V2.3 保留后台导入任务状态，并为常用格式启用真正的增量解析；审计事件会记录导入、解析和后续导出节点。V2.3 keeps import-job state and real incremental parsing for common formats; audit events record import, parsing, and later export milestones.
文件信息File Details
文件格式Format:-
格式来源Source:-
总记录数Records:0
检测到的列Detected columns:
字段映射Field Mapping
数据预览Preview
STEP 02
设置筛选标准Set Screening Criteria
配置纳入、排除与保护条件，作为初筛执行标准。Configure inclusion, exclusion, and protection rules for the screening summary.
配置本轮筛选标准Configure screening criteria
优先完成纳入词、排除词与研究边界设置，再生成初筛结果。Complete inclusion, exclusion, and scope settings before generating the screening summary.
先写纳入词，再补排除词Start with inclusion, then exclusion
确认年份、语言和必填字段Check year, language, and required fields
需要复用时再切到 YAMLUse YAML only when needed
每行一个关键词；留空则不过滤。One keyword per line; leave blank to skip inclusion filtering.
允许语言Allowed languages
必填字段（至少一项）Required fields (at least one)
20%
该参数仅用于预估工作量，不会自动删除任何全文文献。This value is only used for planning and will not automatically exclude any full-text records.
RIS
Field
说明Meaning
TY
Type
AU
Authors
T1
Title
AB
Abstract
PY
Year
DO
DOI
BibTeX
Field
说明Meaning
title
Title
author
Authors
journal
Journal
year
Year
doi
DOI
abstract
Abstract
ENW
Field
映射Maps to
T1
title
AB
abstract
PY
year
AU
authors
T2
journal
DO
doi
RDF
Field
映射Maps to
dc:title
title
dc:creator
authors
dcterms:issued
year
bib:publicationTitle
journal
dcterms:abstract
abstract
标准预览Criteria Preview
STEP 03
初筛结果Screening Summary
核对筛选统计与流程图预览，再决定是否进入全文复核。Review the screening counts and PRISMA preview before continuing to full-text review.
确认结果后进入下一步。Review results before proceeding.
如果统计或保留结果异常，建议返回规则配置调整后再运行。If the counts or retained set look abnormal, return to the rules and rerun.
先看统计，再看图Check stats first
异常时返回规则配置Return to rules if needed
确认后进入全文复核Then enter full-text review
项目会话Project Session
保存当前工作区状态，便于恢复、共享与继续处理。Save the current workspace state so it can be resumed, shared, or continued later.
这里只调整 PRISMA 流程图的呈现风格，不会改变去重、筛选或最终统计结果。This only changes the PRISMA diagram presentation. It does not alter deduplication, screening, or final counts.
导出预览Export Preview
这里是预演环节：可以先单独检查纳入表、排除表、SVG 或报告是否符合预期。Use this area to spot-check included/excluded tables, SVG, or report outputs before the final export step.
如果你还在核查结果是否合理，建议先看统计卡片和 PRISMA 预览；Step 6 再进行正式导出。If you are still validating the result, review the stats cards and PRISMA preview first. Use Step 6 for the formal export.
当前主题：Current theme:柔和版
PRISMA 2020 流程图预览PRISMA 2020 Preview
STEP 04
全文复核Full-text Review
处理全文判断、排除理由与协作复核结果。Review full-text decisions, exclusion reasons, and collaboration outcomes.
开始全文复核Start full-text review
对保留下来的文献进行全文判断，并记录排除理由。Review retained records at full-text level and document exclusion reasons.
先看协作状态Check collaboration state
逐条记录排除理由Record reasons clearly
确认后再生成最终结果Generate results after review
协作状态Collaboration Status
主审查员Primary Reviewer
加载中...
副审查员Secondary Reviewer
加载中...
一致性分析Agreement Analysis
审核说明Review Guidance
需要人工审核的全文文献：Full-text records requiring review:0
研究设计识别Study Design Identification
“自动标注研究设计”的意思是：系统会扫描标题和摘要里的关键词，尝试给每条文献补上可能的研究类型标签，例如 RCT、队列研究、系统综述。Auto-label Study Designs scans titles and abstracts and adds likely study design tags, such as RCT, cohort study, or systematic review.
这只是辅助标注，不会自动决定纳入、排除或修改你的审核决定，仍需你在人工复核时确认。This is support only. It does not automatically include, exclude, or override your review decisions.
批量操作Batch Actions
快捷键Keyboard Shortcuts
1-6快速选择原因Choose reason / Space保留Keep / ↑↓切换文献Navigate records
排除原因模板Exclusion Templates
排除统计Exclusion Statistics
获取全文Full-text obtained
0
已排除Excluded
0
最终纳入Included
0
排除率Exclusion rate
0%
STEP 05
质量评价Quality Assessment
质量评价是在确定纳入研究后，对每篇研究的方法学质量、偏倚风险和证据可信度做一致记录。它帮助判断结论有多可靠，也让后续 PRISMA 报告和审计追溯更清楚。Quality assessment records each included study's methodological quality, risk of bias, and evidence certainty in a consistent way. It helps reviewers judge how reliable the findings are and keeps later PRISMA reporting auditable.
建立质量评价队列Prepare the quality queue
先为最终纳入研究生成质量评价条目，确认研究设计、工具族和证据等级基线，再进入正式导出。Seed quality-assessment entries for included studies, confirm design/tool/evidence defaults, then move to export.
先生成质量评价条目Seed assessment entries
再核对研究设计和工具族Review study design and tool family
最后进入 Step 6 导出Continue to Step 6 for export
当前版本如何处理质量评价How this version handles quality assessment
V2.4 会为最终纳入研究生成结构化质量评价记录、evidence_table.csv 和 GRADE summary 骨架，包含研究设计、正式模板、建议工具族、领域条目、PICOS、效应量和证据等级基线。研究者仍需人工核对和确认最终 GRADE；真实 AI provider 不会默认接入。V2.4 creates structured quality appraisal records, evidence_table.csv, and a GRADE summary scaffold for included studies, including study design, template, recommended tool family, domain items, PICOS, effect fields, and baseline evidence level. Reviewers still confirm final GRADE manually; no real AI provider is connected by default.
待评价研究Included studies
0
未开始Not started
0
进行中In progress
0
已完成Completed
0
研究队列Study queue
证据概览Evidence overview
下一步Next step
这一步已经把纳入研究、建议研究设计、质量工具和证据基线稳定存下来。后续完整表单会直接挂接到这些正式记录上继续扩展。This step now persists included studies together with suggested design, tool, and baseline evidence so later domain forms can extend real records directly.
STEP 06
结果导出Export Deliverables
汇总最终统计并导出图表、结果表与筛选报告。Summarize final counts and export diagrams, study tables, and the screening report.
导出最终交付物Export final deliverables
确认统计结果后，导出图表、结果表与筛选报告。Confirm the final counts, then export diagrams, study tables, and the screening report.
先确认最终统计Check final counts
再选 PRISMA 主题Choose PRISMA theme
最后导出全部结果Export all deliverables
筛选已完成Screening Complete
现在可以把这一轮筛选结果整理成图、表和报告，作为正式项目输出。You can now turn this review cycle into final diagrams, tables, and reports for formal project output.
数据库识别Database identified
0
其他来源Other sources
0
去重后After deduplication
0
标题/摘要筛选Title/abstract screened
0
标题/摘要排除Title/abstract excluded
0
全文评估Full-text assessed
0
全文排除Full-text excluded
0
最终纳入Included
0
最终图表主题Final Diagram Theme
导出文件Export Files
审计包Audit Package
审计包记录完整的筛选过程证据，全部在浏览器本地生成，不会自动上传。可用于方法学附录、审稿回复或研究团队内部分享。The audit package captures the full screening evidence chain. Everything is generated locally in your browser and never uploaded automatically. Use it for methodology appendices, reviewer responses, or team sharing.
每个文件是什么？What is each file?
文件File
用途Purpose
project_manifest.json
项目名称、PRISMA 版本、AI 模式、应用版本等基本信息Project name, PRISMA version, AI mode, app version and other metadata
events.jsonl
按时间顺序记录导入、去重、筛选、复核、质量评价和导出的所有事件Chronological log of all import, dedup, screening, review, quality, and export events
screening_decisions.csv
每条文献的筛选阶段、纳入/排除决定、排除理由、审稿人、冲突状态和 AI 使用情况Per-record screening stage, include/exclude decision, exclusion reason, reviewer, conflict status, and AI usage
exclusion_reasons.csv
标准排除理由分类和每类计数，适合放在 PRISMA 图下方Standardized exclusion reason categories and counts, suitable for PRISMA diagram annotation
prisma_counts.json
从审计数据重新计算的 PRISMA 流程图数字，可直接用于 PRISMA 2020 checklistPRISMA flow diagram counts recalculated from audit data, ready for the PRISMA 2020 checklist
quality_appraisal.csv
V2.4 质量评价导出：每条纳入研究的模板、工具族、领域判断、支持引文、审稿备注和状态V2.4 quality appraisal export: template, tool family, domain judgement, supporting quote, reviewer note, and status for included studies
evidence_table.csv
V2.4-beta 证据表导出：每条纳入研究的 PICOS、效应量、质量判断、证据确定性和备注；GRADE 仍需人工确认V2.4-beta evidence table export: PICOS, effect measure, quality judgement, certainty of evidence, and notes for each included study; GRADE remains human-confirmed
grade_summary.csv
V2.4 GRADE 摘要骨架：按 outcome / PICOS 聚合研究、效应摘要、质量判断摘要和基线确定性；最终 GRADE 和降级理由保留给人工确认V2.4 GRADE summary scaffold: groups studies by outcome / PICOS with effect summary, quality judgement summary, and baseline certainty; final GRADE and downgrade reasons stay human-confirmed
audit_summary.md
可读的审计摘要：事件数量、决策数量、PRISMA 计数、AI 模式、数据边界和风险提示Human-readable audit summary: event counts, decision counts, PRISMA counts, AI mode, data boundary, and risk notes
ai_usage_registry.json
AI 模式、服务边界、允许阶段、数据边界和用户确认记录；这是配置证据，不是最终筛选决定AI mode, provider, allowed stages, data boundary, and user acknowledgement log; configuration evidence, not a final decision ledger
ai_suggestions.jsonl
AI 建议、理由、置信度、输入指纹、人工处理结果、复核时间、人工改写字段、关联的人类筛选决定和计数边界；被拒绝的建议不会进入 PRISMA 计数AI suggestions, rationale, confidence, input hashes, human handling outcomes, reviewed_at, human edit fields, linked human decisions, and prisma_count_boundary; rejected suggestions do not enter PRISMA counts
PRISMA_TRAICE_REPORT.md
V2.3 透明报告：说明是否使用 AI、如何使用、No-AI 状态，以及“最终计数来自人工 ScreeningDecision”的边界V2.3 transparency report covering AI usage, No-AI state, and the boundary that final counts come from human ScreeningDecision records
PRISMA-trAIce 是面向“系统综述中使用 AI”的透明报告要求：需要说明 AI 用在什么环节、是否影响筛选决定，以及人工如何复核。PRISMA-trAIce is a transparency checklist for AI use in systematic reviews: where AI is used, whether it affects screening decisions, and how humans review it.
为什么做Why让审稿人和团队看清 AI 只提供建议，不直接改变最终计数。Reviewers can see that AI only suggests and never changes final counts directly.
怎么做How记录 AI 模式、建议日志、人工接受/拒绝/改写，以及关联的人类筛选决定。Record AI mode, suggestion logs, human accept/reject/edit actions, and linked human screening decisions.
当前不连接真实 AI 服务，只用本地示例建议验证“AI 建议必须由人工确认”的审计闭环。This stage does not connect to a real AI provider; local mock suggestions verify the human-confirmation audit loop.
服务接入边界Provider boundary已准备服务抽象、摘要指纹和脱敏记录；真实服务请求仍然关闭，不会自动发送文献内容或密钥。Provider abstraction, hashes, and redacted trace records are ready; real API dispatch remains disabled and never sends record content or API keys automatically.