PubMed Health⌕ Search

Biomedical subjects

D Partridge

Publications and source records attributed to D Partridge.

8 recordsLinked to original sources
PubMed1999

Artificial neural networks: a potential role in osteoporosis.

Artificial neural networks are computer software systems that recognize patterns in complex data sets. A recent development in neural computing, multiversion systems (MVS), has led to enhanced analytical power, and this was harnessed to demonstrate the value of risk factors in predicting the result of osteoporosis investigations by quantitative ultrasound. 274 women were screened in an open-access osteoporosis service. A conventional risk factor questionnaire was completed for each patient by the osteoporosis specialist nurse. An MVS was trained on 180 randomly selected data sets and tested on the remaining 94. The results were compared with those from logistic regression analysis in predictive power, both from the selected 20-item questionnaire and for a limited 5-item questionnaire comprising age, height, height loss, weight and years since the menopause. The MVS approach predicted the T-score categorization of the patients from the 20-item questionnaire with 83.0% accuracy, whereas logistic regression yielded an accuracy of only 72.8% (P = 0.04). From the 5-item database the MVS yielded a best prediction accuracy of 73.1%, whereas the logistic regression prediction accuracy was 60% (P = 0.04). These results suggest that 20 risk factors can be used by an MVS to predict the outcome of osteoporosis investigations with a power that outperforms conventional statistical methods. Use of this system may improve the selection of patients for osteoporosis investigations, since even with only 5 risk factors the system performs nearly as well as that based on the full 20 factors.

Aged↗
PubMed1996

Engineering multiversion neural-net systems.

In this paper we address the problem of constructing reliable neural-net implementations, given the assumption that any particular implementation will not be totally correct. The approach taken in this paper is to organize the inevitable errors so as to minimize their impact in the context of a multiversion system, i.e., the system functionality is reproduced in multiple versions, which together will constitute the neural-net system. The unique characteristics of neural computing are exploited in order to engineer reliable systems in the form of diverse, multiversion systems that are used together with a "decision strategy" (such as majority vote). Theoretical notions of "methodological diversity" contributing to the improvement of system performance are implemented and tested. An important aspect of the engineering of an optimal system is to overproduce the components and then choose an optimal subset. Three general techniques for choosing final system components are implemented and evaluated. Several different approaches to the effective engineering of complex multiversion systems designs are realized and evaluated to determine overall reliability as well as reliability of the overall system in comparison to the lesser reliability of component substructures.

Algorithms↗
PubMed1987

The relationship of lactoferrin to the anemia of renal cell carcinoma.

This study was designed to examine whether lactoferrin, a glycoprotein contained in neutrophils which binds free iron, mediates the anemia associated with renal cell carcinoma. Preoperative hematocrit, urinalysis, serum iron, total iron binding capacity, and ferritin levels were obtained in 24 patients with hypernephroma. At the time of radical nephrectomy, a tumor specimen was obtained from all 24 patients and corresponding normal renal tissue was obtained from eight patients. Fifteen patients had low serum iron, whereas nine patients had normal serum iron. All tissue samples were snap frozen at the time of surgery and were subsequently sectioned into 3-microns slices using the cryostat. Then all the sectioned specimens were stained with FITC (fluorescein isothiocyanate) and peroxidase conjugated rabbit derived anti-human lactoferrin. Ten of the 15 patients with low serum iron had positive anti-lactoferrin staining in both the FITC and peroxidase systems. None of the tumors from patients with normal serum iron and none of the normal renal parenchyma exhibited positive anti-lactoferrin uptake. Stains for iron in the bone marrow of two patients with low serum iron showed increased iron stores. These studies suggest that lactoferrin mediates the anemia often seen in association with renal cell carcinoma.

Anemia↗
PubMed1984

Computer programs as theories in biology.

Complex theories in biology may be developed, refined, and tested by the use of computer programmed simulations. The computer is recognized as a powerful tool for theory development; it is, in fact, the only means of thoroughly testing and examining a large and intricate theory. A program as a text is a statement of a theory and when run on the computer it is model of that theory. As the program's behavior is then the major argument for the credibility of a large and complex theory, the program itself is the only irrefutable statement of the theory. Bur programs written in the currently available programming languages tend to be incomprehensible. We argue that the program should be the definitive statement of the theory. In addition, the program plus a series of abstractions is a vehicle for effective communication of complex theories in biology. Several techniques of computer science are borrowed, for the purpose of developing a methodology for abstraction and a language for representing abstractions. The arguments are fully illustrated with a recently published biological theory.

Communication↗
PubMed1983

A neural theory of cognitive development.

A physiologically based model of the neocortex has been developed in an attempt to elucidate possible structural and functional mechanisms of the mammalian cortex and account for a wide range of low level cognitive behavior. The model has been constrained by diverse empirical data. At the level of structural details, neuroanatomical and neurophysiological data have been considered and at the level of gross behavior, psychological data has been used. From the theory that groups of reverberating neurons provide a short term memory mechanism and that primary drive reduction triggers consolidation of a memory, a mechanism for selective learning has been developed. Fundamental to the model is the postulate of a novelty drive mechanism that functions in a manner analogous to the more widely accepted primary drives (e.g. hunger and fear). This paper examines the novelty drive mechanism and demonstrates its utility in accounting for a wide range of habituation behaviors. The success of the model is evaluated by comparing its behavior to appropriate empirical data. Finally, it is argued that a computer program is both a theory and a model, and that important advantages accrue from such a viewpoint.

Animals↗
PubMed1975

Urine vapor pattern for olivopontocerebellar degeneration.

Autosomal dominant olivopontocerebellar degeneration was evaluated in a family numbering 414 living members. Urine samples were obtained from 5 affected and 5 unaffected not at-risk family members after 4 days on a synthetic diet. 159 urine vapor constituents were measured by gas-liquid and ion-exchange chromatography and a non-correlation index was calculated. 2 distinct populations of urine vapor patterns were identified which conformed to the normal and disease affected groups, thus a diagnostically useful pattern for this disease in this family has been demonstrated which will be helpful in genetic counselling and potential elimination of the disease.

Adult↗