2 results match your criteria: "1 Metro Health-University of Michigan Health[Affiliation]"
Overfitting and Use of Mismatched Cohorts in Deep Learning Models: Preventable Design Limitations.
Am J Respir Crit Care Med
August 2018
1 Metro Health-University of Michigan Health Wyoming, Michigan and.
Purpose: To present the chronic total occlusion (CTO) crossing approach based on plaque cap morphology (CTOP) classification system and assess its ability to predict successful lesion crossing.
Methods: A retrospective analysis was conducted of imaging and procedure data from 114 consecutive symptomatic patients (mean age 69±11 years; 84 men) with claudication (Rutherford category 3) or critical limb ischemia (Rutherford category 4-6) who underwent endovascular interventions for 142 CTOs. CTO cap morphology was determined from a review of angiography and duplex ultrasonography and classified into 4 types (I, II, III, or IV) based on the concave or convex shape of the proximal and distal caps.