US 12,082,940 B1
Determination of structural characteristics of an object
James C. Earthman, Irvine, CA (US); Cherilyn G. Sheets, Newport Beach, CA (US); and Dennis A. Quan, Jr., Cary, NC (US)
Assigned to Perimetrics, Inc., Redmond, WA (US)
Filed by Perimetrics, LLC
Filed on Dec. 30, 2019, as Appl. No. 16/731,020.
Claims priority of provisional application 62/786,494, filed on Dec. 30, 2018.
Int. Cl. A61B 5/00 (2006.01); A61B 9/00 (2006.01); A61C 9/00 (2006.01); G06F 11/22 (2006.01); G06F 30/10 (2020.01); G06F 30/20 (2020.01); G16H 50/50 (2018.01)
CPC A61B 5/4547 (2013.01) [A61B 9/00 (2013.01); A61C 9/0046 (2013.01); A61C 9/0053 (2013.01); G06F 11/2247 (2013.01); G06F 11/2263 (2013.01); G06F 11/2268 (2013.01); G06F 11/2289 (2013.01); G06F 30/10 (2020.01); G06F 30/20 (2020.01); G16H 50/50 (2018.01)] 20 Claims
OG exemplary drawing
 
1. A system for predicting at least one structural characteristic of an object using a predictive model, comprising:
a compilation of test results from a multitude of objects, each of said test results being generated by a device having a housing with an open end and a longitudinal axis comprising:
an energy application tool mounted inside the housing for movement, said energy application tool having a resting and an active configuration;
a sleeve protruding from the open end of the housing for a distance and having an object contacting portion at its open end adapted for resting said device on at least a portion of said object; and
a drive mechanism supported inside the housing adapted for activating said energy application tool to impact said object when said object contacting portion of said sleeve is resting on at least a portion of said object and for measuring a response after impact and generating a response versus time curve, said response versus time curve having been annotated and validated;
a computer coupled to said device adapted for capturing any of said response, said response versus time curve and said annotation and validation to generate said predictive model having one or more characteristic ranges; and
a transformation function adapted for identifying structural characteristics comprising a combination of loss coefficient and normal fit error from said predictive model;
wherein said predictive model is adapted for predicting said at least one structural characteristic of a newly tested object using said device with said transformation function when both said loss coefficient and said normal fit error in said combination fall into said one or more characteristic ranges without the need for further annotation and validation.