Data Transmission in Remote Computer Assisted Detection

• US 20110243407A1
• Filed: 04/06/2011
• Published: 10/06/2011
• Est. Priority Date: 04/06/2010
• Status: Active Grant

• Current Assignee: Siemens Healthineers AG (Siemens AG)
• Original Assignee: Siemens Corp. (Siemens AG)
• Inventors: Ralovich, Kristof, Sofka, Michal, Zhou, Shaohua Kevin, Paladini, Gianluca, Comaniciu, Dorin, Zhang, Jingdan
• Granted Patent: US 8,811,697 B2
• Time in Patent Office: Days
• US Class Current: 382/128
• CPC Class Codes: G06F 18/214 Generating training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06T 2200/16 involving adaptation to the client's capabilities
• CPC Class Codes: G06T 2207/20016 Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06V 10/774 Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G16H 30/20 for handling medical images, e.g. DICOM, HL7 or PACS
• CPC Class Codes: G16H 40/67 for remote operation

Abstract

For cloud-based computer assisted detection, hierarchal detection is used, allowing detection on data at progressively greater resolutions. Detected locations at coarser resolutions are used to limit the data transmitted at greater resolutions. Data is only transmitted for neighborhoods around the previously detected locations. Subsequent detection using higher resolution data refines the locations, but only for regions associated with previous detection. By limiting the number and/or size of regions provided at greater resolutions based on the previous detection, the progressive transmission avoids transmission of some data. Additionally, or alternatively, lossy compression may be used without or with minimal reduction in detection sensitivity.

Claims

1. In a non-transitory computer readable storage medium having stored therein data representing instructions executable by a programmed processor for data transmission in computer assisted detection, the storage medium comprising instructions for:

receiving, at a computer assisted detection processor and from the remote storage location, first data, the first data comprising down sampled data of data representing a region;

detecting, by the computer assisted detection processor, a plurality of first locations represented by the first data;

requesting and receiving, at the computer assisted detection processor, second data, the second data comprising additional data representing second neighborhoods around the first locations and not representing portions of the region outside of the second neighborhoods;

detecting, by the computer assisted detection processor, a plurality of second locations represented by the second data; and

outputting positions to a client device different than the computer assisted detection processor and the remote storage location, the positions being a function of the second locations.

2. The non-transitory computer readable storage medium of claim 1 wherein receiving the first data comprises receiving the first data in a lossy compression format, and further comprising:

decompressing the first data;

wherein detecting the first locations comprises detecting from the decompressed first data.

3. The non-transitory computer readable storage medium of claim 1 wherein the second locations include one or more of the first locations, the detecting of the second locations ruling out at least one of the first locations.

4. The non-transitory computer readable storage medium of claim 1 wherein the first data and second data are compressed data pursuant to the JPEG 2000 Interactive Protocol.

5. The non-transitory computer readable storage medium of claim 1 wherein the first data comprises compressed data at a first level and the second data comprises compressed data at a second level less than the first level.

6. The non-transitory computer readable storage medium of claim 1 wherein the second data comprises high frequency wavelet components not in the first data.

7. The non-transitory computer readable storage medium of claim 1 further comprising:

requesting and receiving, at the computer assisted detection processor, third data, the third data comprising additional data representing third neighborhoods around the second locations and not representing portions of the region outside of the third neighborhoods; and

detecting, by the computer assisted detection processor, a plurality of third locations represented by the third data;

wherein outputting comprises outputting the positions as the third locations.

8. The non-transitory computer readable storage medium of claim 1 wherein outputting comprises outputting the positions to the client device, the client device comprising a handheld device;

further comprising;

requesting, by the client device, the positions for a patient, wherein the remote storage location comprises a picture archiving server; and

displaying the positions on an image rendered remotely from the client device, the image being of the patient and rendered from the data representing the region.

9. The non-transitory computer readable storage medium of claim 1 wherein detecting the first locations comprises detecting with a first machine-trained detector and wherein detecting the second locations comprises detecting with a second machine-trained detector, wherein the first locations are a collection of locations associated with a threshold probability output by the first machine-trained detector;

further comprising;

bootstrapping the first locations to the second machine-trained detector, the second machine-trained classifier detecting from a union of the second neighborhoods.

10. The non-transitory computer readable storage medium of claim 1 wherein detecting the second locations comprises detecting from a combination of the first and second data for the second neighborhoods.

11. A method for computer assisted detection with data from a remote storage location, the method comprising:

receiving image data representing a patient at progressively greater resolutions; and

detecting, with a processor, candidate locations hierarchally from the image data at the progressively greater resolutions where the image data for each of the resolutions is for the candidate locations detected from a lesser one of the resolutions;

wherein the received image data is free of data representing regions outside of neighborhoods around the candidate locations for any of the progressively greater resolutions.

12. The method of claim 11 wherein receiving comprises receiving first image data at a first resolution level, wherein detecting the candidate locations comprises detecting first candidate locations from the image data at the first resolution level;

further comprising;

feeding back the first candidate locations;

wherein receiving comprises receiving second image data corresponding to the first candidate locations, and wherein detecting the candidate locations comprises detecting second candidate locations from the second image data.

13. The method of claim 11 wherein receiving comprises receiving the image data at the progressively greater resolutions as lossy compressed data, and wherein detecting comprises detecting from uncompressed data derived from the lossy compressed data.

14. The method of claim 11 wherein the progressively greater resolutions comprise at least first and second resolution levels, the first resolution level being more coarse than the second resolution level, and wherein detecting hierarchally comprises detecting from the image data at the first resolution level, using the candidate locations from the detecting the at first resolution level to select the image data at the second resolution level, and detecting the candidate locations at the second resolution level from the selected image data.

15. The method of claim 11 wherein receiving comprises receiving the image data pursuant to the JPEG 2000 Interactive Protocol, the image data at the progressively greater resolutions comprising high frequency wavelet components not in previously received image data at lesser resolutions.

16. The method of claim 11 further comprising:

outputting, to a client handheld device, the candidate locations detected from a greatest resolution of the progressively greater resolutions; and

wherein receiving comprises receiving by a computer assisted detection server from a picture archiving server, the computer assisted detection server comprising the processor.

17. The method of claim 11 wherein detecting comprises detecting with hierarchy of machine-trained detectors.

18. A system for computer assisted detection with medical scan data from a remote storage location, the system comprising:

a database configured to store the medical scan data and configured to output the medical scan data in a lossy compression format, the medical scan data representing a region of a patient; and

a processor configured to receive, through a network, the medical scan data from the database in the lossy compression format, the processor configured to decompress the medical scan data and detect a location in the region from the decompressed medical scan data.

19. The system of claim 18 wherein the processor is configured to perform the receiving the medical scan data at progressively greater resolutions and is configured to perform the detection hierarchally from the medical scan data at the progressively greater resolutions;

wherein the medical scan data received form the progressively greater resolutions is free of data representing regions outside of neighborhoods around detected candidate locations, the location being one of the detected candidate locations.

20. The system of claim 18 wherein the lossy compression format comprises a JPEG 2000 format.