Method and system for computer-aided triage
A system for computer-aided triage can include a router, a remote computing system, and a client application. A method for computer-aided triage can include determining a parameter associated with a data packet, determining a treatment option based on the parameter, and transmitting information to a device associated with a second point of care.
1 . A system for computer-aided triage of a patient at a first point of care, the system comprising:
a set of multiple trained machine learning algorithms, each configured to detect a different condition from a set of different conditions, and each executable by a processing subsystem, the set of multiple trained machine learning algorithms comprising:
a first trained machine learning algorithm configured to detect an aneurysm based on a first feature; and
a second trained machine learning algorithm configured to detect an occlusion based on a second feature; and
the processing subsystem, wherein the processing subsystem is configured to:
determine a suspected condition and a parameter associated with the suspected condition, comprising: concurrently running the first trained machine learning algorithm and the second trained machine learning algorithm of the set of trained machine learning algorithms to process a set of inputs associated with the patient; and
if the parameter exceeds a threshold severity value, automatically initiate a treatment of the patient.
2 . The system of claim 1 , wherein the treatment comprises notifying a specialist of the suspected condition, wherein notifying the specialist comprises bypassing a radiologist workflow of a radiologist reviewing the set of inputs.
3 . The system of claim 2 , wherein the processing subsystem is further configured to receive a secondary input from the specialist, wherein the secondary input initiates a transfer of the patient to a second point of care.
4 . The system of claim 2 , wherein the set of inputs associated with the patient comprises a set of image data in a DICOM format, wherein notifying the specialist further comprises displaying a compressed version of the set of DICOM brain images on a mobile device of the specialist.
5 . The system of claim 1 , wherein the parameter comprises a severity score, wherein the processing subsystem is further configured to:
determine a set of severity scores for each of a plurality of patients; and
reorganize a queue associated with the plurality of patients based on the set of severity scores.
6 . The system of claim 1 , wherein the set of different conditions comprises different neurovascular conditions.
7 . The system of claim 1 , wherein the treatment comprises a transfer of the patient to a specialized care center.
8 . The system of claim 1 , wherein the system further comprises a router configured to forward medical data associated with the patient to a remote computing system.
9 . The system of claim 1 , wherein each trained machine learning algorithm of the set of multiple trained machine learning algorithms comprises a convolutional neural network.
10 . A method for computer-aided triage of a patient, the patient located at a first point of care, the method comprising, during a triage process of a patient:
at a computing system, receiving a set of images associated with the patient;
at the computing system, detecting a potential condition based on the set of images by concurrently processing the set of images with each of a set of multiple trained machine learning algorithms, wherein each of the trained machine learning algorithms is configured to detect a different condition, the set of trained machine learning algorithm comprising:
a first trained machine learning algorithm configured to detect an aneurysm based on a first feature; and
a second trained machine learning algorithm configured to detect an occlusion based on a second feature;
upon detecting the potential condition from the set of images, automatically notifying a specialist associated with a second point of care of the potential condition; and
in response to notifying the specialist, receiving an input from the specialist, the input comprising information related to a transfer of the patient to the second point of care.
11 . The method of claim 10 , wherein the different conditions that each of the trained machine learning algorithms is configured to detect comprises at least one of: a cardiovascular condition, a neurovascular condition, or a vascular condition.
12 . The method of claim 10 , further comprising initiating a transfer of the patient from the first point of care to the second point of care after receipt of the input.
13 . The method of claim 10 , further comprising establishing communication between a set of digital applications operating on a first device of the specialist and on a second device of a user located at the first point of care.
14 . The method of claim 10 , wherein the notification is populated in less than 8 minutes from a time at which the set of images is received at the computing system.
15 . The method of claim 10 , wherein the potential condition comprises a potential cerebral artery occlusion.
16 . The method of claim 10 , further comprising:
determining a set of severity scores for a plurality of patients; and
reorganizing a queue of the plurality of patients based on the set of severity scores.
17 . The method of claim 10 , wherein automatically notifying the specialist associated with the second point of care of the potential condition comprises displaying a version of the set of images on a device associated with the specialist.
18 . The method of claim 17 , wherein the device is a mobile user device.
19 . The method of claim 10 , wherein:
in a standard radiology workflow, a radiologist analyzes the set of images associated with the patient and notifies the specialist based on the analysis, wherein the standard radiology workflow takes a first amount of time; and
upon detecting the potential condition from the set of images, automatically notifying the specialist of the potential condition occurs notified in a second amount of time shorter than the first amount of time.