{
  "abstract": "Background Cardiac and cerebral ischaemic diseases share multiple vascular risk factors (VRFs), largely driving atherosclerosis, and only partially explaining their interconnections. Cardiovascular magnetic resonance (CMR) radiomics, which detects subtle imaging features beyond conventional techniques, can enhance disease characterisation. We previously identified common radiomic features in cardiac and cerebral ischaemia, illustrating heart-brain relationships. 1 However, the extent to which VRFs influence these features remains unclear. This study aims to clarify the role of VRFs in the radiomic characteristics shared by cardiac and cerebral ischaemia.Methods We examined UK Biobank participants with prevalent ischaemic heart disease (IHD, n=781) or ischaemic cerebrovascular disease (n=360)—with no overlapping diagnoses—and their respective subsets (myocardial infarction [MI], n=542; ischaemic stroke [IS], n=119). After defining a radiomics signature for each condition, as previously described, we extracted the top 20 features common to all groups. Multivariable regression models, adjusted for age, sex, and body surface area, were then used to assess associations of these features with hypertension (HTN), adiposity (waist-hip ratio [WHR]), smoking, hypercholesterolaemia, and diabetes.Results VRFs had more significant associations with radiomics in cardiac than in cerebral ischaemia ( table 1). HTN and WHR were consistently linked across all groups. HTN had larger effects on cardiac remodelling in cerebral ischaemia, particularly increasing left ventricular (LV) systolic kurtosis, and that effect was greater in the IS subset (IS: β=0.42, CI [0.30; 0.55], p<0.001; cerebrovascular disease: β=0.23, CI [0.15; 0.32], p<0.001). In cardiac ischaemia, HTN presented smaller but distinct associations, including elevated myocardial volume in diastole (IHD: β=0.08, CI [0.05; 0.11]; p<0.0001; MI: β=0.11, CI [0.06; 0.16]; p =0.0002) and systolic remodelling in MI. WHR showed the most extensive effects in all groups, predominantly involving diastolic-focused remodelling marked by increased LV sphericity. Other VRFs, although with smaller effects, emerged in specific scenarios: smoking in IHD (subtle LV alterations during diastole), hypercholesterolaemia in cerebrovascular disease (systolic LV remodelling), and diabetes in MI (diastolic myocardial changes).Abstract 6-017 Table 1Associations between VRFs and significant radiomics features in cardiac ischaemia (IHD) and cerebrovascular diseasesThe beta values are standardised, and the p-value is corrected for multiple tests. ED = end-diastole; ES = end-systole; CI = confidence interval; IHD = ischaemic heart disease; RV = right ventricle; LV = left ventricle; MYO = myocardium; VRFs = vascular risk factors; WHR = waist-hip ratio.Conclusion Hypertension and adiposity were the main VRFs shaping shared cardiac radiomic features in cardiac and cerebral ischaemic diseases. While adiposity drove similar remodelling patterns, hypertension resulted in distinct effects, indicating that other VRFs, organ-specific processes, and heart-brain interactions also influenced cardiac adaptations. These findings support more comprehensive management strategies in ischaemic patients that extend beyond VRFs control, with CMR radiomics offering a key tool to explore these complexities.Reference Rauseo E, et al. New imaging signatures of cardiac alterations in ischaemic heart disease and cerebrovascular disease using CMR radiomics. Front Cardiovasc Med. 2021 Sep 23;8(6).",
  "authors": [
    {
      "affiliations": [
        "William Harvey Research Institute, NIHR Barts Biomedical Research Centre, Queen Mary University London, Charterhouse Square, London, EC1M 6BQ, UK",
        "Barts Heart Centre, St Bartholomew’s Hospital, Barts Health NHS Trust, West Smithfield, EC1A 7BE, London, UK"
      ],
      "name": "Elisa Rauseo"
    },
    {
      "affiliations": [
        "William Harvey Research Institute, NIHR Barts Biomedical Research Centre, Queen Mary University London, Charterhouse Square, London, EC1M 6BQ, UK",
        "Barts Heart Centre, St Bartholomew’s Hospital, Barts Health NHS Trust, West Smithfield, EC1A 7BE, London, UK",
        "Department of Population Health Sciences, University of Leicester, Leicester UK",
        "PRIME Lab, Scientific Research Center, University of Zakho, Kurdistan Region, Iraq"
      ],
      "name": "Ahmed Salih"
    },
    {
      "affiliations": [
        "Departament de Matemàtiques i Informàtica, Universitat de Barcelona, Artificial Intelligence in Medicine Lab (BCN-AIM), Barcelona, Spain"
      ],
      "name": "Cristian Izquierdo Morcillo"
    },
    {
      "affiliations": [
        "Departament de Matemàtiques i Informàtica, Universitat de Barcelona, Artificial Intelligence in Medicine Lab (BCN-AIM), Barcelona, Spain"
      ],
      "name": "Polyxeni Gkontra"
    },
    {
      "affiliations": [
        "William Harvey Research Institute, NIHR Barts Biomedical Research Centre, Queen Mary University London, Charterhouse Square, London, EC1M 6BQ, UK",
        "Barts Heart Centre, St Bartholomew’s Hospital, Barts Health NHS Trust, West Smithfield, EC1A 7BE, London, UK"
      ],
      "name": "Nay Aung"
    },
    {
      "affiliations": [
        "Digital Environment Research Institute, Queen Mary University of London, UK",
        "School of Electronic Eng. & Computer Science, Queen Mary University of London, UK"
      ],
      "name": "Gregory G Slabaugh"
    },
    {
      "affiliations": [
        "Departament de Matemàtiques i Informàtica, Universitat de Barcelona, Artificial Intelligence in Medicine Lab (BCN-AIM), Barcelona, Spain"
      ],
      "name": "Karim Lekadir"
    },
    {
      "affiliations": [
        "William Harvey Research Institute, NIHR Barts Biomedical Research Centre, Queen Mary University London, Charterhouse Square, London, EC1M 6BQ, UK",
        "Barts Heart Centre, St Bartholomew’s Hospital, Barts Health NHS Trust, West Smithfield, EC1A 7BE, London, UK"
      ],
      "name": "Steffen E Petersen"
    }
  ],
  "title": "6-017 Cardiac and cerebral ischaemic diseases: how vascular risk factors shape shared cardiac radiomics features",
  "uid": "974cfb4b-ce0d-5882-8ca8-b088a0472353"
}