Improving efficiency and operator proficiency during percutaneous coronary interventions utilizing a standardized optical coherence tomography workflow.

OBJECTIVES: This study was designed to compare efficiency and quality metrics between percutaneous coronary intervention (PCI) procedures using optical coherence tomography (OCT) guided by a variable workflow versus a standardized workflow in a real-world population. BACKGROUND: The LightLab (LL) Initiative was designed to evaluate the impact of a standardized OCT workflow during PCI to address barriers to adoption. METHODS: The LL Initiative was a multicenter, prospective, observational study. PCI efficiency data were collected from 1/21/19 to 1/8/21 from 45 physicians at 17 US centers. OCT-guided PCIs were compared between baseline phase (variable workflow; N = 383) and the LL workflow utilization phase (N = 447). The LL workflow uses OCT to assess lesion Morphology, Length and Diameter, and then optimize outcomes by correcting for Medial dissection, stent mal-Apposition, and under-eXpansion (MLD MAX). Matching based on propensity scores was used to control for differences between PCIs. RESULTS: After propensity matching, 291 paired procedures were included. Integration of the LL versus variable workflow resulted in no difference in procedure time (51 min vs. 51 min, p = 0.93). There was a reduction in radiation exposure (1124 mGy vs. 1493 mGy, p < 0.0001) and contrast volume (160 cc vs. 172 cc, p < 0.001). The LL workflow decreased the proportion of underexpanded lesions (34% vs. 54%, p < 0.0001) and improved minimum stent expansion (85% vs. 79%, p < 0.0001). Number of noncompliant balloons used was reduced with the LL workflow. (2.0 vs. 1.7, p < 0.01). CONCLUSIONS: These data suggest that standardizing imaging with the LL workflow may overcome barriers to imaging and improve PCI outcomes without prolonging procedures.

Self-supervised ARTMAP.

Computational models of learning typically train on labeled input patterns (supervised learning), unlabeled input patterns (unsupervised learning), or a combination of the two (semi-supervised learning). In each case input patterns have a fixed number of features throughout training and testing. Human and machine learning contexts present additional opportunities for expanding incomplete knowledge from formal training, via self-directed learning that incorporates features not previously experienced. This article defines a new self-supervised learning paradigm to address these richer learning contexts, introducing a neural network called self-supervised ARTMAP. Self-supervised learning integrates knowledge from a teacher (labeled patterns with some features), knowledge from the environment (unlabeled patterns with more features), and knowledge from internal model activation (self-labeled patterns). Self-supervised ARTMAP learns about novel features from unlabeled patterns without destroying partial knowledge previously acquired from labeled patterns. A category selection function bases system predictions on known features, and distributed network activation scales unlabeled learning to prediction confidence. Slow distributed learning on unlabeled patterns focuses on novel features and confident predictions, defining classification boundaries that were ambiguous in the labeled patterns. Self-supervised ARTMAP improves test accuracy on illustrative low-dimensional problems and on high-dimensional benchmarks. Model code and benchmark data are available from: http://techlab.eu.edu/SSART/.