Kwok Hing Lee

6725231, Apr 20, 2004

Mar 27, 2001

09/818716

Jingkun Hu - Nyack NY
Kwok Pun Lee - Yorktown Heights NY

Koninklijke Philips Electronics N.V. - Eindhoven

G06F 1730

707102, 707 1

A DICOM-to-XML conversion system is provided that converts the DICOM SR standard into a set of XML DTDs and Schemas. By providing a mapping between the DICOM SR standard and XML DTDs and Schemas, DICOM specific XML-based applications can be developed, via a larger field of XML-fluent application developers. Additionally, by providing standard XML DTDs and Schemas for containing DICOM data, other commonly available non-DICOM-related applications, such as accounting and mailing programs, can be structured to use information as required from DICOM reports that are converted to conform to these defined XML DTDs and Schemas. In a preferred embodiment, a two-phase conversion is employed. The DICOM SR specification is parsed and converted directly into a set of ârawâ XML documents. Thereafter, the ârawâ XML documents are transformed into the corresponding XML DTDs and Schemas, via an XSLT processor. Changes to the desired XML DTDs and Schemas, as standards develop, can thus be effected via changes in the corresponding XSLT stylesheets, without modification to the DICOM-to-raw-XML process.

6950985, Sep 27, 2005

Dec 27, 2001

10/029827

Kwok Pun Lee - Flushing NY, US

Koninklijke Philips Electronics, N.V. - Eindhoven

G08F017/00

715513, 7155011

The invention relates to a method of providing DICOM SR constraints within an XML document. An XML document is created containing DICOM SR constraints using declarative language. The document can then be accessed and displayed if desired.

7231652, Jun 12, 2007

Mar 28, 2001

09/819286

Srinivas Gutta - Buchanan NY, US
Kwok Pun Lee - Flushing NY, US
J. David Schaffer - Wappingers Falls NY, US

Koninklijke Philips N.V. - Eindhoven

H04H 9/00
H04N 7/16
H04N 5/445
G06F 13/00
G06F 17/30
G06F 15/18
G06Q 30/00
G06Q 7/00

725 46, 725 9, 725 10, 725 13, 725 53, 706 15, 705 10, 705 14

Artificial intelligence applications require use of training sets containing positive and negative examples. Negative examples are chosen using distributions of positive examples with respect to a dominant feature in feature space. Negative examples should share or approximately share, with the positive examples, values of a dominant feature in feature space. This type of training set is illustrated with respect to content recommenders, especially recommenders for television shows.

7373600, May 13, 2008

Mar 27, 2001

09/818715

Kwok Pun Lee - Flushing NY, US
Jingkun Hu - Nyack NY, US

Koninklijke Philips Electronics N.V. - Eindhoven

G06F 15/00

715523, 715513, 715530

A conversion system converts DICOM SR information from a DICOM-formatted file into an XML representation. By providing a mapping between DICOM SR and XML, the DICOM SR content material can be more easily processed by application programs that are DICOM-specific, such as medical analysis programs, as well as by application programs that are not DICOM-specific, such as routine clerical or data-management programs. In a preferred embodiment, a two-phase conversion is employed. The DICOM information is parsed and Fig converted directly into a “raw” XML data set. Thereafter, the “raw” XML is transformed into a proper XML output form, via an XSLT processor. Changes to the desired XML output form can thus be effected via changes in the corresponding XSLT stylesheets.

7840062, Nov 23, 2010

Nov 21, 2005

11/719634

Lilla Boroczky - Mount Kisco NY, US
Kwok Pun Lee - Flushing NY, US
Luyin Zhao - White Plains NY, US

Koninklijke Philips Electronics, N.V. - Eindhoven

G06K 9/00
G01N 23/04

382159, 382224, 378 62

A method for computer aided detection (CAD) and classification of regions of interest detected within HRCT medical image data includes post-CAD machine learning techniques applied to maximize specificity and sensitivity of identification of a region/volume as being a nodule or non-nodule. The regions are identified by a CAD process, and automatically segmented. A feature pool is identified and extracted from each segmented region, and processed by genetic algorithm to identify an optimal feature subset, which subset is used to train the support vector machine to classify candidate region/volumes found within non-training data.

8265355, Sep 11, 2012

Nov 18, 2005

11/719214

Luyin Zhao - White Plains NY, US
Kwok Pun Lee - Flushing NY, US

Koninklijke Philips Electronics N.V. - Eindhoven

G06K 9/00

382128, 700182

A method for segmenting regions within a medical image includes evaluating a set of candidate segmentations generated from an initial segmentation. Based on distance calculations for each candidate using derivative segmentations, the best candidate is recommended to clinician if it is better than the initial segmentation. This recommender realizes a most stable segmentation that will benefit follow-up computer aided diagnosis (i. e. classifying lesion to benign/malignant).

20030105846, Jun 5, 2003

Nov 30, 2001

10/015708

Luyin Zhao - Briarcliff Manor NY, US
Jin Lu - Croton-on-Hudson NY, US
Kwok Lee - Flushing NY, US

Koninklijke Philips Electronics N.V.

G06F015/173
G06F015/16
G06F015/177

709/221000, 709/219000, 709/225000

A method for obtaining service information over the Internet. The method including: at least one service provider registering a service and corresponding service status with a server and storing the same in a database; a user requesting a service from the server; searching the database for the requested service; and informing the user of the results of the search. If the requested service is found in the database, the user is informed of the corresponding service status of the requested service. Similarly, if the requested service is not found in the database, the user is informed of the same. If the requested service is not found in the database, or the service status for the requested service is unavailable, the request is stored and the user is informed when the service becomes registered or the service status is changed to available.

20090175514, Jul 9, 2009

Nov 21, 2005

11/719672

Luyin Zhao - White Plains NY, US
Kwok Pun Lee - Flushing NY, US
Lilla Boroczky - Mount Kisco NY, US

KONINKLIJKE PHILIPS ELECTRONICS, N.V. - EINDHOVEN

G06K 9/00
G06K 9/62

382128, 382159

A method for computer aided detection (CAD) and classification of regions of interest detected within HRCT medical image data. The method includes post-CAD machine learning techniques applied to maximize specificity and sensitivity of identification of a region/volume as being a nodule or non-nodule. The regions are identified by a CAD process, and automatically segmented. A feature pool is identified and extracted from each segmented region, and processed by genetic algorithm to identify an optimal feature subset, wherein a data stratification method is used to balance the number of cases in different classes. The subset determined by GA is used to train the support vector machine to classify candidate region/volumes found within non-training data.