Analyzing surgical videos to identify a billing coding mismatch
Disclosed herein are apparatus, system, method, and computer-readable medium aspects for using surgical video analysis for improving compliance with medical guidelines and improving processing of medical bills, medical malpractice claims, and insurance claims. Aspects disclosed herein utilize intracorporeal video footage, image analysis, and notifications to optimize correspondences among medical procedure information, medical guidelines, and medical transaction information.
1 . A non-transitory computer readable medium containing instructions that, when executed by at least one processor, cause the at least one processor to execute operations to identify a billing coding mismatch, the operations comprising:
receiving a medical reimbursement code associated with a surgical procedure;
receiving surgical video of the surgical procedure;
performing image analysis on the surgical video to determine whether a match exists between the medical reimbursement code and the surgical video, the image analysis further comprising:
analyzing the surgical video to detect an adverse event in the surgical procedure,
calculating a result value by executing a convolution of at least part of the surgical video, and
determining whether the match exists based on the detected adverse event and the result value; and
when the match is determined not to exist, outputting an indicator of a lack of support in the surgical video for the medical reimbursement code.
2 . The non-transitory computer readable medium of claim 1 , wherein the performing the image analysis comprises analyzing the surgical video to determine a level of complexity of the surgical procedure, and determining whether an alleged reimbursable event associated with the medical reimbursement code took place in the surgical procedure based on the determined level of complexity.
3 . The non-transitory computer readable medium of claim 1 , wherein the performing the image analysis comprises analyzing the surgical video to determine a level of guideline compliance during the surgical procedure, and determining whether an alleged reimbursable event associated with the medical reimbursement code took place in the surgical procedure based on the determined level of guideline compliance.
4 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise accessing a medical record of the patient, and determining whether an alleged reimbursable event associated with the medical reimbursement code took place in the surgical procedure based on the medical record of the patient.
5 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise accessing an audio recording captured during the surgical procedure, and determining whether an alleged reimbursable event associated with the medical reimbursement code took place in the surgical procedure based on the audio recording.
6 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise:
analyzing the surgical video to detect a surgical tool; and
basing the determination of whether the match exists between the medical reimbursement code and the surgical video on the detected surgical tool.
7 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise:
analyzing the surgical video to detect an anatomical structure;
analyzing the surgical video to determine a condition of the anatomical structure; and
basing the determination of whether the match exists between the medical reimbursement code and the surgical video on the condition of the anatomical structure.
8 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise:
analyzing the surgical video to detect an interaction between a surgical tool and an anatomical structure; and
basing the determination of whether the match exists between the medical reimbursement code and the surgical video on the detected interaction between the surgical tool and the anatomical structure.
9 . The non-transitory computer readable medium of claim 8 , wherein the operations further comprise:
analyzing the surgical video to determine a characteristic of the interaction between the surgical tool and the anatomical structure; and
basing the determination of whether the match exists between the medical reimbursement code and the surgical video on the characteristic of the interaction between the surgical tool and the anatomical structure.
10 . The non-transitory computer readable medium of claim 9 , wherein the characteristic is a duration of the interaction.
11 . The non-transitory computer readable medium of claim 9 , wherein the characteristic is a type of the interaction.
12 . The non-transitory computer readable medium of claim 9 , wherein the characteristic is a state of the surgical tool during the interaction.
13 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise:
calculating a convolution of at least part of the surgical video to obtain a result value; and
basing the determination of whether the match exists between the medical reimbursement code and the surgical video on the result value.
14 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise using a machine learning model to analyze the surgical video to determine whether the match exists between the medical reimbursement code and the surgical video.
15 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise:
analyzing the surgical video to classify a mismatch between the medical reimbursement code and the surgical video; and
basing the indicator on a classification of the mismatch between the medical reimbursement code and the surgical video.
16 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise:
analyzing the surgical video to determine an alternative medical reimbursement code; and
outputting an indication of the alternative medical reimbursement code.
17 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise:
analyzing the surgical video to detect in the surgical procedure a first surgical event and a second surgical event; and
further basing the determination of whether the match exists between the medical reimbursement code and the surgical video on whether the first surgical event precedes the second surgical event in the surgical procedure.
18 . The non-transitory computer readable medium of claim 1 , wherein the operations further comprise:
training a machine learning model using training examples to analyze the surgical video, the training to determine whether particular actions occurred.
19 . A method for performing intracorporeal video analysis operations for identifying a billing coding mismatch, comprising:
receiving a medical reimbursement code associated with a surgical procedure;
receiving surgical video of the surgical procedure;
performing image analysis on the surgical video to determine whether a match exists between the medical reimbursement code and the surgical video, the image analysis further comprising:
analyzing the surgical video to detect an adverse event in the surgical procedure,
calculating a result value by executing a convolution of at least part of the surgical video, and
determining whether the match exists based on the detected adverse event and the result value; and
when the match is determined not to exist, outputting an indicator of a lack of support in the surgical video for the medical reimbursement code.
20 . A system for identifying a billing coding mismatch, the system comprising a processor configured to perform steps comprising:
receiving a medical reimbursement code associated with a surgical procedure;
receiving surgical video of the surgical procedure;
performing image analysis on the surgical video to determine whether a match exists between the medical reimbursement code and the surgical video, the image analysis further comprising:
analyzing the surgical video to detect an adverse event in the surgical procedure,
calculating a result value by executing a convolution of at least part of the surgical video, and
determining whether the match exists based on the detected adverse event and the result value; and
when the match is determined not to exist, outputting an indicator of a lack of support in the surgical video for the medical reimbursement code.
21 . A non-transitory computer readable medium containing instructions that, when executed by at least one processor, cause the at least one processor to execute operations to identify a billing coding mismatch and an alleged reimbursable event, the operations comprising:
receiving a medical reimbursement code associated with a surgical procedure;
receiving surgical video of the surgical procedure;
performing image analysis on the surgical video to determine whether a match exists between the medical reimbursement code and the surgical video, the image analysis further comprising:
calculating a result value by executing a convolution of at least part of the surgical video;
when the match is determined not to exist, outputting an indicator of a lack of support in the surgical video for the medical reimbursement code, the match determined not to exist based on the result value;
accessing a medical record of a patient; and
determining whether the alleged reimbursable event associated with the reimbursement code took place in the surgical procedure based on the medical record of the patient.
22 . A non-transitory computer readable medium containing instructions that, when executed by at least one processor, cause the at least one processor to execute operations to identify a billing coding mismatch using a surgical video and an audio recording captured during a surgical procedure, the operations comprising:
receiving a medical reimbursement code associated with the surgical procedure;
receiving the surgical video of the surgical procedure;
performing image analysis on the surgical video to determine whether a match exists between the medical reimbursement code and the surgical video, the image analysis further comprising:
calculating a result value by executing a convolution of at least part of the surgical video;
accessing an audio recording capturing during the surgical procedure; and
when the match is determined not to exist, outputting an indicator of a lack of support in the surgical video for the medical reimbursement code, the matched determined not to exist based on the result value and the audio recording.