{
  "abstract": "Introduction Intracoronary optical coherence tomography (OCT) can identify changes following drugs or devices and high-risk plaques causing major adverse cardiovascular events (MACE), but detailed OCT analysis requires time-consuming expert clinician or core laboratory analysis, whilst artifacts and limited sampling impair reproducibility. Assistive technologies such as artificial intelligence (AI)-based analysis may aid both interpretation and patient management, but many studies exhibit methodological, dataset, and reporting deficiencies that limit their robustness for clinical application.Methods We analysed 366 OCT pullbacks from 297 patients in two linked studies in patients with coronary artery disease. The OCT artifact burden that prevented accurate tissue identification and/or plaque measurements was determined using both ex-vivo (6,916 frames, 14 patients) and clinical trial OCT pullbacks (5,750 frames, 22 patients). 36,212 unselected frames (127 whole pullbacks, 106 patients) were used to design and validate a modular AI-based system that corrects artifact-containing OCT images and automatically identifies plaque composition and risk (AutoOCT), with outputs comprising labelled images and plaque component measurements. Validation of tissue and plaque classification used ex-vivo OCT pullbacks with co-registered histopathology, while external validation used core laboratory analysis of the high-intensity statin (HIS) IBIS-4 (83 baseline/follow-up patients) and natural history CLIMA (62 patients) studies respectively to determine if AI-based systems can detect plaque progression or regression, changes in plaque composition and higher-risk features with drugs, and future MACE.Results Autopsy and clinical pullbacks contained a similar and wide range of OCT artifacts in ~50% of frames, with ~10% preventing accurate measurement of plaque features. Pre-processing corrected >70% of common artifacts and improved sensitivity and accuracy to detect fibrous plaques and fibroatheroma. AutoOCT had a plaque classification accuracy of 83% vs. histology( figure 1)(table 1), equivalent to expert clinician readers, and replicated core laboratory plaque composition changes after HIS, including reduced lesion lipid arc (13.3° vs. 12.5°) and increased minimum fibrous cap thickness (FCT, 18.9µm vs. 24.4µm)(figure 2). AutoOCT also identified high-risk plaque features leading to MACE including minimal lumen area <3.5 mm2, Lipid arc >180°, and FCT <75µm, comparable to the CLIMA core laboratory (table 2).Abstract 2-022 Figure 1AutoOCT performance in high-risk lesionsLeft to right, Plaque components for higher-risk plaque-types measured on histology sections, co-registered OCT frames by AutoOCT, and expert OCT reader, with attention map output from the plaque classification system. Lipid and calcium arcs are labeled in yellow and blue respectively as degrees, and FCT by red line in microns.Abstract 2-022 Figure 2AutoOCT validation of drug effects against expert core laboratory(A-B) Bland-Altman plots of mean (x axis) and difference (y axis) with histograms of mean (top) and difference (right) for measurements of lipid arc (n=1218)(A) and minimum fibrous cap thickness (n=1384)(B). (C-D) Example fibroatheroma lesions that show regression of TCFA (C) or progression of ThCFA (D) after statin therapy. Fibrous caps are outlined in red.Abstract 2-022 Table 1Accuracy of AutoOCT and expert reader plaque classification compared with histology Histological Classification AutoOCT Low RiskThCFATCFAFibrocalcific Sensitivity, (%)72.6%70.2%27.3%12.5%Specificity, (%)88.3%70.4%88.7%96.7%PPV, (%)80.4%57.9%33.4%20.1%NPV, (%)83.0%80.3%85.4%94.3%Diagnostic Accuracy, (%)82.0% (0.012) 70.3% (0.001) 78.1% (0.001) 91.4% (<0.001) Expert OCT Reader Sensitivity, (%)72.7%53.2%31.8%50.0%Specificity, (%)87.8%72.8%84.0%91.7%PPV, (%)76.9%53.2%29.2%28.6%NPV, (%)85.2%72.9%85.6%96.5%Diagnostic Accuracy, (%)82.4%65.6%75.0%89.1%Low risk defined as normal vessel, adaptive intimal thickening (AIT), or pathological intimal thickening (PIT); TCFA, thin-cap fibroatheroma; ThCFA, thick-cap fibroatheroma; PPV, positive predictive value; NPV, negative predictive value. P values shown in brackets demonstrate non-inferiority between AutoOCT and Expert reader for each plaque-type.Abstract 2-022 Table 2Accuracy of AutoOCT to detect higher-risk plaque features compared to expert core laboratory Sensitivity (%)Specificity (%)PPV (%)NPV (%) Core Laboratory Minimum lumen area <3.5 mm2 27.786.06.896.9Minimum fibrous cap thickness <75µm40.683.98.697.4Maximum lipid arc extension >180°46.965.64.897.1 AutoOCT Minimum lumen area <3.5 mm2 36.780.06.597.1Minimum fibrous cap thickness <75µm30.086.77.997.0Maximum lipid arc extension >180°56.756.74.697.2NPV, negative predictive value; PPV, positive predictive valueConclusions Artifact-corrected, AI-based analysis of intracoronary OCT allows rapid analysis of all available data, identifies tissue and plaque types, and measures features of plaque stabilisation and high-risk plaques. AI-based OCT analysis may assist real-time clinical interpretation and augment clinician or core laboratory analysis for trials of drug/device efficacy and identifying high-risk lesions.",
  "authors": [
    {
      "affiliations": [
        "Section of Cardiorespiratory Medicine, University of Cambridge, UK"
      ],
      "name": "Benn Jessney"
    },
    {
      "affiliations": [
        "Section of Cardiorespiratory Medicine, University of Cambridge, UK"
      ],
      "name": "Xu Chen"
    },
    {
      "affiliations": [
        "Section of Cardiorespiratory Medicine, University of Cambridge, UK"
      ],
      "name": "Sophie Gu"
    },
    {
      "affiliations": [
        "Section of Cardiorespiratory Medicine, University of Cambridge, UK"
      ],
      "name": "Yuan Huang"
    },
    {
      "affiliations": [
        "Department of Pathology, Royal Papworth Hospital, Cambridge, UK"
      ],
      "name": "Martin Goddard"
    },
    {
      "affiliations": [
        "Monash University, Melbourne, Australia"
      ],
      "name": "Adam Brown"
    },
    {
      "affiliations": [
        "Swansea University Medical School, UK"
      ],
      "name": "Daniel Obaid"
    },
    {
      "affiliations": [
        "Interventional Cardiology, MedStar Washington Hospital Center, DC, USA"
      ],
      "name": "Michael Mahmoudi"
    },
    {
      "affiliations": [
        "Faculty of Medicine, University of Southampton, UK"
      ],
      "name": "Hector M Garcia Garcia"
    },
    {
      "affiliations": [
        "Department of Cardiology, Royal Papworth Hospital, Cambridge, UK"
      ],
      "name": "Stephen P Hoole"
    },
    {
      "affiliations": [
        "Department of Cardiology, Royal Papworth Hospital, Cambridge, UK"
      ],
      "name": "Charis Costopoulos"
    },
    {
      "affiliations": [
        "Cardiology Department, Bern University Hospital, University of Bern, Switzerland"
      ],
      "name": "Lorenz Räber"
    },
    {
      "affiliations": [
        "Department of Cardiology, San Giovanni Addolorata Hospital, Rome"
      ],
      "name": "Francesco Prati"
    },
    {
      "affiliations": [
        "Department of Applied Mathematics and Theoretical Physics, University of Cambridge, UK"
      ],
      "name": "Carola-Bibiane Schönlieb"
    },
    {
      "affiliations": [
        "Galway University Hospital, Ireland",
        "Cardiovascular Research Centre for Advanced Imaging and Core Laboratory (CORRIB), University of Galway"
      ],
      "name": "Yoshinobu Onuma"
    },
    {
      "affiliations": [
        "Cardiovascular Research Centre for Advanced Imaging and Core Laboratory (CORRIB), University of Galway"
      ],
      "name": "Patrick Serruys"
    },
    {
      "affiliations": [
        "Section of Cardiorespiratory Medicine, University of Cambridge, UK",
        "Department of Applied Mathematics and Theoretical Physics, University of Cambridge, UK"
      ],
      "name": "Michael Roberts"
    },
    {
      "affiliations": [
        "Section of Cardiorespiratory Medicine, University of Cambridge, UK"
      ],
      "name": "Martin Bennett"
    }
  ],
  "title": "2-022 Artifact-corrected artificial intelligence-LED intracoronary oct analysis identifies plaque progression and vulnerability, drug efficacy and patient events",
  "uid": "c0a651d6-f53a-52fd-b9cf-a40f42ac0c73"
}