SELECT data_key, data_value
FROM xf_data_registry
WHERE data_key IN ('options', 'languages', 'contentTypes', 'codeEventListeners', 'deferredRun', 'simpleCache', 'addOns', 'defaultStyleProperties', 'routeFiltersIn', 'routeFiltersOut', 'routesPublic', 'nodeTypes', 'bannedIps', 'discouragedIps', 'styles', 'displayStyles', 'userBanners', 'smilies', 'bbCode', 'threadPrefixes', 'userTitleLadder', 'reportCounts', 'moderationCounts', 'userModerationCounts', 'notices', 'userFieldsInfo')Run Time: 0.000562
| Select Type | Table | Type | Possible Keys | Key | Key Len | Ref | Rows | Extra |
|---|
| SIMPLE | xf_data_registry | range | PRIMARY | PRIMARY | 27 | | 26 | Using where |
SELECT data_key, data_value
FROM xf_data_registry
WHERE data_key IN ('brListenerClasses', 'brBriviumAddOns')Run Time: 0.000155
| Select Type | Table | Type | Possible Keys | Key | Key Len | Ref | Rows | Extra |
|---|
| SIMPLE | xf_data_registry | range | PRIMARY | PRIMARY | 27 | | 2 | Using where |
SELECT *
FROM xf_brivium_addon
ORDER BY addon_id
Run Time: 0.000145
| Select Type | Table | Type | Possible Keys | Key | Key Len | Ref | Rows | Extra |
|---|
| SIMPLE | xf_brivium_addon | index | | PRIMARY | 77 | | 1 | |
INSERT INTO xf_data_registry
(data_key, data_value)
VALUES
(?, ?)
ON DUPLICATE KEY UPDATE
data_value = VALUES(data_value)
Params: brBriviumAddOns, a:0:{}
Run Time: 0.000156
SELECT cache_value
FROM xf_permission_combination
WHERE permission_combination_id = ?
Params: 1
Run Time: 0.000153
| Select Type | Table | Type | Possible Keys | Key | Key Len | Ref | Rows | Extra |
|---|
| SIMPLE | xf_permission_combination | const | PRIMARY | PRIMARY | 4 | const | 1 | |
SELECT thread.*
,
user.gender, user.avatar_date, user.gravatar,
NULL AS thread_read_date,
0 AS thread_reply_banned,
0 AS thread_is_watched,
'' AS draft_message, NULL AS draft_extra
FROM xf_thread AS thread
LEFT JOIN xf_user AS user ON
(user.user_id = thread.user_id)
WHERE thread.thread_id = ?
Params: 314322
Run Time: 0.000418
| Select Type | Table | Type | Possible Keys | Key | Key Len | Ref | Rows | Extra |
|---|
| SIMPLE | thread | const | PRIMARY | PRIMARY | 4 | const | 1 | |
| SIMPLE | user | const | PRIMARY | PRIMARY | 4 | const | 1 | |
SELECT node.*, forum.*
,
permission.cache_value AS node_permission_cache,
NULL AS forum_read_date
FROM xf_forum AS forum
INNER JOIN xf_node AS node ON (node.node_id = forum.node_id)
LEFT JOIN xf_permission_cache_content AS permission
ON (permission.permission_combination_id = 1
AND permission.content_type = 'node'
AND permission.content_id = forum.node_id)
WHERE node.node_id = ?
Params: 25
Run Time: 0.000371
| Select Type | Table | Type | Possible Keys | Key | Key Len | Ref | Rows | Extra |
|---|
| SIMPLE | forum | const | PRIMARY | PRIMARY | 4 | const | 1 | |
| SIMPLE | node | const | PRIMARY | PRIMARY | 4 | const | 1 | |
| SIMPLE | permission | const | PRIMARY | PRIMARY | 35 | const,const,const | 1 | |
SELECT post.*
,
thread.*, thread.user_id AS thread_user_id, thread.username AS thread_username,
thread.post_date AS thread_post_date,
post.user_id, post.username, post.post_date,
bb_code_parse_cache.parse_tree AS message_parsed, bb_code_parse_cache.cache_version AS message_cache_version,
user.*, IF(user.username IS NULL, post.username, user.username) AS username,
user_profile.*,
user_privacy.*,
signature_parse_cache.parse_tree AS signature_parsed, bb_code_parse_cache.cache_version AS signature_cache_version,
session_activity.view_date AS last_view_date,
0 AS like_date
FROM xf_post AS post
INNER JOIN xf_thread AS thread ON
(thread.thread_id = post.thread_id)
LEFT JOIN xf_bb_code_parse_cache AS bb_code_parse_cache ON
(bb_code_parse_cache.content_type = 'post' AND bb_code_parse_cache.content_id = post.post_id)
LEFT JOIN xf_user AS user ON
(user.user_id = post.user_id)
LEFT JOIN xf_user_profile AS user_profile ON
(user_profile.user_id = post.user_id)
LEFT JOIN xf_user_privacy AS user_privacy ON
(user_privacy.user_id = post.user_id)
LEFT JOIN xf_bb_code_parse_cache AS signature_parse_cache ON
(signature_parse_cache.content_type = 'signature' AND signature_parse_cache.content_id = post.user_id)
LEFT JOIN xf_session_activity AS session_activity ON
(post.user_id > 0 AND session_activity.user_id = post.user_id AND session_activity.unique_key = CAST(post.user_id AS BINARY))
WHERE (
(post.thread_id = ? AND (post.position >= 0 AND post.position < 10) )
)
AND (post.message_state IN ('visible'))
ORDER BY post.position ASC, post.post_date ASCParams: 314322
Run Time: 0.002184
| Select Type | Table | Type | Possible Keys | Key | Key Len | Ref | Rows | Extra |
|---|
| SIMPLE | thread | const | PRIMARY | PRIMARY | 4 | const | 1 | Using filesort |
| SIMPLE | post | ref | thread_id_post_date,thread_id_position | thread_id_post_date | 4 | const | 1 | Using index condition; Using where |
| SIMPLE | bb_code_parse_cache | eq_ref | content_type_id | content_type_id | 31 | const,xenforo.cc.post.post_id | 1 | Using where |
| SIMPLE | user | eq_ref | PRIMARY | PRIMARY | 4 | xenforo.cc.post.user_id | 1 | |
| SIMPLE | user_profile | eq_ref | PRIMARY | PRIMARY | 4 | xenforo.cc.post.user_id | 1 | |
| SIMPLE | user_privacy | eq_ref | PRIMARY | PRIMARY | 4 | xenforo.cc.post.user_id | 1 | |
| SIMPLE | signature_parse_cache | eq_ref | content_type_id | content_type_id | 31 | const,xenforo.cc.post.user_id | 1 | Using where |
| SIMPLE | session_activity | eq_ref | PRIMARY | PRIMARY | 22 | xenforo.cc.post.user_id,func | 1 | Using where |
INSERT INTO xf_thread_view
(thread_id)
VALUES
(?)
Params: 314322
Run Time: 0.000089
INSERT INTO `xf_session` (`session_id`, `session_data`, `expiry_date`) VALUES (?, ?, ?)
Params: 6c554e58420b808de7e903781735b4c7, , 1714907506
Run Time: 0.001038
INSERT INTO xf_session_activity
(user_id, unique_key, ip, controller_name, controller_action, view_state, params, view_date, robot_key)
VALUES
(?, ?, ?, ?, ?, ?, ?, ?, ?)
ON DUPLICATE KEY UPDATE
ip = VALUES(ip),
controller_name = VALUES(controller_name),
controller_action = VALUES(controller_action),
view_state = VALUES(view_state),
params = VALUES(params),
view_date = VALUES(view_date),
robot_key = VALUES(robot_key)
Params: 0, , , XenForo_ControllerPublic_Thread, Index, valid, thread_id=314322, 1714903906,
Run Time: 0.000108
INSERT INTO xf_bb_code_parse_cache
(content_type, content_id, parse_tree, cache_version, cache_date)
VALUES (?, ?, ?, ?, ?)
ON DUPLICATE KEY UPDATE parse_tree = VALUES(parse_tree),
cache_version = VALUES(cache_version),
cache_date = VALUES(cache_date)
Params: post, 320950, a:7:{i:0;s:1079:"快速准确的病情评估对治疗新冠肺炎患者极为重要。当地时间5月14日，华中科技大学同济医学院附属同济医院等机构在学术期刊《自然-机器智能》发文表示，利用3个生物标志物，机器学习算法可以至少提前10天预测新冠肺炎患者个体的死亡率，准确率超过90%。这3个生物标志物分别是乳酸脱氢酶、淋巴细胞和超敏C反应蛋白水平。
结合这些标志物，论文提出了一个简单且易操作的决策规则，可以快速预测风险最高的新冠肺炎患者，使他们能够被优先考虑，并有可能降低死亡率。
上述论文题为“An interpretable mortality prediction model for COVID-19 patients”，第一作者为华中科技大学同济医学院附属同济医院急诊科副主任医师严丽。华中科技大学同济医学院附属同济医院急诊与重症医学科主任李树生、华中科技大学同济医学院附属同济医院麻醉科徐慧和华中科技大学人工智能与自动化学院教授袁烨共为论文通讯作者。
";i:1;a:4:{s:3:"tag";s:3:"url";s:6:"option";s:67:"https://static.cnbetacdn.com/article/2020/0517/37688dbf25514d8.jpeg";s:8:"original";a:2:{i:0;s:75:"[URL='https://static.cnbetacdn.com/article/2020/0517/37688dbf25514d8.jpeg']";i:1;s:6:"[/URL]";}s:8:"children";a:1:{i:0;a:4:{s:3:"tag";s:3:"img";s:6:"option";N;s:8:"original";a:2:{i:0;s:5:"[IMG]";i:1;s:6:"[/IMG]";}s:8:"children";a:1:{i:0;s:97:"https://rdimg.com/proxy/https://static.cnbetacdn.com/thumb/article/2020/0517/37688dbf25514d8.jpeg";}}}}i:2;s:333:"
基于3个生物标志物及其绝对阈值的决策规则
严丽等人收集了中国武汉485名新冠肺炎患者的血液标本，从中寻找可靠且有意义的死亡风险标志物。这些样本的采集时间为2020年1月10日至2020年2月18日，采集地点在华中科技大学同济医学院附属同济医院。
";i:3;a:4:{s:3:"tag";s:3:"url";s:6:"option";s:67:"https://static.cnbetacdn.com/article/2020/0517/c48e8b2dab7c72b.jpeg";s:8:"original";a:2:{i:0;s:75:"[URL='https://static.cnbetacdn.com/article/2020/0517/c48e8b2dab7c72b.jpeg']";i:1;s:6:"[/URL]";}s:8:"children";a:1:{i:0;a:4:{s:3:"tag";s:3:"img";s:6:"option";N;s:8:"original";a:2:{i:0;s:5:"[IMG]";i:1;s:6:"[/IMG]";}s:8:"children";a:1:{i:0;s:97:"https://rdimg.com/proxy/https://static.cnbetacdn.com/thumb/article/2020/0517/c48e8b2dab7c72b.jpeg";}}}}i:4;s:251:"
病例分类
样本被分成两个部分，其中375例用于建模，另外110例用于模型测试。用于建模的375例病例，有201例在康复后出院，其余174例死亡。 用于测试的110例病例中，97例存活，13例死亡。
";i:5;a:4:{s:3:"tag";s:3:"url";s:6:"option";s:67:"https://static.cnbetacdn.com/article/2020/0517/b37ba7634fd50fb.jpeg";s:8:"original";a:2:{i:0;s:75:"[URL='https://static.cnbetacdn.com/article/2020/0517/b37ba7634fd50fb.jpeg']";i:1;s:6:"[/URL]";}s:8:"children";a:1:{i:0;a:4:{s:3:"tag";s:3:"img";s:6:"option";N;s:8:"original";a:2:{i:0;s:5:"[IMG]";i:1;s:6:"[/IMG]";}s:8:"children";a:1:{i:0;s:97:"https://rdimg.com/proxy/https://static.cnbetacdn.com/thumb/article/2020/0517/b37ba7634fd50fb.jpeg";}}}}i:6;s:1725:"
机器学习模型在测试数据集中的性能表现
作者根据机器学习算法设计了一套数学建模方法，用于识别最能预测患者死亡率的生物标志物。作者在算法中将问题变成一个分类任务，输入数据包括基本信息、症状、血液标本，以及普通、重症、危重症患者的实验室检查结果，如肝功能、肾功能、凝血功能、电解质和炎症因子。
最终，该模型将乳酸脱氢酶、淋巴细胞、超敏C 反应蛋白水平选为识别新冠肺炎病危患者的3个关键生物标志物。作者认为，这一结果与当前的医学知识相符。
作者在论文中分析称，乳酸脱氢酶水平较高本身与不同疾病出现的组织损伤有关，包括肺炎等肺部疾病。对于新冠肺炎危重患者，乳酸脱氢酶水平的升高表明肺损伤的程度增加。超敏C反应蛋白水平升高是急性呼吸窘迫综合征预后不良的重要标志，反映了持续的炎症状态。此外，论文表示，“我们的结果还表明淋巴细胞或能作为一个潜在的治疗靶标，这一假设得到了临床研究结果的支持。”论文写道，淋巴细胞减少是新冠肺炎患者的共同特征，可能是与疾病严重程度和致死率相关的关键因素。
作者总结称，他们的模型可以通过简单直观且易解读的临床检查来精准、快速地量化新冠患者的死亡风险。作者称，模型所涉及的3个生物标志物在任何一家医院都可以很容易地收集到。“在拥挤的医院和医疗资源短缺的情况下，这个简单的模型可以帮助医生快速确定患者的优先顺序，特别是在医疗资源有限的疫情期间。”";}, 1621686945, 1714903906
Run Time: 0.000947