Device for analyzing thermal data based on breast surface temperature for the detection for use in determining cancerous conditions
A device for providing improved classification of breast tissue wherein the tissue is classified as one of cancerous and non-cancerous tissue includes sensors for sensing temperature of a subject breast tissue over a predetermined period and generating signals in response thereto and a computer operatively connected to the sensors for receiving and manipulating the signals operatively associated with a normalizer for applying a normalization to each the signal by a function of X/Y where X is one temperature reading of one of the signals for a particular subject and Y is the maximum temperature obtained from a predetermined number sets of temperature readings obtained a plurality of subject breast tissue from the subject to provide a set of normalized signals used by a trained artificial classifier to produce a signal indicative of cancerous and non-cancerous tissue of the subject.
1. A device for providing improved classification of breast tissue wherein the tissue: is classified as one of cancerous and non-cancerous tissue, which comprises:
means for sensing temperature of a subject breast tissue over a predetermined period and generating signals in response thereto;
means operatively associated with said means for sensing for receiving and manipulating said signals employing a computer based device operatively associated with means for applying a normalization to each said signal, said normalization including a function of X/Y where X is one temperature reading of one of said signals for a particular subject and Y is the maximum temperature obtained from a predetermined number sets of temperature readings obtained a plurality of subject breast tissue from said subject, said normalization to provide a set of normalized signals and said computer based device operatively associated with a trained artificial classifier and wherein said set normalized signals are input through said trained artificial classifier to produce a signal indicative of cancerous and non-cancerous tissue of the subject.
2. The device for providing improved classification of breast tissue of claim 1 , wherein said computer based device is operatively associated with a plurality of trained artificial classifiers and wherein said set normalized signals are input through said trained artificial classifiers to produce signals indicative of cancerous and non-cancerous tissue of the subject.
3. The device for providing improved classification of breast tissue of claim 1 , wherein said classifier is further trained using said normalized signals.
4. The device for providing improved classification of breast tissue of claim 2 , wherein said classifiers are selected from a group of Back-Propagation, Radial Basis Function, Gaussian Mixture Model, Fuzzy Network and Support Vector Machine.
5. A method for providing improved classification of breast tissue wherein the tissue is classified as one of cancerous and non-cancerous tissue, which comprises:
(a) sensing temperature of a subject breast tissue over a predetermined period and generating signals in response thereto;
(b) employing a computer based device having means for receiving and manipulating said signals, applying a normalization to each said signal, said normalization including a function of X/Y where X is one temperature reading of a given set of readings for a particular subject and Y is the maximum temperature obtained from a predetermined number sets of temperature readings obtained a plurality of subject breast tissue from said subject, said normalization to provide a set of normalized signals and said computer based device operatively associated with a trained artificial classifier and wherein said set normalized signals are input through said trained artificial classifier to produce a signal indicative of cancerous and non-cancerous tissue of the subject.
6. The method for providing improved classification of breast tissue of claim 5 , wherein said computer based device is operatively associated with a plurality of trained artificial classifiers and wherein said set normalized signals are input through said trained artificial classifiers to produce signals indicative of cancerous and non-cancerous tissue of the subject.
7. The method for providing improved classification of breast tissue of claim 5 , wherein said classifier is further trained using said normalized signals.
8. The method for providing improved classification of breast tissue of claim 6 , wherein said classifiers are selected from a group of Back-Propagation, Radial Basis Function, Gaussian Mixture Model, Fuzzy Network and Support Vector Machine.