Systems and methods for detecting impairment of an individual
System and methods are provided for detecting impairment of an individual. The method involves operating a processor to: receive at least one image associated with the individual; and identify at least one feature in each image. The method further involves operating the processor to, for each feature: generate an intensity representation for that feature; apply at least one impairment analytical model to the intensity representation to determine a respective impairment likelihood; and determine a confidence level for each impairment likelihood based on characteristics associated with at least the applied impairment analytical model and that feature. The method further involves operating the processor to: define the impairment of the individual based on at least one impairment likelihood and the respective confidence level.
1. A method for detecting impairment of an individual, the method comprising operating a processor to:
receive at least two images associated with the individual, wherein a first image depicts a first portion of the individual and a second image depicts a second portion of the individual, the second portion of the individual being different from the first portion of the individual, and the first image being different from the second image;
identify at least one feature in each image of the at least two images, wherein the first image comprises a first feature type and the second image comprises a second feature type, the second feature type being different from the first feature type;
for each feature identified in each image:
generate an intensity representation for a region of that image associated with the feature, the intensity representation corresponding to a frequency distribution of pixel intensities for pixels in the region and each pixel in the region is associated with a heat intensity of the feature;
select at least one impairment analytical model related to the feature;
apply the at least one impairment analytical model to the intensity representation to determine a respective at least one impairment likelihood;
determine a feature reliability indicator for the feature, the feature reliability indicator representing a level of reliability of that feature for determining impairment of the individual;
determine a confidence level for each of the at least one impairment likelihood based on the feature reliability indicator and characteristics associated with the corresponding at least one impairment analytical model and that feature; and
define the impairment of the individual based on the at least one impairment likelihood and the respective confidence level generated from the at least one feature in each image of the at least two images.
2. The method of claim 1 , wherein determining the confidence level for each of the at least one impairment likelihood, based on the feature reliability indicator and the characteristics associated with the corresponding at least one impairment analytical model and the feature, comprises:
determining whether an image quality of the at least two images satisfies a quality threshold;
generating a quality indicator according to whether the image quality satisfies the quality threshold; and
determining the confidence level for each of the at least one impairment likelihood based at least on the quality indicator.
3. The method of claim 1 , wherein determining the confidence level for each of the at least one impairment likelihood, based on the feature reliability indicator and the characteristics associated with the corresponding at least one impairment analytical model and the feature, comprises:
determining an image reliability indicator associated with an image view of the at least two images; and
determining the confidence level for each of the at least one impairment likelihood based at least on the image reliability indicator.
4. The method of claim 1 , wherein determining the confidence level for each of the at least one impairment likelihood, based on the feature reliability indicator and the characteristics associated with the corresponding at least one impairment analytical model and the feature, comprises:
determining at least one model reliability indicator associated with the at least one impairment analytical model; and
determining the confidence level for each of the at least one impairment likelihood based at least on the at least one model reliability indicator.
5. The method of claim 1 , further comprising operating the processor to:
receive at least one audio recording involving the individual;
identify at least one audio property of the at least one audio recording to analyze;
for each audio property:
select at least one audio analytical model for that audio property;
apply the at least one audio analytical model to the at least one audio recording to determine a respective impairment likelihood of the individual; and
determine the confidence level for each impairment likelihood based on characteristics associated with at least the applied audio analytical model and that audio property.
6. The method of claim 1 , wherein defining the impairment of the individual comprises generating an impairment indicator indicating an impairment level of the individual based on the at least one impairment likelihood and the respective confidence level generated from the at least one feature in each image of the at least two images.
7. The method of claim 1 , wherein receiving the at least two images associated with the individual comprises receiving:
the first image depicting a first view of a portion of the individual; and
the second image depicting a second view of the portion of the individual, the second view being different from the first view.
8. A system for detecting impairment of an individual, the system comprising:
a data storage to store one or more impairment analytical models; and
a processor operable to:
receive, via a network, at least two images associated with the individual, wherein a first image depicts a first portion of the individual and a second image depicts a second portion of the individual, the second portion of the individual being different from the first portion of the individual, and the first image being different from the second image;
identify at least one feature in each image of the at least two images, wherein the first image comprises a first feature type and the second image comprises a second feature type, the second feature type being different from the first feature type;
for each feature identified in each image:
generate an intensity representation for a region of that image associated with the feature, the intensity representation corresponding to a frequency distribution of pixel intensities for pixels in the region and each pixel in the region is associated with a heat intensity of the feature;
select at least one impairment analytical model related to the feature from the one or more impairment analytical models;
apply the at least one impairment analytical model the intensity representation to determine a respective at least one impairment likelihood;
determine a feature reliability indicator for the feature, the feature reliability indicator representing a level of reliability of that feature for determining impairment of the individual;
determine a confidence level for each of the at least one impairment likelihood based on the feature reliability indicator and characteristics associated with the corresponding at least one impairment analytical model and that feature; and
define the impairment of the individual based on the at least one impairment likelihood and the respective confidence level generated from the at least one feature in each image of the at least two images.
9. The system of claim 8 , wherein the processor is operable to:
determine whether an image quality of the at least two images satisfies a quality threshold;
generate a quality indicator according to whether the image quality satisfies the quality threshold; and
determine the confidence level for each of the at least one impairment likelihood based at least on the quality indicator.
10. The system of claim 8 , wherein the processor is operable to:
determine an image reliability indicator associated with an image view of the at least two images; and
determine the confidence level for each of the at least one impairment likelihood based at least on the image reliability indicator.
11. The system of claim 8 , wherein the processor is operable to:
determine a model reliability indicator associated with the at least one impairment analytical model; and
determine the confidence level for each of the at least one impairment likelihood based at least on the model reliability indicator.
12. The system of claim 8 , wherein the processor is operable to:
receive at least one audio recording involving the individual;
identify at least one audio property of the at least one audio recording to analyze;
for each audio property:
select at least one audio analytical model for that audio property;
apply the at least one audio analytical model to the at least one audio recording to determine a respective impairment likelihood of the individual; and
determine the confidence level for each impairment likelihood based on characteristics associated with at least the applied audio analytical model and that audio property.
13. The system of claim 8 , wherein the processor is operable to generate an impairment indicator indicating an impairment level of the individual based on the at least one impairment likelihood and the respective confidence level generated from the at least one feature in each image of the at least two images.
14. The system of claim 8 , wherein the processor is operable to receive:
the first image depicting a first view of a portion of the individual; and
the second image depicting a second view of the portion of the individual, the second view being different from the first view.