Method of and apparatus for detecting diseased tissue by sensing two bands of infrared radiation
A method of and apparatus for detecting diseased tissue based upon infrared imaging in two different bands of infrared wavelengths is described. The use of two series of infrared images taken in two different bands of infrared wavelengths increases sensitivity to the subtle temperature changes caused by diseased skin and tissue, especially in the case of cancerous tissue. By sensing skin temperature, the homogeneity thereof, the time variations thereof and the correlation between the two series of infrared images, the present invention decreases the rate of false positives and false negatives. The increased discrimination due to two series of infrared images allows for reliable detection of diseased or cancerous tissue even in the presence of skin tone variations such as birthmarks, tattoos and freckles. The present invention finds special application in the field of breast cancer detection where subtle skin temperature variations may readily be sensed using two series of infrared images.
1. A method of detecting diseased tissue comprising:
recording first and second series of infrared images of a predetermined area of tissue, the first and second series of infrared images recorded in respective first and second bands of infrared wavelengths, the second band of infrared wavelengths different from the first band of infrared wavelengths;
converting the first and second series of infrared images into corresponding first and second series of thermal images having a plurality of subareas;
determining a first plurality of average temperature values for each of the plurality of subareas, each of the first plurality of average temperature values for each of the plurality of subareas determined from a corresponding one of the first series of thermal images;
determining a first average temperature using the first plurality of average temperature values;
determining a second plurality of average temperature values for each of the plurality of subareas, each of the second plurality of average temperature values for each of the plurality of subareas determined from a corresponding one of the second series of thermal images;
determining a second average temperature using the second plurality of average temperature values; and
analyzing the first and second pluralities of average temperature values for each of the plurality of subareas,
wherein when a spatial distribution of the first plurality of average temperature values corresponding to a cluster comprising at least six adjacent subareas is less than about 20% or more than about 100% of the first average temperature, tissue corresponding to the cluster is preliminarily determined to be diseased, and
wherein when the tissue corresponding to the cluster is preliminarily determined to be diseased and when a spatial distribution of the second plurality of average temperature values corresponding to the cluster is less than about 20% or more than about 100% of the second average temperature, tissue corresponding to the cluster is confirmed to be diseased.
2. A method of detecting diseased tissue in accordance with claim 1 , wherein analyzing the first and second pluralities of average temperature values for each of the plurality of subareas further includes:
determining a first contributing frequency of the first plurality of average temperature values for each of the plurality of subareas using the first series of thermal images;
determining first lower and upper threshold frequencies using the first contributing frequency of each of the subareas;
determining a second contributing frequency of the second plurality of average temperature values for each of the plurality of subareas using the second series of thermal images; and
determining second lower and upper threshold frequencies using the second contributing frequency of each of the subareas,
wherein when a spatial distribution of first contributing frequencies of the cluster is less than the first lower threshold frequency or more than the first upper threshold frequency, tissue corresponding to the cluster is preliminarily determined to be diseased, and
wherein when tissue corresponding to the cluster is preliminarily determined to be diseased and when a spatial distribution of second contributing frequencies of the cluster is less than the second lower threshold frequency or more than the second upper threshold frequency, tissue corresponding to the cluster is confirmed to be diseased.
3. A method of detecting diseased tissue in accordance with claim 1 , wherein analyzing the first and second pluralities of average temperature values for each of the plurality of subareas further includes:
determining a first contributing frequency of the first plurality of average temperature values for each of the plurality of subareas using the first series of thermal images;
determining a first average amplitude using the first contributing frequency for the plurality of subareas;
determining a second contributing frequency of the second plurality of average temperature values for each of the plurality of subareas using the second series of thermal images; and
determining a second average amplitude using the second contributing frequency for the plurality of subareas,
wherein when a spatial distribution of first amplitudes of first contributing frequencies of the cluster is less than about 20% or more than about 100% of the first average amplitude, tissue corresponding to the cluster is preliminarily determined to be diseased, and
wherein when tissue corresponding to the cluster is preliminarily determined to be diseased and when a spatial distribution of second amplitudes of second contributing frequencies of the cluster is less than about 20% or more than about 100% of the second average amplitude, tissue corresponding to the cluster is confirmed to be diseased.
4. A method of detecting diseased tissue in accordance with claim 1 , wherein analyzing the first and second pluralities of average temperature values for each of the plurality of subareas further includes analyzing the first and second pluralities of average temperature values for each of the plurality of subareas using a fast Fourier transform analysis.
5. An apparatus for detecting diseased tissue comprising:
an imager for recording first and second series of infrared images of a predetermined area of tissue, the first and second series of infrared images recorded in respective first and second bands of infrared wavelengths, the second band of infrared wavelengths different from the first band of infrared wavelengths;
a converter for converting the first and second series of infrared images into corresponding first and second series of thermal images having a plurality of subareas;
an averager for determining a first plurality of average temperature values for each of the plurality of subareas, each of the first plurality of average temperature values for each of the plurality of subareas determined from a corresponding one of the first series of thermal images, the averager for determining a first average temperature using the first plurality of average temperature values, the averager for determining a second plurality of average temperature values for each of the plurality of subareas, each of the second plurality of average temperature values for each of the plurality of subareas determined from a corresponding one of the second series of thermal images, and the averager for determining a second average temperature using the second plurality of average temperature values; and
an analyzer for analyzing the first and second pluralities of average temperature values for each of the plurality of subareas,
wherein when a spatial distribution of the first plurality of average temperature values corresponding to a cluster comprising at least six adjacent subareas is less than about 20% or more than about 100% of the first average temperature, tissue corresponding to the cluster is preliminarily determined to be diseased, and
wherein when the tissue corresponding to the cluster is preliminarily determined to be diseased and when a spatial distribution of the second plurality of average temperature values corresponding to the cluster is less than about 20% or more than about 100% of the second average temperature, tissue corresponding to the cluster is confirmed to be diseased.
6. An apparatus for detecting diseased tissue in accordance with claim 5 , further comprising a frequency analyzer for determining a first contributing frequency of the first plurality of average temperature values for each of the plurality of subareas using the first series of thermal images, the frequency analyzer for determining first lower and upper threshold frequencies using the first contributing frequency of each of the subareas, the frequency analyzer for determining a second contributing frequency of the second plurality of average temperature values for each of the plurality of subareas using the second series of thermal images, and the frequency analyzer for determining second lower and upper threshold frequencies using the second contributing frequency of each of the subareas,
wherein the analyzer is further adapted for analyzing the first and second contributing frequencies for each of the plurality of subareas,
wherein when a spatial distribution of first contributing frequencies of the cluster is less than the first lower threshold frequency or more than the first upper threshold frequency, tissue corresponding to the cluster is preliminarily determined to be diseased, and
wherein when tissue corresponding to the cluster is preliminarily determined to be diseased and when a spatial distribution of second contributing frequencies of the cluster is less than the second lower threshold frequency or more than the second upper threshold frequency, tissue corresponding to the cluster is confirmed to be diseased.
7. An apparatus for detecting diseased tissue in accordance with claim 5 , further comprising a frequency analyzer for determining a first contributing frequency of the first plurality of average temperature values for each of the plurality of subareas using the first series of thermal images, the frequency analyzer for determining a first average amplitude using the first contributing frequency for the plurality of subareas, the frequency analyzer for determining a second contributing frequency of the second plurality of average temperature values for each of the plurality of subareas using the second series of thermal images, and the frequency analyzer for determining a second average amplitude using the second contributing frequency for the plurality of subareas,
wherein the analyzer is further adapted for analyzing the first and second contributing frequencies for each of the plurality of subareas,
wherein when a spatial distribution of first amplitudes of first contributing frequencies of the cluster is less than about 20% or more than about 100% of the first average amplitude, tissue corresponding to the cluster is preliminarily determined to be diseased, and
wherein when tissue corresponding to the cluster is preliminarily determined to be diseased and when a spatial distribution of second amplitudes of second contributing frequencies of the cluster is less than about 20% or more than about 100% of the second average amplitude, tissue corresponding to the cluster is confirmed to be diseased.
8. An apparatus for detecting diseased tissue in accordance with claim 5 , wherein the analyzer further uses a fast Fourier transform analysis.