ADNI-Study-Data

梳理ADNI文档资料:表格数据 tabular data + 文档及压缩文件

Study Info: Data & Databases
Key ADNI tables merged into one table

aka ADNIMERGE.csv ☆☆☆☆☆

患者基本信息

  • RID (Participant roster ID) ex. 2、PTID (Original study protocol) ex. 011_S_0002、VISCODE (Visit code) ex. bl、SITE ex. 11、COLPROT (Study protocol of data collection) ex. ADNI1、ORIGPROT (Original study protocol) ex. ADNI1、EXAMDATE (Date) ex. 9/8/2005、DX (Diagnosis) ex. CN

  • AGE 54.4-91.4、PTGENDER (Sex) Female \ male、PTEDUCAT (Education) 4-20、PTETHCAT (Ethnicity)PTRACCAT (Race)PTMARRY (Marital)

生物标记物量化值

  • ABETA (CSF ABETA)APOE4 (0, 1, 2)FDG (Average FDG-PET of angular, temporal, and posterior cingulate)PIB (Average PIB SUVR of frontal cortex, anterior cingulate, precuneus cortex, and parietal cortex)AV45 (Reference region - florbetapir mean of whole cerebellum. Regions defined by Freesurfer; see Jagust lab PDF on LONI for details)TAU (CSF TAU)PTAU (CSF PTAU)

认知评估

  • CDRSB (Clinical dementia rating sum of boxes)MMSE (Mini-mental state evaluation)FAQ (Functional Activities Questionnaire)MOCA (Montreal cognitive assessment)LDELTOTAL (Logical Memory - Delayed Recall)DIGITSCOR (Digit Symbol Substitution)

  • ADAS11ADAS13ADASQ4 (ADAS Delayed Word Recall)

  • mPACCdigit (ADNI modified Preclinical Alzheimer's Cognitive Composite (PACC) with Digit Symbol Substitution)mPACCtrailsB (ADNI modified Preclinical Alzheimer's Cognitive Composite (PACC) with Trails B)TRABSCOR (Trails B)

  • RAVLT (forgetting)RAVLT_immediate (sum of 5 trials)RAVLT_learning (RAVLT Learning (trial 5 - trial 1))RAVLT_forgetting (RAVLT Forgetting (trial 5 - delayed))RAVLT_perc_forgetting (RAVLT Percent Forgetting)

  • EcogPtMemEcogPtLangEcogPtVisspatEcogPtPlanEcogPtOrganEcogPtDivattEcogPtTotalEcogSPMemEcogSPLangEcogSPVisspatEcogSPPlanEcogSPOrganEcogSPDivattEcogSPTotal

解剖结构量化值

  • IMAGEUIDVentricles 心室Hippocampus 海马体WholeBrainEntorhinal 嗅内区Fusiform 梭状回MidTemp 颞中回ICV → 可以和 PET数据的定量数据 一起分析,对结构向数据功能向数据的结合使用可以有效降低自然衰老带来的个体间差异

Other

  • FSVERSION (FreeSurfer Software Version)FLDSTRENG (MRI Field Strength)Month (Months since baseline)M (Months since baseline)

And 上述各信息的bl值

Key ADNI tables merged into one table - Dictionary

aka ADNIMERGE_DICT.csv





Data Dictionary [ADNI1,GO,2,3]

aka DATADIC.csv 为 ADNIMERGE_DICT.csv 条目解释进行补充

Merged ADNI 1/GO/2 Methods (PDF)

aka ADNIMERGE_Methods_21030429.pdf

ALL Diagnosis
Diagnosis and Symptoms Checklist [ADNI1,GO,2]

Total 4884 items:ADNI1 (4201 items) + ADNIGO (683 items)

  • header

Phase ID RID SITEID VISCODE VISCODE2 USERDATE USERDATE2 EXAMDATE AXNAUSEA AXVOMIT AXDIARRH AXCONSTP AXABDOMN AXSWEATN AXDIZZY AXENERGY AXDROWSY AXVISION AXHDACHE AXDRYMTH AXBREATH AXCOUGH AXPALPIT AXCHEST AXURNDIS AXURNFRQ AXANKLE AXMUSCLE AXRASH AXINSOMN AXDPMOOD AXCRYING AXELMOOD AXWANDER AXFALL AXOTHER AXSPECIF update_stamp

Diagnostic Summary - Baseline Changes [ADNI1,GO,2,3]

Total 12161 items:ADNI1 (3871 items) + ADNI2 (5666 items) + ADNI3 (2149 items) + ADNIGO (475 items)

  • header

Phase ID RID SITEID VISCODE VISCODE2 USERDATE USERDATE2 EXAMDATE BCPREDX BCADAS BCMMSE BCMMSREC BCNMMMS BCNEUPSY BCNONMEM BCFAQ BCCDR BCDEPRES BCSTROKE BCDELIR BCEXTCIR BCEXTSP BCCORADL BCCORCOG BCSUMM update_stamp

Diagnostic Summary [ADNI1,GO,2,3]

Total 12156 items:ADNI1 (3868 items) + ADNI2 (5671 items) + ADNI3 (2142 items) + ADNIGO (475 items)

  • header

Phase ID RID SITEID VISCODE VISCODE2 USERDATE USERDATE2 EXAMDATE DXCHANGE DXCURREN DXCONV DXCONTYP DXREV DXNORM DXNODEP DXMCI DXMDES DXMPTR1 DXMPTR2 DXMPTR3 DXMPTR4 DXMPTR5 DXMPTR6 DXMDUE DXMOTHET DXMOTHSP DXDSEV DXDDUE DXAD DXADES DXAPP DXAPROB DXAMETASP DXAOTHRSP DXAPOSS DXAATYSP DXAMETSP DXAOTHSP DXPARK DXPARKSP DXPDES DXPCOG DXPATYP DXPOTHSP DXDEP DXDEPSP DXOTHDEM DXODES DXOOTHSP DXCONFID DIAGNOSIS update_stamp

summary
ADNI1ADNIGOADNI2ADNI3Total
D14201683//4884
D238714755666214912161
D338684755671214212156
D1: Diagnosis and Symptoms Checklist [ADNI1,GO,2]
D2: Diagnostic Summary - Baseline Changes [ADNI1,GO,2,3]
D3: Diagnostic Summary [ADNI1,GO,2,3]
ALL PET Image Acquisition
FDG PET Scan Information [ADNI1]

1957 items = bl (431) + m06 (402) + m12 (376) + m18 (173) + m24 (330) + m36 (224) + m48 (21)

ID RID SITEID VISCODE USERDATE USERDATE2 RECNO PMCONDCT PMREASON PMOTHSPE PMCONSENT PMTYPE EXAMDATE PMSCANNER PMGEMODEL PMSIEMENS PMSIEMODEL PMPHILLIPS PMPHMODEL PMQCTIME PMBLTIME PMBLGLUC PMFDGTIME PMFDGDOS PMFDGVOL PMINJTIME PMGLUCOM PMSCTIME PMSCANCOM PMVARIAT PMVARSP PMPROTID PMFRAME PMDEVIAT PMMOTION PMMOTSP PMMALFUN PMMALSP PMOTHER PMOTHSP PMRECON PMSUBSET PMSUBSPE PMITERAT PMITERSPE PMRAMLA PMFILTER PMMODEON PMSMOOTH PMDECAY PMSCATTR PMATTEN PMWITHDRAW PMWITHTIME PMCOUNT PMCOUNTIME PMBG1TIME PMBG1VOL PMBG1DUR PMBG1COUNT PMP1DTIME PMP1CTIME PMP1BGL PMP1VOL PMP1DUR PMP1COUNT PMP2DTIME PMP2CTIME PMP2BGL PMP2VOL PMP2DUR PMP2COUNT PMP3DTIME PMP3CTIME PMP3BGL PMP3VOL PMP3DUR PMP3COUNT PMP4DTIME PMP4CTIME PMP4BGL PMP4VOL PMP4DUR PMP4COUNT PMP5DTIME PMP5CTIME PMP5BGL PMP5VOL PMP5DUR PMP5COUNT PMBG2CTIME PMBG2VOL PMBG2DUR PMBG2COUNT PMPIPVOL PMPIPVOLSP PMCTIME PMCTIMESP PMPHACT PMPHVOL PMPHCOUNT PMALIQVOL PMALIQACT PMBLOOD PMBLOOD2 PMTRNSFR PMTRNDATE PMTRNCOM PMARCHIVE PMARCMED PMARCCOM PMLPDONE PMLPINTER update_stamp

FDG PET Scan Information [ADNIGO,2]

3376 items = bl (131) + m36 (9) + m48 (146) + m60 (56) + v03 (792) + v06 (380) + v11 (337) + v21 (819) + v31 (214) + v41(456) + v51 (36)

3376 items = ADNI2 (3034) + ADNIGO (342)

Phase ID RID SITEID VISCODE VISCODE2 USERDATE USERDATE2 RECNO PMCONDCT PMREASON PMOTHSPE RADTRACER EXAMDATE PMSCANNER PMGEMODEL PMSIEMENS PMSIEMODEL PMPHILLIPS PMPHMODEL PMQCTIME PMBLTIME PMBLGLUC PMFDGTIME PMFDGDOS PMFDGVOL PMINJTIME PMGLUCOM PMSCTIME PMSCANCOM PMVARIAT PMVARSP PMPROTID PMFRAME PMDEVIAT PMMOTION PMMOTSP PMMALFUN PMMALSP PMOTHER PMOTHSP PMRECON PMSUBSET PMSUBSPE PMITERAT PMITERSPE PMRAMLA PMSHARP PMFILTER PMMODEON PMSMOOTH PMATTEN PMTRNSFR PMTRNDATE PMTRNCOM PMARCHIVE PMARCMED PMARCCOM PMLPDONE PMLPINTER update_stamp

FDG PET Scan Information [ADNI3]

VISCODE : 539 items = bl (291) + init (248)
VISCODE2 : 539 items = bl (290+1) + m48 (27) + m60 (50) + m72 (98) + m84 (23) + m96 (5) + m120 (6) + m132 (27) + m144 (10) + m156 (2)

Phase ID RID SITEID VISCODE VISCODE2 USERDATE USERDATE2 DONE NDREASON NDSPECIFY RADTRACER SCANDATE QCTIME PMBLTIME PMBLGLUC ASSAYTIME FDGDOS INJTIME PMGLUCOM SCANTIME SCANDIFF VARIAT VARIATSPEC PROTID MOTION OTHERVAR RECON SHARP ARCHIVE LONI TRANDATE update_stamp

Summary
PET Metadata Listing [ADNI1,GO,2]
PET Scanner Smoothing Table

PET Image Analysis
BAI - PET NMRC FDG Summaries Dictionary [ADNI1,GO,2,3]

  • HCI:Hypometabolic Convergence Index 代谢下降指数
BAI - PET NMRC FDG Summaries [ADNI1,GO,2,3]

3648 items = bl (1471) + m06 (368) + m12 (339) + m18 (154) + m24 (660) + m36 (190) + m48 (127) + m60 (110) + m72 (125) + m84 (67) + m96 (14)

BAI - PET NMRC FDG Summaries Methods (PDF)

NYU FDG-PET Hippocampus (pons normalized) Dictionary [ADNI1]

NYU FDG-PET Hippocampus (pons normalized) [ADNI1]


612 items

NYU FDG-PET Hippocampus (pons normalized) Methods (PDF)

Hippocampal Glucose Metabolism Sampling, the NYU HIPMASK
NYU HIPMASK海马葡萄糖代谢采样

UC Berkeley - FDG Analysis Dictionary [ADNI1,GO,2]

UC Berkeley - FDG Analysis [ADNI1,GO,2,3]

VISCODE2:18170 items = bl (7280 + 5) + m06 (1840) + m12 (1695) + m18 (770) + m24 (3300) + m36 (950) + m48 (635) + m60 (550) + m72 (625) + m84 (335) + m96(70)→ 18170/5 = 3634

UC Berkeley - FDG Methods (PDF)

UU - PET AD Subjects Cerebral Metabolic Pattern of Glucose Uptake Consistent with Frontotemporal Dementia in Baseline FDG-PET (PDF)
PET Image Quality
FDG PET QC [ADNI3]

VISCODE : 346 items = bl (207) + init (139)
VISCODE2 : 346 items = bl (206+1) + m48 (5) + m60 (31) + m72 (59) + m84 (16) + m96 (3) + m120 (7) + m132 (13) + m144 (4) + m156 (1)

Phase ID RID SITEID VISCODE VISCODE2 USERDATE USERDATE2 SCANDATE LONIUPDT LONIUID REVDT ALLFRAME UNUSABL UNRSN UNRSNSPEC SCANQLTY REPROCREQ PROCERR PROCERRSPEC RESCANREQ update_stamp

初步分析
时间顺序(早-晚) :SCANDATE(ex.1/26/2017) — LONIUPDT (ex. 3/21/2017) — REVDT (ex. 3/21/2017) — USERDATE = USERDATE2 (ex. 3/24/2017)


写在最后:若本文对您有所帮助,请点个赞啦 ٩(๑•̀ω•́๑)۶

### 处理和使用特定日期命名的 CSV 文件 对于名为 `DXSUM_PDXCONV_ADNIALL_24Sep2023.csv` 和 `ANDI1GO_postprocess_10_09_2023.csv` 的文件,以及另一个文件 `ADNIMERGE_05Jan2024.csv`,可以按照以下方法进行数据处理和分析。 #### 数据加载 为了读取这些 CSV 文件并将其转换为可操作的数据结构,通常会使用 Python 中的 Pandas 库。以下是加载 CSV 文件的一个基本示例: ```python import pandas as pd # 加载第一个文件 dxsum_file_path = "DXSUM_PDXCONV_ADNIALL_24Sep2023.csv" dxsum_data = pd.read_csv(dxsum_file_path) # 加载第二个文件 andi_file_path = "ANDI1GO_postprocess_10_09_2023.csv" andi_data = pd.read_csv(andi_file_path) # 加载第三个文件 adni_merge_file_path = "ADNIMERGE_05Jan2024.csv" adni_merge_data = pd.read_csv(adni_merge_file_path) ``` 上述代码片段展示了如何通过指定路径来加载三个不同的 CSV 文件到内存中的 DataFrame 对象中[^5]。 #### 数据预处理 一旦数据被成功加载至 DataFrame 结构中,则需执行一系列清理与准备步骤以便后续分析工作顺利开展。这可能涉及缺失值填充、重复记录移除以及其他标准化过程等操作。 - **检查缺失值** 使用 `.info()` 或者 `.isnull().sum()` 方法可以帮助识别每列中存在的任何空白项。 ```python print(dxsum_data.isnull().sum()) ``` - **删除冗余条目** 如果发现某些行完全相同或者部分字段一致而其余不同之处无意义时,应该考虑去除这类多余的信息。 ```python adni_merge_data.drop_duplicates(inplace=True) ``` - **重命名列名以提高可读性** 当原始表头不够直观清晰时,重新定义它们有助于增强理解程度。 ```python dxsum_data.rename(columns={"OldColumnName": "NewColumnName"}, inplace=True) ``` 以上所有动作均应依据实际业务需求灵活调整实施策略[^6]。 #### 合并与连接多个表格 如果几个独立来源之间存在关联关系(比如共同键),那么可以通过 merge 函数实现横向拼接;反之若是单纯追加新观测单位则适合采用 concat 方式纵向堆叠起来形成更大规模的整体视图。 ```python merged_df = pd.merge(dxsum_data, adni_merge_data, on='CommonKeyColumn') combined_df = pd.concat([andi_data, merged_df], ignore_index=True) ``` 这里假设两个主要对象间共享某一公共属性作为匹配条件完成融合任务的同时还将另一组额外补充资料附加进来构建最终综合版本[^7]。 --- ###
评论 33
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

爱学习的卡比兽

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值