AI system for predicting reading time and reading complexity for reviewing 2D/3D breast images
Examples of the present disclosure describe systems and methods for predicting the reading time and/or reading complexity of a breast image. In aspects, a first set of data relating to the reading time of breast images may be collected from one or more data sources, such as image acquisition workstations, image review workstations, and healthcare professional profile data. The first set of data may be used to train a predictive model to predict/estimate an expected reading time and/or an expected reading complexity for various breast images. Subsequently, a second set of data comprising at least one breast image may be provided as input to the trained predictive model. The trained predictive model may output an estimated reading time and/or reading complexity for the breast image. The output of the trained predictive model may be used to prioritize mammographic studies or optimize the utilization of available time for radiologists.
1. A system comprising:
a processor; and
memory coupled to the processor, the memory comprising computer executable instructions that, when executed by the processor, performs a method comprising:
collecting a first set of data, wherein the first set of data comprises:
first mammographic exam data for one or more patients;
user profile data for one or more mammographic exam readers of the first mammographic exam data; and
evaluation data for the one or more mammographic exam readers;
providing the first set of data to a predictive model, wherein the first set of data is used to train the predictive model to determine a reading time for the first mammographic exam data;
collecting a second set of data, wherein the second set of data comprises at least second mammographic exam data for a patient, wherein the second mammographic exam data includes breast image data and one or more factors determined according to processing of the breast image data;
applying the second set of data to the trained predictive model;
receiving, from the trained predictive model, an estimated reading time for the second mammographic exam data based on the one or more factors determined according to processing of the breast image data; and
displaying the estimated reading time.
2. The system of claim 1 , wherein the method further comprises:
using the estimated reading time to optimize a workload distribution for a plurality of mammographic exam readers.
3. The system of claim 1 , wherein the first mammographic exam data comprises at least one of: pixel image data and image header data.
4. The system of claim 3 , wherein the pixel image data is used to determine at least one of: breast tissue patterns, breast density, breast complexity, or breast thickness.
5. The system of claim 3 , wherein the image header data is indicative of at least one of: type of study used to collect the mammographic exam data, image resolution of the mammographic exam data, or type of processing used create the mammographic exam data.
6. The system of claim 3 , wherein the user profile data comprises at least one of: an amount of experience of the reader, an age of the reader, or expertise of the reader.
7. The system of claim 3 , wherein the evaluation data comprises at least one of: time of day the evaluation data was collected, type of tools used to collect the evaluation data, usage data for the tools used to collect the evaluation data, and hanging protocol used to generate the evaluation data.
8. The system of claim 1 , wherein the predictive model is at least one of:
a neural network; or
a support vector machine.
9. The system of claim 1 , wherein the predictive model is further trained to output a complexity score, the complexity score representing an estimated complexity of reading the second mammographic exam data.
10. The system of claim 1 , wherein the first set of data further comprises reading time estimates provided by the one or more mammographic exam readers.
11. The system of claim 1 , wherein the one or more factors determined according to processing of the breast image data include one or more of: a number of lesions detected in the breast image data, a type of anomaly identified in the breast image data, a location of evaluation within the breast image data, a determination of symmetry between the patient's breasts, a number of image slices generated, a breast density, a breast thickness, a breast area, a breast tissue composition structure, a breast tissue patterns, and a number of computer-aided detection markers.
12. A method of predicting reading time of a mammographic exam, the method comprising:
collecting a first set of data, wherein the first set of data comprises:
first mammographic exam data for one or more patients;
user profile data for one or more mammographic exam readers of the first mammographic exam data; and
evaluation data for the one or more mammographic exam readers;
providing the first set of data to a predictive model, wherein the first set of data is used to train the predictive model to determine a reading time for the first mammographic exam data;
collecting a second set of data, wherein the second set of data comprises at least second mammographic exam data for a patient, wherein the second mammographic exam data includes breast image data and one or more factors determined according to processing of the breast image data;
applying the second set of data to the trained predictive model;
receiving, from the trained predictive model, an estimated reading time for the second mammographic exam data based on the one or more factors determined according to processing of the breast image data; and
displaying the estimated reading time.
13. The method of claim 12 , wherein the first mammographic exam data comprises at least one of: pixel image data and image header data, the image header data indicating at least one of:
a type of study performed;
an image resolution;
a type of hardware system used to collect the first mammographic exam data; or
an image processing method used to collect the first mammographic exam data.
14. The method of claim 12 , wherein the predictive model is further trained to output a complexity score, the complexity score representing an estimated complexity of reading the second mammographic exam data.
15. The method of claim 14 , wherein the complexity score is based on at least one of:
breast shape;
breast density; or
number of detected lesions.
16. The method of claim 12 , wherein the predictive model is further trained to output at least one of:
an identification of a recommended mammographic exam reader; or
an available time slot for reading the second mammographic exam data.
17. The method of claim 12 , wherein the estimated reading time is used to automatically determine a workload distribution for a mammographic exam reader.
18. The method of claim 12 , wherein the estimated reading time is used to automatically assign the second mammographic exam data to a mammographic exam reader.
19. The method of claim 18 , further comprising:
determining an actual reading time required by the mammographic exam reader to read the second mammographic exam data;
providing the actual reading time to the trained predictive model; and
updating the trained predictive model based on the actual reading time.
20. A computing device comprising:
a user interface;
a processor;
memory comprising executable instructions that enable the processor to:
receive from a user, via the user interface, a reading time estimate for first mammographic exam data;
collect a first set of data, wherein the first set of data comprises:
the first mammographic exam data;
user profile data for one or more mammographic exam readers of the first mammographic exam data;
evaluation data for the one or more mammographic exam readers; and
the reading time estimate for mammographic exam data;
provide the first set of data to a predictive model, wherein the first set of data is used to train the predictive model to determine a reading time for the first breast image data;
collect a second set of data, wherein the second set of data comprises at least second mammographic exam data, wherein the second mammographic exam data includes breast image data and one or more factors determined according to processing of the breast image data;
apply the second set of data to the trained predictive model based on the one or more factors determined according to processing of the breast image data;
receive, from the trained predictive model, an estimated reading time for the second mammographic exam data; and
display the estimated reading time.