Contour-based determination of malignant tissue in a thermal image
What is disclosed is a system and method for contour-based determination of malignant tissue in a thermal image of a patient for cancer screening. In one embodiment, the method involves receiving a thermal image for cancer screening. Pixels in the thermal image with a higher temperature value are displayed in a first color and pixels with a lower temperature value are displayed in a second color. Pixels with temperature values between the lower and higher temperature values are displayed in gradations of color between the first and second colors. The thermal image is then analyzed to identify a patch of pixels with an elevated temperature relative to a temperature of pixels associated with surrounding tissue. Thereafter, tissue associated with the identified patch is determined to be malignant or non-malignant based a measure of irregularity calculated for boundary contour encompassing that patch of pixels.
1. A method for contour-based determination of malignant tissue in a thermal image of a patient for cancer screening, the method comprising:
receiving a thermal image of a patient for cancer screening, wherein the thermal image is captured by a thermal imaging camera, the thermal imaging camera comprising:
an array of sensors that converts infrared energy into electrical signals on a per-pixel basis;
a lens that focuses the infrared energy from the patient's breast onto the array of sensors, wherein the array of sensors detects temperature values from the patient's breast; and
a specialized processor that processes the detected temperature values into at least one block of pixels to generate the thermal image,
wherein pixels in the thermal image with a higher temperature value being displayed in a first color and pixels with a lower temperature value being displayed in a second color, pixels with temperature values between the lower and higher temperature values being displayed in gradations of color between the first and second colors;
analyzing the thermal image to identify a patch of pixels with an elevated temperature relative to a temperature of pixels associated with surrounding tissue;
determining a boundary contour of the identified patch of pixels;
calculating a measure of irregularity of the boundary contour, wherein the measure of irregularity is calculated by
selecting a plurality of points along a best-fit ellipse around the pixel patch;
calculating a distance between the points along the boundary contour and the points along a best-fit ellipse; and
determining, in response to the distance being at or above a threshold, that a shape of the boundary contour of the pixel patch is determined to be irregular, and regular otherwise;
determining, based on the measure of irregularity, that tissue associated with the pixel patch is malignant and non-malignant otherwise; and
displaying the irregular boundary contour that represents the malignant tissue and the regular boundary contour that represents the non-malignant tissue in different colors.
2. The method of claim 1 , wherein calculating the measure of irregularity comprises:
calculating a centroid of the pixel patch;
calculating an area of the pixel patch;
calculating an area of a circle with a radius equal to a distance from the centroid to a farthest point on the boundary contour; and
determining, in response to the ratio of an area of the circle to the area of the pixel patch being above a threshold, that the shape of the boundary contour of the pixel patch is determined to be irregular, and regular otherwise.
3. The method of claim 1 , wherein calculating the measure of irregularity comprises:
calculating a centroid of the pixel patch;
selecting a plurality of points along a best-fit circle around the pixel patch;
calculating a first distance of a line from each of the selected points to the centroid;
calculating a second distance of each of the lines from a point along the boundary contour to the centroid;
calculating a standard deviation of the first and second distances; and
determining, in response to an amount of the deviation being at or above a threshold, that the shape of the boundary contour of the pixel patch is determined to be irregular, and regular otherwise.
4. The method of claim 1 , wherein calculating the measure of irregularity comprises:
selecting a plurality of points along a best-fit polynomial curve with an N-degree polynomial around the pixel patch, where N≦5 and N≧2;
calculating a distance between the points along the best fit polynomial curve and the points on the boundary contour; and
determining, in response to the distance being at or above a threshold, that the shape of the boundary contour of the pixel patch is determined to be irregular, and regular otherwise.
5. The method of claim 1 , wherein the image is obtained from any imaging modality.
6. The method of claim 1 , wherein the image is one of: a breast area, and a non-breast area.
7. The method of claim 1 , wherein, in response to said tissue being malignant, performing any of: initiating an alert, and signaling a professional.
8. The method of claim 1 , wherein said images are a streaming video and a determination of malignancy occurs in real-time.
9. A system for contour-based determination of malignant tissue in a thermal image of a patient for cancer screening, the system comprising:
a thermal imaging camera, wherein a thermal image is captured by the thermal imaging camera, the thermal imaging camera comprising:
an array of sensors that converts infrared energy into electrical signals on a per-pixel basis;
a lens that focuses the infrared energy from the patient's breast onto the array of sensors, wherein the array of sensors detects temperature values from the patient's breast; and
a specialized processor that processes the detected temperature values into at least one block of pixels to generate the thermal image;
a storage device; and
a processor in communication with said storage device, said processor executing machine readable instructions for performing:
receiving a thermal image of a patient for cancer screening from said thermal imaging camera, pixels in the thermal image with a higher temperature value being displayed in a first color and pixels with a lower temperature value being displayed in a second color, pixels with temperature values between the lower and higher temperature values being displayed in gradations of color between the first and second colors;
analyzing the thermal image to identify a patch of pixels with an elevated temperature relative to a temperature of pixels associated with surrounding tissue;
determining a boundary contour of the identified patch of pixels;
calculating a measure of irregularity of the boundary contour, wherein the measure of irregularity is calculated by
selecting a plurality of points along a best-fit ellipse around the pixel patch;
calculating a distance between the points along the boundary contour and the points along a best-fit ellipse; and
determining, in response to the distance being at or above a threshold, that a shape of the boundary contour of the pixel patch is determined to be irregular, and regular otherwise;
determining, based on the measure of irregularity, that tissue associated with the pixel patch is malignant and non-malignant otherwise; and
displaying the irregular boundary contour that represents the malignant tissue and the regular boundary contour that represents the non-malignant tissue in different colors.
10. The system of claim 9 , wherein calculating the measure of irregularity comprises:
calculating a centroid of the pixel patch;
calculating an area of the pixel patch;
calculating an area of a circle with a radius equal to a distance from the centroid to a farthest point on the boundary contour; and
determining, in response to the ratio of an area of the circle to the area of the pixel patch being above a threshold, that the shape of the boundary contour of the pixel patch is determined to be irregular, and regular otherwise.
11. The system of claim 9 , wherein calculating the measure of irregularity comprises:
calculating a centroid of the pixel patch;
selecting a plurality of points along a best-fit circle around the pixel patch;
calculating a first distance of a line from each of the selected points to the centroid;
calculating a second distance of each of the lines from a point along the boundary contour to the centroid;
calculating a standard deviation of the first and second distances; and
determining, in response to an amount of the deviation being at or above a threshold, that the shape of the boundary contour of the pixel patch is determined to be irregular, and regular otherwise.
12. The system of claim 9 , wherein calculating the measure of irregularity comprises:
selecting a plurality of points along a best-fit polynomial curve with an N-degree polynomial around the pixel patch, where N≦5 and N≧2;
calculating a distance between the points along the best fit polynomial curve and the points on the boundary contour; and
determining, in response to the distance being at or above a threshold, that the shape of the boundary contour of the pixel patch is determined to be irregular, and regular otherwise.
13. The system of claim 9 , wherein the image is obtained from any imaging modality.
14. The system of claim 9 , wherein the image is one of: a breast area, and a non-breast area.
15. The system of claim 9 , wherein, in response to said tissue being malignant, performing any of: initiating an alert, and signaling a professional.
16. The system of claim 9 , wherein said images are a streaming video and a determination of malignancy occurs in real-time.