds to step 718, wherein the child node is processed. Preferably, the child node is processed by a recursive call to the node processing routine of the Findings Editor 214. Specifically, the child node is preferably processed by beginning processing at step 710 for the specific child node, which may be a parent node to other nodes.

[0078] If, however, in step 716, a determination is made that there are no more child nodes to process, processing proceeds to step 720, wherein the closing list tags associated with the parent node are outputted.

[0079] When all of a child nodes have been processed, the parent node outputs its own closing tags, as shown by block 722. It is noted that the original parent node is the node that begins the process at each level in the hierarchical tree. Thus, the process is recursive and each node is handled relative to its position in the hierarchy, and a node may be both a parent and a child, because these terms refer to relative relationships with other nodes, not absolute positions in the tree.

[0080] Performing the pre-order tree traversal discussed above results in a well-formed XML document, in which the hierarchical structure of the nodes is represented by the embedding of elements within each other and the content of the decision tree is represented as elements within each node.

[0081] XML provides a flexible format that allows for certain modifications of the hierarchical structure, without having to change the processors of the data. For example, it is possible to add a new property to a node without having to modify the parser that reads decision trees. Another advantage of XML is that it is widely used for transferring data between organizations. Therefore, integration with other systems is easier due to the use of XML for decision trees. Finally, there are many third-party tools available for processing XML, which makes it easier to develop applications to process decision trees.

[0082] For illustrative purposes only, FIG. 8 illustrates a Report Template and a dynamically created Expanded Report that documents the structured analysis of a particular radiological study in accordance with an embodiment of the present invention. As discussed above, the Expanded Report is preferably created on the right-hand side as the user selects and de-selects items in the Report Template on the left-hand side. The template 800 contains an initial layer 802 of choices that comprises the following: Organ morphology (Om) 804, Organ physiology (Op), Specific pathology (Sp), General pathology (Gp) 806, and Negatives (N) 808.

[0083] The Organ morphology choice 804 defines pathological entities that describe pathological architectural alterations of organs or organ systems, as well as pathological changes in the tissue characteristics of organs or organ systems.

[0084] The Organ physiology choice defines pathological entities that describe physiological alterations of organs or organ systems.

[0085] The Specific pathology choice defines a list of specific diseases of specific organs or specific organ systems or well-known syndromes. Upon demand by the user, the reporting system may provide specific information regarding a specific disease entity with respect to a disease synopsis or specific radiological features.

[0086] The General pathology choice 806 defines and provides analysis of unknown disease processes of organs or organ systems and the reporting system may provide differential diagnoses of general disease entities upon demand by the user.

[0087] The Negatives choice 808 defines an exhaustive list of pertinent negatives.

[0088] The comments choice (not shown) allows the user to enter comments by using an input device (e.g., voice input, mouse, keyboard, etc.) that is connected to the interface devices 110.

[0089] Each major division is composed of a set of choices consisting of minor divisions. The minor divisions are contained within the second layer 810, and the minor divisions, include the following: *Number(Nb) 812, Pathology(P) 814, *Morphology(M) 816, *Chronicity(C) 818, *Dimensions(D) 820, Anatomic site(AS) 822, Tissue characteristics(Tc) (not shown), Interactions(I) 824, Associations(A) 826, *Trend(T) 828 and RADITFS (not shown).

[0090] The starred choices are the primary modifiers and are the same for all templates in all parts of the human body for all modalities. The unstarred choices are the secondary modifiers and are specific to a particular anatomic site within the human body and are particular to a specific modality.

[0091] The Number choice 812 is the first function in the output of the Report Formatter 218, and all other primary modifier functions follow this function. The Number choice 812 consists of a drill-down list comprised of the following: Possible, Single, One, Two, Three, Multiple, Numerous or other quantitative descriptors. The Number node 812 is a category, a way of organizing specific choices. Those choices (possible, single, two, etc.) represent possible findings. For example, if the pathology is a mass, then a radiologist may specify the number of masses as possible mass, single mass, two masses, etc. Modifiers of a multiple selection may be selected and attached consisting of identification of largest of two, 2nd largest of multiple, 3rd largest of three, etc. These selections are more specific descriptions of the term multiple. For example, a radiologist might simply specify multiple masses or something more specific, such as 2nd largest of multiple masses.

[0092] The Pathology choice is a pathological process involving organ morphology 804 or organ physiology or specific pathology or general pathology and is the primary navigational tool. When a user selects a disease process under Pathology 814 in any of the above categories, the present invention provides the user with options that enable the creation of the output and provides access to other information. Further, when the Pathology choice 814 is highlighted under specific pathology, the user may select a Synopsis choice (not shown). The Synopsis choice displays a medical summary of the disease entity, including the Etiology, Incidence, Pathophysiology, Clinical features, Imaging characteristics and References of a specific disease entity.

[0093] The order in which different Pathology entities appear in the Impression section 832 is determined by the user's assignment of a relative sequence number (not shown) to a Pathology entity. Once the order has been specified by the user, a summary statement of the Pathology entity and associated modifiers will appear in the Impression section 832 in the relative order specified by the user

[0094] The output in the Findings section 830 consists of a certain combination of primary modifiers and a Pathology entity 814. For example, in FIG. 8, the Findings section 830 states: Pathology: New 5 cm RUL cavitating mass.

[0095] The output in the Impression section 832 consists of a variable subset of the primary modifiers. The subset may consist of Number 812, Pathology 814, Anatomic site, Morphology 816, Chronicity 818, Dimensions 820 and Trend 828 in a proscribed output order. The subset must include a Pathology entity 814.

[0096] The Differential Diagnosis section (not shown) contains a drill-down differential diagnosis list that may be selected for any Pathology entity. Up to three selections may be made from the list, and the output is ordered according to the user's selection order. The output of the selections is reformatted in a specific proscribed manner and is attached to the Impression section 832. When a Differential Diagnosis entity is highlighted the user may select Synopsis (not shown) and display a medical summary of the Differential Diagnosis entity.

[0097] A standard Recommendation list (not shown) may be selected for any Pathology entity. The output of the Recommendation is reformatted in a specific proscribed manner and is attached to the specific Impression 832 following the Differential Diagnosis. When the Differential Diagnosis option is not selected, then the Recommendation is attached directly to the Impression 832.

[0098] The Morphology function 816 is a primary modifier and immediately precedes the Pathology entity in the Findings 830 and Impression 832 output analytical algorithm. The Morphology function 816 defines the type of lesion, the shape of the lesion, the margins of the lesion and the internal structure or matrix of the lesion.

[0099] The Anatomic (pre) site is a primary modifier and immediately precedes the Morphology function 816 in the Findings 830 and Impression 832 output analytical algorithm.

[0100] The Chronicity function 818 is a primary modifier and immediately precedes the pre-Anatomic site function in the Findings 830 and Impression 832 output analytical algorithm. The Chronicity function 818 defines the age of the lesion and defines age as hyper-acute, acute, sub-acute, chronic, old and indeterminate aged Pathology 814.

[0101] The Dimension function 820 is a primary modifier and immediately precedes the Chronicity function 118 in the Findings 830 and Impression 832 output analytical algorithm. The Dimension function 820 defines the size of the lesion qualitatively or quantitatively.

[0102] The detailed post-Anatomic site 822 function is a secondary modifier and follows the Pathology 814 entity and its primary modifiers and is contained in the same sentence.

[0103] The Tissue characteristics (not shown), Interactions 824 and Associations 826 functions are special functions with special attachment tools, which may secondarily modify a Pathology 814 entity in a subsequent section. The Tissue characteristics function displays general tissue characteristics utilized in the description of a general or specific Pathology 814 entity.

[0104] When a Pathology entity under Specific pathology (not shown) is highlighted the user may select Tissue characteristics and display the classic Tissue characteristics of that specific disease entity.

[0105] The Interaction function 824 describes the direct effects of the Pathology 814 entity upon immediately surrounding anatomical structures. The Interaction function 824 consist of a drill-down list of Interaction descriptors coupled to a recursive anatomic site list which displays in a proscribed manner in the output analytical algorithm in the Findings sections 830. The Interaction function 824 secondarily modifies the Pathology entity 814 in a subsequent section.

[0106] The Associations function 826 displays general radiological features of a General pathology pathological entity and specific detailed radiological features of a Specific pathology pathological entity. The specific pathology entity is coupled with the specific Associations for that specific pathology entity and selected features may be selected by the user and displayed in a proscribed manner in the output analytical algorithm in the Findings section 830.

[0107] The Trend function 828 is a secondary modifier which describes the course of disease over time and stands alone. The Trend function may be selected for each and every Pathology entity.

[0108] RADITFS function (not shown) is selected in the Impression and displayed in the Impression 832 only. RADITFS or RADIOLOGIC ANALYSIS DETAILED IN THE findings SECTION is attached to the Impression 832 following differential diagnosis and Recommendations.

[0109] This concludes the description of the preferred embodiment of the invention. The foregoing description of the embodiments of the invention have been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed. Many modifications and variations are possible in light of the above teaching. It is intended that the scope of the invention be limited not by this detailed description, but rather by the claims appended hereto.