Image data pre-processing

• US 10,163,028 B2
• Filed: 01/23/2017
• Issued: 12/25/2018
• Est. Priority Date: 01/25/2016
• Status: Active Grant

• Current Assignee: Koninklijke Philips N.V.
• Original Assignee: Koninklijke Philips N.V.
• Inventors: Rubens, Eran, Fadida-Specktor, Bella, Zino, Eliahu, Melamed, Menachem, Meir, Eran, Weinberg, Netanel, Kaplinsky, Maria
• Primary Examiner(s): Tran, Phuoc
• Application Number:
  US15/412,109
• Publication Number:
  US 20170213101A1
• Time in Patent Office: 701 Days
• CPC Class Codes: G06F 18/217 Validation; Performance evaluation; Active pattern learning techniques
• CPC Class Codes: G06T 11/206 Drawing of charts or graphs
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 5/00 Image enhancement or restoration
• CPC Class Codes: G06T 5/60 using machine learning, e.g. neural networks
• CPC Class Codes: G16H 30/20 for handling medical images, e.g. DICOM, HL7 or PACS

Abstract

A computer-implemented method (40) of reducing processing time of an application for visualizing image data, the application being one of a plurality of applications selectable by a user and each of the plurality of applications comprising a pre-processing algorithm for pre-processing the image data. The method comprises predicting which one of the pre-processing algorithms is to be performed in response to a selection of an application by the user by:

• • extracting a feature vector from the image data, metadata, and/or additional data associated with the image data,
  • supplying the feature vector as input to a machine learned model, and
  • receiving an algorithm identifier as output from the machine learned model, the algorithm identifier identifying the pre-processing algorithm,
  • the method further comprising:
  • using the algorithm identifier to select (42) the pre-processing algorithm, thereby obtaining a selected pre-processing algorithm, and
  • pre-processing (43) the image data using the selected pre-processing algorithm.

Claims

1. A computer-implemented method for visualizing medical image data, comprising:

extracting a feature vector from the medical image data, metadata associated with the medical image data, patient data associated with the medical image data, or a combination thereof;

supplying the feature vector as input to a machine learned model configured to predict which one of a plurality of applications is to be selected by a user, wherein each of the plurality of applications comprises a pre-processing algorithm for pre-processing the medical image data and a corresponding advanced visualization process, the pre-processing algorithm reduces subsequent processing time for the corresponding advanced visualization process;

receiving an algorithm identifier as output from the machine learned model, the algorithm identifier identifying a selected pre-processing algorithm of a predicted selection of the user from the plurality of applications, wherein the predicted selection is generated by the machine learned model before obtaining a user selection; and

applying the selected pre-processing algorithm to the image data before applying the corresponding advanced visualization process of the predicted selection when the predicted selection corresponds to the user selection.

2. The method according to claim 1,

wherein the image data is DICOM image data,

wherein the metadata is a DICOM header of the DICOM image data, and

wherein the extracting step comprises extracting DICOM text tags from the DICOM header.

3. The method according to claim 2,

wherein the DICOM text tags are free-text DICOM tags, and

wherein the extracting step further comprises processing the free-text DICOM tags to obtain a set of processed DICOM text tags.

4. The method according to claim 3, wherein the extracting step further comprises using a random forest machine learning technique to generate the feature vector from the set of processed DICOM text tags.

5. The method according to claim 1, wherein the machine learned model is produced by a machine learning system.

6. The method according to claim 5,

wherein the machine learning system comprises a machine learning engine configured to build the machine learned model using training data comprising a training input vector and a corresponding training output vector, wherein the training input vector comprises a feature vector, and the training output vector comprises at least one algorithm identifier, wherein the feature vector is extracted from stored data in the machine learning system, the stored data comprising stored medical image data, metadata associated with stored medical image data, patient data associated with the stored image data, or a combination thereof.

7. The method according to claim 6,

wherein the stored data is associated with medical image data from a single picture archiving and communication system.

8. The method according to claim 1, wherein at least one of the pre-processing algorithms is configured to leave the image data unaltered.

9. A computer readable medium comprising transitory or non-transitory data representing instructions arranged to cause a processor to carry out the method according to claim 1.

10. The method according to claim 1, wherein at least one of the plurality of advanced visualization processes generates at least one of measurements and 3D-views of body parts from the image data.

11. The method according to claim 1, wherein the medical image data are generated by a medical scanner selected from the group consisting of a CT (computer tomography) scanner, an MR (magnetic resonance) scanner, and an X-ray scanner.

12. A device configured to visualize medical image data, comprising:

an input interface configured to access the medical image data, metadata associated with the medical image data, patient data associated with the medical image data, or a combination thereof;

a memory comprising application data representing the plurality of applications and instruction data representing a set of instructions, wherein each of the plurality of applications comprises a pre-processing algorithm for pre-processing the medical image data and a corresponding advanced visualization process, the pre-processing algorithm reduces subsequent processing time for the corresponding advanced visualization process; and

a processor configured to communicate with the input interface and the memory and to execute the set of instructions, wherein the set of instructions, when executed by the processor, cause the processor to;

extracting a feature vector from the medical image data, the metadata, the patient data or the combination thereof,

supplying the feature vector as input to a machine learned model configured to predict which one of the plurality of applications is to be selected by a user, wherein the machine learned model is stored as model data in the memory, and

receiving an algorithm identifier as output from the machine learned model, the algorithm identifier identifying a selected pre-processing algorithm of a predicted selection of the user from the plurality of applications, wherein the predicted selection is generated by the processor executing the machine learning model before obtaining a user selection; and

applying the selected pre-process algorithm to the image data before applying the corresponding advanced visualization process to the predicted selection when the predicted selection corresponds to the user selection.

13. The device according to claim 12 being a workstation or imaging apparatus.