US 9,326,698 B2
Method for automatic, unsupervised classification of high-frequency oscillations in physiological recordings
Justin Blanco, Annapolis, MD (US); and Brian Litt, Bala Cynwyd, PA (US)
Assigned to The Trustees of the University of Pennsylvania, Philadelphia, PA (US)
Filed by Justin Blanco, Annapolis, MD (US); and Brian Litt, Bala Cynwyd, PA (US)
Filed on Feb. 17, 2012, as Appl. No. 13/385,413.
Claims priority of provisional application 61/444,560, filed on Feb. 18, 2011.
Prior Publication US 2012/0245481 A1, Sep. 27, 2012
Int. Cl. A61B 5/04 (2006.01); A61B 5/0476 (2006.01); A61B 5/048 (2006.01); A61B 5/00 (2006.01)
CPC A61B 5/048 (2013.01) [A61B 5/4094 (2013.01); A61B 5/7264 (2013.01); A61B 5/0476 (2013.01); A61B 5/7203 (2013.01); A61B 5/7267 (2013.01); A61B 5/7275 (2013.01)] 28 Claims
OG exemplary drawing
 
1. A method for automatically detecting, classifying, and processing oscillatory signals for use as a biomarker for a physiologic physiological event in a patient, the method comprising:
detecting and classifying oscillatory signals representative of discrete physiologic events in a patient's body using a high sensitivity, low specificity detector;
processing, using a processor, the detected signals by:
testing the detected signals in the context of surrounding background activity to identify anomalous discrete physiologic events that are sufficiently different from the surrounding background activity; and
automatically clustering said anomalous discrete physiologic events into clusters of anomalous physiologic events having correlative morphological, time, or location characteristics;
determining at least one cluster of anomalous physiologic events as a biomarker indicative of at least one region of the patient's body that is associated with a medical condition; and
providing treatment, using a treatment device responsive to said biomarker, of said medical condition to the at least one region of the patient's body identified by said biomarker.