Apparatus for providing evaluation of bedsore stages and treatment recommendations using artificial intelligence and operation method thereof
Inventors
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12086987-B2
Publication Date
2024-09-10
Expiration Date
Abstract
Provided are a device for managing bedsores and an operating method of the same. The operating method includes acquiring image data of a plurality of existing bedsores, acquiring existing bedsore-related information corresponding to the image data of the plurality of existing bedsores, training a convolutional neural network (CNN) with relationships between the image data of the plurality of existing bedsores and the existing bedsore-related information to acquire a machine learning model, acquiring bedsore image data of a current patient, applying the machine learning model to the bedsore image data of the current patient to determine information on a bedsore or bedsore treatment information of the current patient, and outputting the information on the bedsore or the bedsore treatment information of the current patient.
Core Innovation
The invention provides an operating method and a device for managing a bedsore. The method acquires image data of a plurality of existing bedsores and acquires existing bedsore-related information corresponding to the image data, including at least one of bedsore stage information, bedsore diagnosis information, bedsore location information in the image data, and existing bedsore treatment information. The method trains a convolutional neural network (CNN) with relationships between the image data and the existing bedsore-related information to acquire a machine learning model.
The method acquires bedsore image data of a current patient and applies the machine learning model to determine information on a bedsore or bedsore treatment information of the current patient. The information on the bedsore includes at least one of bedsore stage information and bedsore diagnosis information, and the bedsore treatment information includes dressing information for treating a bedsore. The results are output as information on the bedsore or the bedsore treatment information.
In one implementation, the CNN training is refined using a region-based convolutional neural network (R-CNN) approach. Image-data preprocessing includes preprocessing and annotation processing, and the training uses a region proposal network (RPN) with relationships between the image data and bedsore stage information. In another implementation, the treatment determination includes a dressing recommendation algorithm that determines required dressing information from bedsore stage information and bedsore diagnosis information, and the dressing recommendation algorithm is determined using a deep decision tree boosting model.
Claims Coverage
Independent claims are directed to an operating method and a bedsore management device. The claim set covers four inventive features: CNN-based training on existing bedsore image data and related information, determination of current-patient bedsore stage/diagnosis and dressing treatment information, region-based CNN training using RPN and annotation processing, and a dressing recommendation algorithm determined by deep decision tree boosting.
Bedsore image-to-information learning with existing bedsore-related information
Acquiring image data of a plurality of existing bedsores; acquiring existing bedsore-related information corresponding to the image data; training a convolutional neural network (CNN) with relationships between the image data and the existing bedsore-related information to acquire a machine learning model; and outputting information on a bedsore and/or bedsore treatment information of a current patient based on applying the machine learning model to the current patient's bedsore image data.
Determination of bedsore stage/diagnosis and treatment dressing information for a current patient
Applying the machine learning model to bedsore image data of a current patient to determine information on a bedsore or bedsore treatment information; the information on the bedsore includes at least one of bedsore stage information and bedsore diagnosis information; and the bedsore treatment information includes dressing information for treating the bedsore.
Region-based CNN training using RPN and annotation processing
Preprocessing and annotation processing for image data of a plurality of existing bedsores and training a region-based convolutional neural network (R-CNN) using a region proposal network (RPN) with relationships between the image data and bedsore stage information to acquire relationship information and associated outputs including bedsore location information and bedsore stage information.
Dressing recommendation determined by deep decision tree boosting
Determining the required dressing information using a dressing recommendation algorithm based on bedsore stage information and bedsore diagnosis information, where the dressing recommendation algorithm is determined using a deep decision tree boosting model.
Overall, the independent-claim coverage requires a CNN-based model trained on existing bedsore image data paired with bedsore-related information to determine bedsore stage/diagnosis and dressing treatment information for a current patient, with additional coverage for R-CNN training using RPN/annotation processing and treatment determination via a dressing recommendation algorithm determined by deep decision tree boosting.
Stated Advantages
Documented Applications
No documented applications found
Interested in licensing this patent?