Virtualized computing platform for inferencing, advanced processing, and machine learning applications
In various examples, a virtualized computing platform for advanced computing operations—including image reconstruction, segmentation, processing, analysis, visualization, and deep learning—may be provided. The platform may allow for inference pipeline customization by selecting, organizing, and adapting constructs of task containers for local, on-premises implementation. Within the task containers, machine learning models generated off-premises may be leveraged and updated for location specific implementation to perform image processing operations. As a result, and using the virtualized computing platform, facilities such as hospitals and clinics may more seamlessly train, deploy, and integrate machine learning models within a production environment for providing informative and actionable medical information to practitioners.
1 . A web-based service, comprising:
a user interface to allow one or more users of the web-based service to select one or more neural network data processing programs from a container registry and one or more selected neural networks from a model registry, wherein the model registry comprises a plurality of neural networks trained on data in different locations to perform different inferencing tasks; and
cause the web-based service to use the selected one or more neural networks to use information generated by the one or more selected neural network data processing programs to implement a sequence of processing operations, wherein the sequence of processing operations include a combination of:
inferencing using the one or more selected neural networks,
generating a visualization based, at least in part, on data generated from the inferencing by the one or more selected neural networks, and
transmitting display data corresponding to the visualization to a computing system for display by the computing system, wherein the computing system comprises an augmented reality (AR) system or a virtual reality (VR) system.
2 . The web-based service of claim 1 , wherein the web-based service receives data from an on-premise computing system, prepares the data for the one or more selected neural network data processing programs, applies the data to each of the one or more selected neural network data processing programs, and prepares the data generated from the one or more selected neural network data processing programs for use by the one or more selected neural networks at the on-premise computing system.
3 . The web-based service of claim 2 , wherein the data received from the on-premise computing system is to be processed before the data is to be used by the one or more selected neural network data processing programs to generate output to train the one or more selected neural networks.
4 . The web-based service of claim 3 , wherein the on-premise computing system is a medical imaging device.
5 . The web-based service of claim 4 , wherein the data received from the medical imaging device comprises raw medical image data.
6 . The web-based service of claim 5 , further comprising:
processing the raw medical image data, using the one or more selected neural network data processing programs, to generate an image.
7 . The web-based service of claim 1 , further comprising:
implementing a deployment pipeline within a computing system to include the one or more selected neural networks and the one or more selected neural network data processing programs.
8 . The web-based service of claim 7 , wherein at least one of the one or more selected neural network data processing programs is hosted remotely from the computing system.
9 . The web-based service of claim 1 , wherein the one or more selected neural networks includes performing operations comprising using at least one of: a computer vision algorithm, an object detection algorithm, an image reconstruction algorithm, and a gene sequencing algorithm.
10 . The web-based service of claim 1 , further comprising:
performing one or more operations on data generated by the one or more selected neural networks before transmission to the one or more users.
11 . The web-based service of claim 1 , wherein data to be used by the one or more selected neural network data processing programs include at least one of: digital imaging and communications in medicine (DICOM) data, remote procedure call (RPC) data, data substantially compliant with a REST interface, data substantially compliant with a file-based interface, raw data, and radiological information system (RIS) data.
12 . A system, comprising one or more processors comprising:
circuitry to implement a sequence of processing operations using information generated by one or more selected neural networks performing one or more selected neural network data processing programs, and to execute a web-based service, the web-based service comprising:
a user interface to allow one or more users of the web-based service to select the one or more selected neural network data processing programs from a container registry and the one or more selected neural networks from a model registry, wherein the model registry comprises a plurality of neural networks trained on data in different locations to perform different inferencing tasks and wherein the sequence of processing operations include a combination of:
inferencing using the one or more selected neural networks,
generating a visualization based, at least in part, on data generated from the inferencing by the one or more selected neural networks, and transmitting display data corresponding to the visualization to a computing system for display by the computing system, wherein the computing system comprises an augmented reality (AR) system or a virtual reality (VR) system.
13 . The system of claim 12 , wherein the one or more processors are further to implement a processing pipeline that is customized by:
receiving a selection of at least the one or more selected neural networks from a pool of neural networks for processing data; and
configuring the processing pipeline to include at least one of the one or more selected neural networks and at least one of the one or more selected neural network data processing programs.
14 . The system of claim 13 , wherein the one or more selected neural network data processing programs includes a pre-processing operation for preparing the data for the one or more selected neural networks.
15 . The system of claim 13 , wherein the one or more processors are further to:
receive a request from a computing device to perform one or more operations on the data; and
use the processing pipeline to apply at least one of the one or more selected neural network data processing programs to perform the one or more operations on the data in the request.
16 . The system of claim 12 , wherein the one or more selected neural networks includes at least one of: a computer vision algorithm, an object detection algorithm, an image reconstruction algorithm, and a gene sequencing algorithm.
17 . The system of claim 16 , wherein the one or more selected neural networks is used to perform one or more tasks including an object detection task, a feature detection task, a segmentation task, a reconstruction task, a calibration task, or an image enhancement task.
18 . The system of claim 12 , wherein the one or more processors are further to use a visualization service to display the data generated from the inferencing by the one or more selected neural networks.
19 . One or more processors, comprising:
circuitry to implement a sequence of processing operations using information generated by one or more selected neural networks performing one or more selected neural network data processing programs, and to execute a web-based service, the web-based service comprising:
a user interface to allow one or more users of the web-based service to select the one or more selected neural network data processing programs from a container registry and the one or more selected neural networks from a model registry, wherein the model registry comprises a plurality of neural networks trained on data in different locations to perform different inferencing tasks and wherein the sequence of processing operations include a combination of:
inferencing using the one or more selected neural networks,
generating a visualization based, at least in part, on data generated from the inferencing by the one or more selected neural networks, and
transmitting display data corresponding to the visualization to a computing system for display by the computing system, wherein the computing system comprises an augmented reality (AR) system or a virtual reality (VR) system.
20 . The one or more processors of claim 19 , wherein the circuitry is further to enable the one or more users to implement a pipeline by selecting the one or more selected neural networks from the model registry and selecting the one or more selected neural network data processing programs from the container registry.
21 . The one or more processors of claim 20 , wherein the model registry includes one or more machine learning models trained to perform at least one processing task with respect to digital imaging and communications in medicine (DICOM) data, radiology information system (RIS) data, clinical information system (CIS) data, remote procedure call (RPC) data, data substantially compliant with a representation state transfer (REST) interface, data substantially compliant with a file-based interface, or raw data.
22 . The one or more processors of claim 19 , wherein the circuitry is further to:
receive a request to perform one or more operations on data;
identify a format of the data before data is used to train the one or more selected neural networks; and
use at least one of the one or more selected neural network data processing programs to perform the one or more operations on the data, based on the identified format, before data is used to train the one or more selected neural networks.
23 . The one or more processors of claim 22 , wherein the one or more operations on the data comprise converting the data to another format suitable for processing by the one or more selected neural networks.
24 . The one or more processors of claim 19 , wherein the circuitry is further to:
identify a format of the data generated by the one or more selected neural networks; and
use at least one of the one or more selected neural network data processing programs to process the data generated by the one or more selected neural networks, based on the identified format, before transmitting the data to a display device.