Search

Artificial Adaptive Systems include Artificial Neural Networks (ANNs or simply neural networks as they are commonly known). The philosophy of neural networks is to extract from data the underlying model that relates this data as an input/output (domain/range) pair. This is quite different from the way most mathematical modeling processes operate. Most mathematical modeling processes normally impose on the given data a model from which the input to output relationship is obtained. For example, a linear model that is a “best fit” in some sense, that relates the input to the output is such a model. What is imposed on the data by artificial neural networks is an a priori architecture rather than an a priori model. From the architecture, a model is extracted. It is clear, from any process that seeks to relate input to output (domain to range), requires a representation of the relationships among data. The advantage of imposing an architecture rather than a data model, is that it allows for the model to adapt. Fundamentally, a neural network is represented by its architecture. Thus, we look at the architecture first followed by a brief introduction of the two types of approaches for implementing the architecture—supervised and unsupervised neural networks. Recall that Auto-CM, which we discuss in Chap. 3, is an unsupervised ANN while K-CM, discussed in Chap. 6, is a supervised version of Auto-CM. However, in this chapter, we show that, in fact, supervised and unsupervised neural networks can be viewed within one framework in the case of the linear perceptron. The chapter ends with a brief look at some theoretical considerations.

We have looked at how to visualize the relationships among the elements of a dataset in Chap. 4. This chapter is devoted to the use of Auto-CM in the transformation of datasets for the purpose of extracting further relationships among data elements. The first transformation we call the FS-Transform, which looks beyond an all or nothing, binary relationship that is typical of most ANNs. The extraction of these perhaps more subtle relationships can be thought of as gradual relationships, zero denoting no relationship is present and one denoting a full/complete relationship that is absolutely present. It is thus, akin to a fuzzy set. The second transformation is one, which “morph” the delineation between records and variables within records that we call Hyper-Composition.

We look at how to use Auto-CM in the context of datasets that are changing in time. We modify our approach while keeping the original philosophy of Auto-CM.

This paper uses computer modeling to estimate the progressive decline in oxygenation that occurs in the human diabetic retina after years of slowly progressive ischemic insult. An established model combines diffusion, saturable consumption, and blood capillary sources to determine the oxygen distribution across the retina. Incorporating long-term degradation of blood supply from the retinal capillaries into the model yields insight into the effects of progressive ischemia associated with prolonged hyperglycemia, suggesting time-scales over which therapeutic mitigation could have beneficial effect. A new extension of the model for oxygen distribution introduces a feedback mechanism for vasodilation and its potential to prolong healthy retinal function. A model for retinal oxygenation estimates ischemia associated with diabetes.The model employs diffusion, saturable consumption, and a nonlinear transport.The model's results reinforces the need for tighter glycemic control.A new extension includes vasodilation as a feedback response to ischemia.

In radiation shielding design, one is usually faced with a set of conflicting goals that are navigated by an experienced physicist. If one has abundant space, the task is simplified because concrete is relatively inexpensive and will provide adequate shielding for high energy photons and neutrons, when applicable. However, if space is constrained (which is usually the case), the design becomes more difficult since one will likely have to employ combinations of steel, lead, and concrete, or other new materials—each with different properties and costs. Very experienced shielding designers can draw upon previous plans, but they do not know if their design is optimal in any sense. We have constructed a linear program that minimizes the cost of the shielding materials and minimizes the dose at the protection point or the shielding thickness subject to space constraints and to Federal or State regulations regarding the allowable exposure to individuals adjacent to the radiotherapy vault. In spite of what appears to be a simple model, the solution may require iterations of the optimization to arrive at the optimal solution.

Purpose: To compute and to verify experimentally Cherenkov radiation production inside the eye from direct and indirect high energy x‐ray irradiation of the orbit and that it exceeds detectability of Cherenkov radiation in scotopic vision. We show that the Cherenkov yield for direct and indirect irradiation exceeds the detection threshold of 6.4×107 photons/m2s. Methods: In photon and electron beamradiotherapy of intraorbital and periorbital tissues, patients commonly report having light sensations during treatment. The explanation usually offered is that ‘nerve stimulation’ is occurring. The radiotherapy literature reports the effect to be the result of ‘phosphenes’. Although phosphenes may play a role, we propose instead that patients are predominantly seeing Cherenkov radiation resulting from electrons inside the eye having kinetic energy exceeding the Cherenkov radiation threshold of 0.26 MeV. We consider calculations for direct irradiation of the eye from a portal image and indirect irradiation from treatment of peri‐orbital tissues and show that it exceeds threshold detection. Results: We calculate the Cherenkov yield using analytic methods and the measurements were in good agreement with our calculations. The Cherenkov radiation from a distilled water phantom, a square plastic phantom and an anthropomorphic plastic phantom were readily visible with a high quality CCD video camera. A digital camera captures the images from the console monitor. Threshold detection of Cherenkov radiation emanating from the water phantom through the video camera was determined and compared to a measurable source. The calculations for the water phantom match reasonably well with measurements and are roughly 1 lux and 3 lux respectively. Conclusion: We show images of Cherenkov radiation emanating from routine radiotherapytreatments and demonstrate that measurements corresponding to these images match well with calculations. We show that the Cherenkov component generated inside ocular media is sufficient to be detected by the patient.