Data processing: artificial intelligence – Neural network – Learning task
Reexamination Certificate
1999-11-29
2003-12-02
Starks, Jr., Willbert L. (Department: 2121)
Data processing: artificial intelligence
Neural network
Learning task
C600S300000, C600S544000
Reexamination Certificate
active
06658396
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally concerns the construction, training and use of neural networks for the optimization of the administration of drugs (and, for the invention, drug equivalents such as food and exercise) in respect of patient characteristics.
The present invention particularly concerns the construction, training and use of neural networks to better recognize any of (1) optimal patient dosage of a single drug, (2) optimal patient dosage of one drug in respect of the patient's concurrent usage of another drug, (3a) optimal patient drug dosage in respect of patient characteristics, (3b) sensitivity of patient recommended drug dosage to patient characteristics, (4a) expected outcome versus patient drug dosage, (4b) sensitivity of expected outcome to drug dosage, (5) expected outcome(s) from drug dosage(s) other than projected optimal dosage, from which expected outcome(s) costs both human and economic may be separately predicted.
2. Description of the Prior Art
2.1 Drug Dosage Estimation by Drug Developers and Physician Practitioners
Many ailments exist in society for which no absolute cure exists. These aliments include, to name a few, certain types of cancers, certain types of immune deficiency diseases and certain types of mental disorders. Although society has not found an absolute cure for these and many other types of disease, the use of drugs has reduced the negative effect of these disorders.
Generally the developers of drugs have two goals. First, they try to alter the drug user's biochemistry to correct the physiological nature of the illness. Second, they try to reduce the drug's negative side effects on the user. To accomplish these goals, drug developers utilize time consuming and scientifically advanced methods. These expensive efforts yield an extremely high cost for many drugs.
Unfortunately, when these costly drugs are distributed they are usually accompanied by only a crude system for assisting a doctor in determining an appropriate drug dosage for a patient. For instance, the annually printed
Physician's Desk Reference
summarizes experimentally determined reasonable drug dosage ranges found in the research literature. These ranges are general. The same dosage range is given for all patients.
Other publications exist which provide general methods to assist a doctor in determining an appropriate dosage. These references and manuals are not, however, directed towards providing a precise dosage range to match a specific patient. Rather, they provide a broad range of dosages based on an averaging of characteristics over an entire population of patients. The correlations between distinguishing patient characteristics and actual required dosages are never obtained, even in the original research.
Faced with the task of minimizing side effects and maximizing drug performance, doctors sometimes refine the dosage they prescribe for a given individual by trial and error. This method suffers from a variety of deleterious consequences. During the period that it takes for trial and error to find an optimal drug dosage for a given patient, the patient may suffer from unnecessarily high levels of side effects or low or totally ineffective levels of relief. Furthermore, the process wastes drugs, because it either prescribes a greater amount of drug than is needed or prescribes such a small amount of drug that it does not produce the desired effect. The trial and error method also unduly increases the amount of time that the patient and doctor must consult.
2.2 The Need for Drug Dosage Optimization
The past few decades have produced research identifying numerous factors that influence the clinical effects of medication. Age, gender, ethnicity, weight, diagnosis and diet have all been found to influence both the pharmacokinetics and pharmacodynamics of drugs. As a result, it is now acknowledged that women, minorities, and the elderly often require considerably lower doses of some medications than their male Caucasian counterparts. Furthermore, it is possible that patient variables have potentially varying strengths of influence for each case, and each drug. For example, weight may be of greater importance than age for a Caucasian male while the converse may be true for an African American female. See Lawson, W. B. (1996). The art and science of psychopharmacotherapy of African Americans.
Mount Sinai Journal of Medicine,
63, 301-305. See also Lin, K. M., Poland, R. E., Wan, Y., Smith, M. W., Strickland, T. L., & Mendoza, R. (1991). Pharmacokinetic and other related factors affecting psychotropic responses in Asians.
Psychopharmacology Bulletin,
27, 427-439. See also Mendoza, R., Smith, M. W., Poland, R., Lin, K., Strickland, T. (!991). Ethnic psychopharmacology: The Hispanic and Native American perspective.
Psychopharmacology Bulletin,
27, 449-461. See also Roberts, J., & Tumer, N. (1988). Pharmacodynamic basis for altered drug action in the elderly.
Clinical Geriatric Medicine,
4, 127-149. See also Rosenblat, R., & Tang, S. W. (1987). Do Oriental psychiatric patients receive different dosages of psychotropic medication when compared with Occidentals?
Canadian Journal of Psychiatry,
32, 270-274. See also Dawkins, K., & Potter, Z. (1991). Gender differences in pharmacokinetics and pharmacodynamics of psychotropics: Focus on women.
Psychopharmacology Bulletin,
27, 417-426.
The large number of potentially interacting variables to consider, in addition to the wide therapeutic windows of many drugs (including psychotropic drugs) have resulted in prescribing practices that rely mainly upon trial-and-error and the experience of the prescribing clinician.
The compensation process can be quite lengthy while drug consumers experiment with varying dosages. New methods are needed to reduce the time to compensation for patients (including psychiatric patients), thus alleviating their suffering more quickly as well as reducing the cost of hospitalization. The optimization of drug dosages would also help avoid unnecessarily high dosages, reducing the severity of the many side effects that typically accompany such medications and increasing the likelihood of long-term compliance with the prescribed regimen.
For decades, researchers have recognized the need for finding new methods of accounting for inter-individual differences in drug response. See, for example, Smith, M., & Lin, K. M. (1996); A biological, environmental, and cultural basis for ethnic differences in treatment; In P. M. Kato, & T. Mann (Eds.),
Handbook of Diversity Issues in Health Psychology
(pp. 389-406); New York: Plenum Press; and also Lenert, L., Sheiner, L., & Blaschke, T. (1989). Improving drug dosing in hospitalized patients: automated modeling of pharmacokinetics for individualization of drug dosage regimens;
Computational Methods in Programs Biomedical,
30, 169-176.
However, a practical solution to tailoring drug regimens has yet to be implemented on a widespread basis.
2.3 Existing Pharmacological Software
Pharmacological software currently in use attempts to provide guidelines for drug dosages, but most software programs merely access databases of information rather than compute drug dosages. At best, these databases rely upon existing research that groups subjects in a few gross categories (e.g., the elderly, or children), and they usually do not include information regarding such relevant characteristics as weight or ethnicity.
The few analytical software products that make use of computer algorithms base their recommendations primarily upon blood plasma concentrations of the drug of interest. See, for example, Tamayo, M., Fernandez de Gatta, M., Garcia, M., & Dominguez, G. (1992); Dosage optimization methods applied to imipramine and desipramine in enuresis treatment;
Journal of clinical pharmacy and therapeutics,
17, 55-59; and also Lacarelle B., Pisano P., Gauthier T., Villard P. H., Guder F., Catalin J., & Durand A. (1994); Abbott PKS system: a new version for applied pharmacokinetics including Bayes
Arouh Scott
Diamond Cornelius
Tang Sharon S.
Fuess & Davidenas
Starks, Jr. Willbert L.
LandOfFree
Neural network drug dosage estimation does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Neural network drug dosage estimation, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Neural network drug dosage estimation will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3125757