Aortic dissection has acute onset and high mortality, with immunoinflammatory response driving lesion progression. Current perioperative anti-inflammatory therapies are mostly empirical and poorly targeted, and AI-assisted typing lacks a complete clinical translation pathway. This study integrates multi-dimensional data to construct an AI immunoinflammatory subtyping system, enabling rapid subtyping and establishing a "subtyping-target-treatment" closed loop for emergency needs. Using a prospective multicenter RCT, 300 patients are randomly divided into two groups: the experimental group receives subtyping-based precision therapy, while the control group uses empirical strategies (treatment of physician's choice). It observes 7-day postoperative SOFA score, SIRS and other prognostic indicators to provide evidence-based support for precision treatment.

Inclusion Criteria:

1. Confirmed diagnosis of acute aortic dissection by contrast-enhanced Computed Tomography Angiography (CTA) or Magnetic Resonance Angiography (MRA);
2. Planned emergency surgical treatment (including open surgery and endovascular repair);
3. Aged 18-80 years old, regardless of gender;
4. Time from onset to hospital admission ≤ 72 hours;
5. Signed informed consent form by the patient or their authorized agent, with willingness to cooperate with the follow-up of the study.

Exclusion Criteria:

1. Complicated with underlying diseases that affect immunoinflammatory status, such as severe infections (e.g., sepsis, infective endocarditis), autoimmune diseases, malignant tumors, chronic liver diseases, and chronic kidney diseases (uremic stage);
2. Long-term use of immunosuppressants or glucocorticoids before surgery (continuous use for ≥ 2 weeks);
3. Pregnant or lactating women;
4. Patients with mental disorders or cognitive dysfunction who cannot cooperate with the study.