Data processing: artificial intelligence – Knowledge processing system – Knowledge representation and reasoning technique
Reexamination Certificate
1999-08-20
2001-06-12
Powell, Mark R. (Department: 2122)
Data processing: artificial intelligence
Knowledge processing system
Knowledge representation and reasoning technique
C706S045000, C706S047000, C600S300000
Reexamination Certificate
active
06247004
ABSTRACT:
BACKGROUND AND SUMMARY OF THE INVENTION
In many areas of human endeavor (arts), including medicine (human and veterinary, and particularly specialties therewithin), the biological sciences, geology, automobile repair, etc., it is necessary to diagnose a condition or event from the universe of known conditions or events relating to that art. The use of a computer to assist in that diagnosis has long been considered desirable, and in many art areas computers are used in one form or another to facilitate diagnosis. For most arts, however, truly effective computer utilization to effect diagnosis is still not a reality. In some situations this is due to an inability to effectively use the search and comparison functions of a computer because an appropriate data base is difficult to construct and/or utilize.
According to the most basic aspects of the present invention, a new tool is provided for constructing and utilizing data bases that may be searched and analyzed by a computer in an effective manner, coined “Diagnostic Mapping Language”, or DML. The basic concept behind DML is that all events or conditions have characteristic markers that make them unique and therefore identifiable. A novel grammar is utilized so as to build a data base of individual master maps, each master map corresponding to a condition or event to be diagnosed. A diagnosis is secured by a methodical description of its markers, and by utilization of a computer to compare the markers of a condition or event to be diagnosed with the individual master maps to determine the degree of resemblance between the description input and the conditions or events represented by the master maps.
The DML grammar is based upon three classes of words, namely:
Main: word that describes the most important feature of a sentence. In the case of a medical text it may be an organ, a body part, a secretion, a chemical, etc.
Descriptor: a word that further describes the Main word, by adding a qualification.
Complement: a word that further describes the Descriptor by adding a qualification.
In the description of the invention which follows, diagnosis of a human medical condition using a data base containing a plurality of master maps of medical conditions within a particular specialty (such as nephrology) is provided. However, it should be understood that the principles utilized in the examples are applicable to many other arts where a diagnosis is desired, including the veterinary medicine, biological sciences, geology, and automotive repair arts given as examples above. In the examples that follow some features of human medicine interface with computer technology are also provided that are not necessarily based upon the effective utilization of DML to construct and access data bases, but rather have advantages that are applicable to the effective practice of medicine using a computer regardless of the manner in which the data base is constructed or accessed.
According to the broadest aspects of the present invention, a method of creating a computer searchable data base for use with a computer having a display screen, an input device (e. g. keyboard, mouse, voice responsive signal generator, or any other conventional or known device for putting data or commands into a computer), and processor, for ultimately diagnosing a condition or event from a large plurality of possible conditions or events that may exist in a particular art. The method comprises the steps of: (a) Determining what key words (including, possibly, phrases) describe the characteristic features of each of the possible conditions or events of the large plurality of possible conditions. (b) Creating in computer searchable form a master map for each of the large plurality of possible conditions or events using the key words (including, possibly, phrases) from step (a) that accurately describe each of the conditions or events, to provide a data base containing all the master maps. And (c) providing search access to the master maps data base in the computer so that by a user manipulating the input device to provide a narrative containing some of the key words from step (a), the computer processor will compare the key words input to the key words associated with each master map to determine what conditions or events represented by master maps having the highest degree of similarity with the narrative input, and display those conditions or events on the display screen.
Step (a) may be practiced by dividing the key words into three different classifications of words, main words that describe the most important feature of a sentence that accurately describes an aspect of a condition or event, descriptor words that further describe a main word by adding a qualification, and complements that further describe a descriptor by adding a qualification. Step (b) may be practiced by providing a plurality of distinctly searchable elements for each condition, each element including at least a main word, and where they accurately exist, a descriptor and complement associated with that main word. Step (b) may be further practiced by assigning to each of the elements one of at least two possible logics, a first logic that requires a main and at least a descriptor to also be present before the element is recognized by the processor when searching the data base, and a second logic that requires only the main to be present, but recognizes a descriptor and complement if present.
Steps (a) and (b) may also be practiced by assigning key words to laboratory test results related to each condition or event, including the name of a test, the normal numerical value of the test results if the condition or event doesn't exist, and a flag indicating that the value of inputted data is above normal, below normal, or text including at least one of normal, abnormal, positive, negative, absent, or present. Steps (a) and (b) may also further be practiced by assigning either a “yes” or “no” to individual imaging possibilities related to a particular condition.
The art may be a human medical specialty, where the conditions or events are possible diseases or medical conditions of that medical specialty, and where the main words describe the anatomy or body fluid or tissue involved with the disease or medical condition to be addressed, the descriptor describes the deviation from normal of the main word, and the complement adds specificity to the main or descriptor words. In that case, step (c) may be practiced by placing the master maps data base on a CD ROM along with biopsy images, medical imaging images, or both biopsy images and medical imaging images, associated with a plurality of the master maps, the images capable of display on the display screen by the user manipulating the input device.
According to another aspect of the invention, a method of diagnosing a condition or event from a large plurality of possible conditions or events that may exist in a particular art, each condition or event being provided in a master map which describes that condition or event using a plurality of art specific key words, is provided. The method is practiced using a computer having a display screen, input device, and processor, and comprises the following steps: (a) Providing search access to the master maps data base in the computer so that by a user manipulating the input device to provide a narrative containing some of the key words from the master maps, the computer processor will compare the key words input to the key words associated with each master map. (b) Using the input device, inputting a narrative description that can facilitate diagnosis, the narrative description using some of the key words. (c) Using the input device, activating the computer processor to recognize the key words from the input narrative description, compare the key words to the key words of each of the master maps, to determine what conditions or events represented by master maps have the highest degree of similarity with the narrative input. And (d) displaying on the display screen at least the two conditions represented by master maps
Khatri Anh
Nixon & Vanderhye P.C.
Powell Mark R.
LandOfFree
Universal computer assisted diagnosis does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Universal computer assisted diagnosis, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Universal computer assisted diagnosis will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2467838