训练预测脓毒血症疾病发展的文本大模型的JSON数据集样例
[
{
"patient_id": "001",
"age": 45,
"gender": "male",
"medical_history": {
"diabetes": true,
"hypertension": true,
"heart_disease": false
},
"symptoms": {
"fever": true,
"chills": true,
"confusion": false,
"hypotension": true,
"tachycardia": true,
"tachypnea": true
},
"lab_results": {
"white_blood_cell_count": 15000,
"blood_culture": "positive",
"creatinine": 1.2,
"lactate": 2.5
},
"medications": {
"antibiotics": ["ceftriaxone"],
"vasopressors": ["norepinephrine"],
"fluids": ["normal saline"]
},
"vital_signs": {
"heart_rate": 110,
"blood_pressure": 90,
"respiratory_rate": 24,
"temperature": 38.5,
"oxygen_saturation": 92
},
"disease_progression": "severe"
},
{
"patient_id": "002",
"age": 32,
"gender": "female",
"medical_history": {
"diabetes": false,
"hypertension": false,
"heart_disease": false
},
"symptoms": {
"fever": true,
"chills": true,
"confusion": true,
"hypotension": true,
"tachycardia": true,
"tachypnea": true
},
"lab_results": {
"white_blood_cell_count": 18000,
"blood_culture": "positive",
"creatinine": 1.5,
"lactate": 3.0
},
"medications": {
"antibiotics": ["ceftriaxone", "vancomycin"],
"vasopressors": ["norepinephrine"],
"fluids": ["normal saline"]
},
"vital_signs": {
"heart_rate": 120,
"blood_pressure": 85,
"respiratory_rate": 26,
"temperature": 39.0,
"oxygen_saturation": 90
},
"disease_progression": "critical"
},
{
"patient_id": "003",
"age": 58,
"gender": "male",
"medical_history": {
"diabetes": true,
"hypertension": true,
"heart_disease": true
},
"symptoms": {
"fever": true,
"chills": false,
"confusion": true,
"hypotension": true,
"tachycardia": true,
"tachypnea": true
},
"lab_results": {
"white_blood_cell_count": 20000,
"blood_culture": "positive",
"creatinine": 2.0,
"lactate": 3.5
},
"medications": {
"antibiotics": ["ceftriaxone", "vancomycin"],
"vasopressors": ["norepinephrine", "epinephrine"],
"fluids": ["normal saline"]
},
"vital_signs": {
"heart_rate": 125,
"blood_pressure": 80,
"respiratory_rate": 28,
"temperature": 39.5,
"oxygen_saturation": 88
},
"disease_progression": "critical"
},
{
"patient_id": "004",
"age": 28,
"gender": "female",
"medical_history": {
"diabetes": false,
"hypertension": false,
"heart_disease": false
},
"symptoms": {
"fever": true,
"chills": true,
"confusion": false,
"hypotension": false,
"tachycardia": true,
"tachypnea": true
},
"lab_results": {
"white_blood_cell_count": 12000,
"blood_culture": "negative",
"creatinine": 1.0,
"lactate": 1.8
},
"medications": {
"antibiotics": ["ceftriaxone"],
"vasopressors": false,
"fluids": ["normal saline"]
},
"vital_signs": {
"heart_rate": 100,
"blood_pressure": 100,
"respiratory_rate": 20,
"temperature": 38.0,
"oxygen_saturation": 95
},
"disease_progression": "mild"
}
]
说明
patient_id: 患者的唯一标识符。
age: 患者的年龄。
gender: 患者的性别。
medical_history: 患者的病史,包括糖尿病、高血压和心脏病等。
symptoms: 患者的症状,包括发热、战栗、混乱、低血压、心动过速和呼吸加快等。
lab_results: 实验室检查结果,包括白细胞计数、血培养、肌酐素和乳酸等。
medications: 患者正在使用的药物,包括抗生素、血管加压剂和输液等。
vital_signs: 生命体征,包括心率、血压、呼吸频率、体温和氧饱和度等。
disease_progression: 疾病的进展程度,分为“mild”(轻度)、“severe”(严重)和“critical”(危重)等级。
使用方法
数据预处理:根据需要对数据进行清洗、标准化和匿名化处理。
模型训练:将数据用于训练文本大模型,特别是针对脓毒血症疾病进展的预测任务。
模型评估:使用适当的评估指标(如准确率、灵敏度、特异性等)评估模型性能。
临床应用:将训练好的模型集成到临床决策支持系统中,帮助医生预测患者的疾病进展,指导个性化治疗方案的制定。