{
  "abstract": "Introduction Systemic sclerosis (SSc) is a complex autoimmune disease with significant morbidity and mortality, primarily due to progressive multiorgan involvement. Early identification of organ-specific complications, such as interstitial lung disease (ILD), cardiac dysfunction, and renal involvement, is essential for optimizing treatment. However, predicting organ involvement remains a significant clinical challenge. Advances in artificial intelligence (AI) offer novel opportunities to integrate diverse datasets for enhanced prediction accuracy.Material and Methods 60 patients were included in the study (all females), mean age 53.08 ± 12.35 years, incorporating demographic data, disease characteristics (e.g., duration, specific autoantibody profiles), organ-specific findings (e.g., pulmonary, renal, cardiac involvement), HRCT imaging, and biomarkers of inflammation. A significant proportion of patients are enrolled at EUSTAR centers 254 and 255.Results A total of 60 patients with systemic sclerosis (SSc) were included (diffuse cutaneous: 46.7%; limited cutaneous: 53.3%). Mean disease duration was 9.7 ± 5.8 years. Organ involvement was observed in 68.3% of patients, most commonly ILD (56.7%), digital ulcers (63.3%), gastrointestinal involvement (46.7%), cardiac (18.3%), PAH (15.0%), and renal (6.7%). Clinical outcomes showed improvement in 75.0% and progression in 25.0%.Multivariate analysis demonstrated a significant association between overall organ involvement and longer disease duration (OR 8.8, 95% CI 2.7–28.5, p<0.05). PAH was strongly associated with the diffuse cutaneous subtype (OR 5.0, 95% CI 2.0–26.5, p<0.05). ILD correlated with disease duration (OR 8.8, p<0.05), while an inverse association was observed with anti-centromere antibodies (OR 1.32, 95% CI 0.45–3.8, p<0.05).AI-based prediction yielded similar frequencies: ILD 40–60%, gastrointestinal involvement 80%, renal 5–10%, PAH 10–15%, and digital ulcers 30–50%.Conclusions This study highlights the potential of AI in predicting organ involvement in SSc by integrating clinical, imaging, and laboratory data. Significant associations were identified, including the correlation of organ involvement and ILD with disease duration and the link between PAH and the diffuse cutaneous subtype. Conversely, ILD was negatively associated with anti-centromere antibodies.AI predictions aligned closely with observed outcomes, demonstrating its utility in estimating risks for ILD, gastrointestinal involvement, kidney involvement, PAH and digital ulcers. These findings underscore AI’s promise in enhancing early detection and guiding clinical decision-making in SSc.",
  "authors": [
    {
      "affiliations": [
        "Mikaelyan Institute of Surgery, Varus med, Yerevan, Armenia"
      ],
      "name": "Nune Manukyan"
    },
    {
      "affiliations": [
        "Varus Med, Yerevan, Armenia"
      ],
      "name": "Valentina Vardanyan"
    },
    {
      "affiliations": [
        "Nairi medical center, Yerevan, Armenia"
      ],
      "name": "Vahan Mukuchyan"
    },
    {
      "affiliations": [
        "Mikaelyan Institute of Surgery, Varus med, Yerevan, Armenia"
      ],
      "name": "Anna Sargsyan"
    },
    {
      "affiliations": [
        "Mikaelyan Institute of Surgery, Varus med, Yerevan, Armenia"
      ],
      "name": "Syuzanna Vardanyan"
    },
    {
      "affiliations": [
        "Mikaelyan Institute of Surgery, Varus med, Yerevan, Armenia"
      ],
      "name": "Sona Mardoyan"
    }
  ],
  "title": "P.304 Artificial intelligence-based prediction of organ involvement in systemic sclerosis",
  "uid": "7f2a4c6b-7c29-5b1c-8cae-6a3e51cc6316"
}