Methods for collecting and analyzing thermal data based on breast surface temperature to determine suspect conditions
First Claim
1. A method for classifying breast tissue of a subject as suspect tissue or non-suspect tissue, comprising the steps of:
- placing a plurality of temperature sensors on a breast of the subject at a set of predetermined positions;
sensing a set of temperature readings from the plurality of temperature sensors over a predetermined period; and
sending the set of temperature readings to an ensemble of classifier systems including two or more classifier systems, the ensemble of classifier systems deriving two or more output labels from the set of temperature readings identifying the breast tissue as suspect tissue or non-suspect tissue, and combining the two or more output labels into a single diagnosis identifying the breast tissue as suspect tissue or non-suspect tissue, wherein a weight of each label from the two or more output labels determines a relative contribution of the label towards the single diagnosis, the ensemble of classifier systems trained using one or more actual temperature readings including an above average temperature reading corresponding to suspect tissue, an average temperature reading corresponding to non-suspect tissue, and an anomalous temperature reading corresponding to non-suspect tissue.
4 Assignments
0 Petitions
Accused Products
Abstract
A portable computing device or microprocessor/storage system including temperature sensors used to collect temperature readings of a breast tissue of a subject. The device would collect data from the sensors at regular time intervals over a period of time. All of the generated temperature data is stored in the portable computing or storage device. The sensors are placed on the greatest areas of interest on the breast, based on where most cancers develop, by using a sensor placeholder. The sensor placeholder would be lobate shaped, with the sensor placeholder aligning with the glandular regions of the breast where cancers are most likely to develop. The temperature data is then analyzed by one or more classifier systems and classified as either suspect or non-suspect tissue.
25 Citations
27 Claims
-
1. A method for classifying breast tissue of a subject as suspect tissue or non-suspect tissue, comprising the steps of:
-
placing a plurality of temperature sensors on a breast of the subject at a set of predetermined positions; sensing a set of temperature readings from the plurality of temperature sensors over a predetermined period; and sending the set of temperature readings to an ensemble of classifier systems including two or more classifier systems, the ensemble of classifier systems deriving two or more output labels from the set of temperature readings identifying the breast tissue as suspect tissue or non-suspect tissue, and combining the two or more output labels into a single diagnosis identifying the breast tissue as suspect tissue or non-suspect tissue, wherein a weight of each label from the two or more output labels determines a relative contribution of the label towards the single diagnosis, the ensemble of classifier systems trained using one or more actual temperature readings including an above average temperature reading corresponding to suspect tissue, an average temperature reading corresponding to non-suspect tissue, and an anomalous temperature reading corresponding to non-suspect tissue. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A method for classifying breast tissue of a subject as suspect tissue or non-suspect tissue, comprising the steps of:
-
receiving a set of temperature readings from a plurality of temperature sensors positioned over a breast of the subject, the set of temperature readings including temperature readings over a predetermined period of time; inputting the set of temperature readings into an ensemble of classifier systems including two or more classifier systems; deriving two or more output labels based on the set of temperature readings identifying the breast tissue as suspect tissue or non-suspect tissue with the ensemble of classifier systems, the ensemble of classifier systems trained using one or more actual temperature readings including an above average temperature reading corresponding to suspect tissue, an average temperature reading corresponding to non-suspect tissue, and an anomalous temperature reading corresponding to non-suspect tissue; and combining the two or more output labels into a single diagnosis identifying the breast tissue as suspect tissue or non-suspect tissue, wherein a weight of each label from the two or more output labels determines a relative contribution of the label towards the single diagnosis. - View Dependent Claims (12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27)
-
Specification