Buckets:
| { | |
| "n": 5, | |
| "note": "论文口径 vs 本项目做法；引文为短句摘录用于方法核对", | |
| "items": [ | |
| { | |
| "source_id": "steinmetz2019", | |
| "topic": "单元入选标准", | |
| "paper_quote": "Units containing low-amplitude spikes, spikes with inconsistent waveform shapes, and/or refractory period contamination were labeled as 'multi-unit activity' and not included for further analysis.", | |
| "what_we_did": "桶内 npz 是作者 curation 之后的产物；我们在此之上又加了自己的 SNR>=3 过滤（build_l2_region.py）", | |
| "impact": "我们的额外过滤与原文口径不一致，但方向是更严格；未能量化其剔除量（桶内只有分箱计数，无 SNR 逐单元表）", | |
| "severity": "low" | |
| }, | |
| { | |
| "source_id": "steinmetz2019", | |
| "topic": "每区最少单元数", | |
| "paper_quote": "Neurons were only included for further analysis when at least 13 neurons passing the above criteria were identified as coming from the same brain region, in the same experiment.", | |
| "what_we_did": "MIN_UNITS = 8", | |
| "impact": "实测：>=8 得 276 个 cell / 68 区，>=13 得 267 / 67；只有 9 个 cell（3.3%）落在 8-12 区间；两阈值下区中位的 Spearman = +0.966。结论不受影响", | |
| "severity": "none", | |
| "evidence": "framework/qc_impact.json" | |
| }, | |
| { | |
| "source_id": "dandi000070", | |
| "topic": "单元是否 single-unit", | |
| "paper_quote": "These array-based datasets involved, respectively, 146 and 218 single isolations (each a mix of single and multi-unit isolations).", | |
| "what_we_did": "桶内每个片段 192 个单元 = 96 电极 × 2 阵列，即每电极恰好 1 个单元", | |
| "impact": "这批 'unit' 不是精心 curation 的 single unit，更可能是每电极的代表性信号（含 multi-unit）。这解释了放电率偏高（27–29 Hz）与 ISI 违反率偏高，也意味着 rho_half / PMd-M1 对比反映的是电极级群体信号，不是严格单细胞", | |
| "severity": "high", | |
| "evidence": "framework/isi_quality2.json" | |
| }, | |
| { | |
| "source_id": "dandi000070", | |
| "topic": "PMd / M1 解剖边界", | |
| "paper_quote": "While there are important differences in their average response properties (delay period activity is more common in PMd), these differences are far from absolute: M1-like neurons are frequently found in caudal PMd and vice versa.", | |
| "what_we_did": "按 DANDI electrodes 表把前 96 电极记为 PMd、后 96 记为 M1，并按此做区级对比（PMd > M1 1.20×，64/64 同向）", | |
| "impact": "作者本人承认两群体会混。每个 96 通道阵列上都有另一区的神经元，所以 1.20× 的差异是**电极位置**的差异，不能直接读成'区类型'的差异；同向 100% 说明位置效应稳定，不说明细胞类型不同", | |
| "severity": "medium" | |
| }, | |
| { | |
| "source_id": "dandi000130", | |
| "topic": "DMFC 是什么", | |
| "paper_quote": "We recorded from 617 and 741 units in monkey H and G, respectively in the dorsomedial frontal cortex (DMFC), comprising supplementary eye field (SEF), presupplementary motor area (Pre-SMA), and dorsal portion of the supplementary motor area (SMA).", | |
| "what_we_did": "72 个电极统一标 region='DMFC'，区级不变量按单一区计算", | |
| "impact": "DMFC 是三个亚区的复合（SEF / Pre-SMA / dorsal SMA）。我们标的 DMFC 属于粗粒度标签，其区级 rho_half 是三个亚区的混合，不能与小鼠单区直接对比", | |
| "severity": "medium" | |
| } | |
| ] | |
| } |
Xet Storage Details
- Size:
- 3.56 kB
- Xet hash:
- 3c3df9b5017159415ccb53d00fdeb8cf63ce59ece2d3506c64f500b5fcc18a9f
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.