Author + information
- Tom Stuckey1,
- Narendra Singh2,
- Robi Goswami3,
- Jeremiah Depta4,
- Roger Gammon5,
- John Steuter6,
- Michael Roberts7,
- Sunny Gupta8 and
- William Sanders JR.9
- 1Lebauer Research, Greensboro, North Carolina, United States
- 2Atlanta Heart Specialists, Atlanta, Georgia, United States
- 3Piedmont Heart Institute, Atlanta, Georgia, United States
- 4Rochester General Hospital, Rochester, Georgia, United States
- 5Austin Heart, Austin, Texas, United States
- 6Bryan Heart Institute, Lincoln, Nebraska, United States
- 7Lexington Cardiology, West Columbia, South Carolina, United States
- 8Analytics 4 Life, Toronto, Ontario, Canada
- 9Analytics 4 Life, Morrisville, North Carolina, United States
Combined with machine learning, cardiac phase space tomography analysis (cPSTA) evaluates thoracic physiological signals, without the use of radiation, exercise, or pharmacological stress, to identify algorithmic signatures that are associated with the presence of flow-limiting coronary artery disease (CAD). The ongoing Coronary Artery Disease Learning & Algorithm Development (CADLAD) study is determining the diagnostic performance of cPSTA in assessing CAD among patients with chest pain, referred for coronary angiography (ANGIO). This analysis focuses on obese and elderly patients, since conventional CAD detection pathways may be less accurate in these patients.
This prospective, multicenter, non-significant risk study was designed to develop and test machine-learned algorithms to assess the presence of CAD (defined as one or more ≥ 70% stenosis, or fractional flow reserve < 0.80). The overall interim results of CADLAD have been submitted elsewhere for presentation. This analysis focuses on CADLAD elderly (>65 years of age) and obese (body mass index or BMI > 30). cPSTA signals were collected at rest, prior to ANGIO. Features (mathematical and tomographic) were extracted from the signals, used for machine learning, and blindly, prospectively tested in a validation cohort.
ANGIO results were paired with cPSTA data from 513 subjects to generate a machine-learned algorithm to assess for significant CAD. A separate validation cohort of 94 subjects was used to prospectively test the accuracy. This analysis focused on subjects: > 65 vs. < 65 years of age and those with a BMI of > 30 vs. <30 (Table).
|Age < 65||Age > = 65||BMI < 30||BMI > = 30|
|Sensitivity (95% CI)||100% (100%,100%)||86% (56%,100%)||92% (50%, 100%)||83% (46%, 100%)|
|Specificity (95% CI)||63% (49%, 75%)||67% (40%, 88%)||67% (44%, 84%)||67% (51%, 79%)|
|AUC (95% CI) (Area Under the receiver-operator characteristic Curve)||0.79 (0.66, 0.88)||0.72 (0.50, 0.88)||0.80 (0.62, 0.92)||0.78 (0.64, 0.88)|
|NPV (95% CI) (Negative Predictive Value)||100% (100%,100%)||83% (50%,100%)||94% (68%,100%)||94% (79%,100%)|
|TP (True Positive)||11||12||12||10|
|TN (True Negative)||34||10||16||30|
|FP (False Positive)||20||5||8||15|
|FN (False Negative)||0||2||1||2|
These initial data, while of limited power, suggest that resting cPSTA imaging performs well overall, and in both the elderly and obese subgroups. The study is ongoing and is designed to test the utility of cPSTA in a large population and other important subgroups.
IMAGING: Imaging: Non-Invasive