Postoperative Cognitive Dysfunction(POCD) is commonly seen in cardiac surgery, which may lead to poor pognosis. Cerebral small vessel disease(CVSD) is refer as the main resource of delirium among elderly people. In the study, CVSD will be diagnosed using multimodal MRI. And we want to select a high correlating COPD biomarker through CyTOF. We also want to investigate a medical model to select the high risk patients who may suffer from POCD after cardiac surgery.

This study will investigate whether preoperative CSVD and CyTOF will help to predict the occurencce of Postoprative Cognitive Dysfunction(POCD) in 1 year after surgery. The investigators perform a prospective cohort study among selective cardiac surgery patients between 50 to 85 years old. All participates have pre-operative head multimodel MRI in Renji Hospital Affiliated to Shanghai Jiaotong University. Incusion and exclusion criteria and relevant data were collected. Patients' age, gender, specific surgical methods, educational background, BMI, tobacco and alcohol history, and systematic medical history were collected before surgery, specific history (heart disease, heart failure, arrhythmia, cornoary artery stent implantation, and history of cardiac surgery), medical history (blood pressure medication, anti-arrhythmic drugs, diuresis, anti-arrhythemic drugs, diuresis, anticoagulant drugs), laboratory examination, auxiliary examination(blood routine, blood coagulation, liver and kidney function, etc.), and Cognitive Function Assessment Scale(MMSE, MOCA, ADL, GDSS). Perioperative indicators were also collected, including use of anesthetics, anestetic time,vital signs, etc. Blood sample should be taken before and after surgery. Postopeative data including: POD whithin 5 days after surgery, cardiovascular drugs use in ICU, mechanical ventilation time, length of stay in ICU, occurrence of complicatiosns of patients, Cognitive function Assessment Scale 1 year after surgery. POD was assessed by a professionally trained clinician twice daily at 8-12 hour intervals, using CAM-ICU scale, if diagnosis with POD, DRS-98 is used to evaluate th severity of POD. POCD was assessed by a professionally trained clinican at 1 year after surgery. The incidence of POD and POCD was analyzed between groups, and the association between each index and delirium and POCD was analyzed by logistic regresssion. The waste blood from routine peripheral blood examination was collected before operation, at the end of operation, on the second day after operation and one year after opertion for the detection of biomarkers by CyTOF, and the composite model of MRI and biomarkers was established.

Inclusion Criteria:

1. Age from 50 years to 85 years;
2. The patient is going to have selective cardiac surgery;
3. Written informed consent is obtained before the surgery.

Exclusion Criteria:

1. mental illness;
2. Already have other diseases that can cause dementia, such as Alzheimer's disease, Lewy body dementia, Frontotemporal dementia, progressive supranuclear palsy, Parkinson's disease, Creutzfeldt-Jakob disease, Huntington's disease, alcohol and drug dependence, Neurosyphilis, systemic lupus erythematosus; or preoperative MRI shows hippocampal and temporal lobe atrophy more than three levels
3. Have suffered from other brain diseases (such as overt stroke, multiple sclerosis, central nervous system infection with sequelae, etc.);
4. Preoperative MRI indicates covert stroke MMSE or MoCA scale cannot be completed due to other reasons (such as hearing impairment or visual impairment)
5. MRI contraindications before or after surgery or patients who cannot tolerate MRI imaging
6. Stage 3 or 4 malignant tumors, and high malignancy and poor prognosis cancer, such as pancreatic cancer, gallbladder cancer, and bile duct cancer.

The incidence of POCD after cardiac surgery varies greatly in different studies. Accroding to JAMA and Lancet papers, the incidence of POCD in patients one year after cardiac surgeries is 20-31%. According to the Events per Variable method, the main objective model establishment will explore 5 Events of independent variables. Assuming EPV is 10, the incidence of POCD is 26%, so the number of POCD patients after surgeries should be 5*10=50. Take 10% loss of follow-up rate into account, the total sample size required is 50/26%*110%=211

All IPD that underlie results in a publication will be shared starting 6 months after publication.

All the IPD and any additional supporting information will become available starting 6 months after publication for at least 5 years.

Requests to access all the IPD and additional supporting information will be addressed and reviewed by the corresponding author of the related publication. Access criteria include that the requests shall be submitted by a researcher on an institutional headed paper; the requests shall come with all the detailed contact information of the researcher and the administration office of the institution.