OSCR

How Long Is the Brain Perfusable After Global Ischemia? A Systematic Review.

Code ↔ Paper

6 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 6 matches
  1. [1] § 3. Results and Discussion › 3.1. Study Selection ↔ ischemia_review_code.R, lines 118–139 · score 0.88 · xenon enhanced CT, Xe clearance, Xe CT, India ink, antipyrine, autoradiography
  2. [2] § 3. Results and Discussion › 3.1. Study Selection ↔ ischemia_review_code.R, lines 118–139 · score 0.81 · laser Doppler, laser speckle, functional capillary, intravital, microscopy, sidestream
  3. [3] § 2. Review Methods › 2.5. Data Extraction ↔ ischemia_review_code.R, lines 506–641 · score 0.69 · adequacy criterion, LOESS smoothers, pure blood, adequate perfusion, carbon black, interactive
  4. [4] § 3. Results and Discussion › 3.4. Qualitative Reperfusion Outcomes ↔ ischemia_review_code.R, lines 690–717 · score 0.68 · Wilcoxon rank sum, reactive hyperemia, absent, hypoperfusion, median, qualitative
  5. [5] § 3. Results and Discussion › 3.9. Study Quality ↔ ischemia_review_code.R, lines 155–187 · score 0.57 · publication year, quality scores, Blinding, Spearman, temperature, ischemia duration
  6. [6] § 2. Review Methods › 2.6. Data Synthesis ↔ ischemia_review_code.R, lines 316–377 · score 0.55 · pure blood, carbon black, quality assessments, LOESS, Smoothing, perfusion quality

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

R · 717 lines · 23 KB · no license · 6 matches

  1. library(dplyr)
  2. library(tidyr)
  3. library(ggplot2)
  4. library(patchwork)
  5. library(stringr)
  6. library(plotly)
  7. library(htmlwidgets)
  8. library(htmltools)
  9. library(glue)
  11. load_tsv <- function(path, encoding = "") {
  12. d <- read.delim(
  13. path,
  14. sep = "\t",
  15. header = TRUE,
  16. stringsAsFactors = FALSE,
  17. check.names = FALSE,
  18. na.strings = c("", "NA"),
  19. fileEncoding = encoding
  20. )
  21. d[-1, ]
  22. }
  24. ext <- load_tsv(
  25. "Data Extraction Fixed - Data Extraction2.tsv",
  26. encoding = "UTF-16LE"
  27. )
  29. sq <- load_tsv(
  30. "Data Extraction Fixed - Study Quality.tsv",
  31. encoding = "UTF-8"
  32. )
  34. parse_pct <- function(x) {
  35. x <- trimws(x); x[x %in% c("NR", "Cannot convert")] <- NA
  36. suppressWarnings(as.numeric(sub("%$", "", sub("^~", "", x))))
  37. }
  38. parse_brains <- function(x) {
  39. x <- trimws(x); x[x %in% c("NR", "Cannot convert", "N/A")] <- NA
  40. suppressWarnings(as.numeric(sub("^([0-9.]+).*", "\\1", sub("^~", "", x))))
  41. }
  42. parse_num <- function(x) suppressWarnings(as.numeric(sub("^~", "", trimws(x))))
  43. parse_n <- function(x) {
  44. x <- sub("^~", "", trimws(x))
  45. suppressWarnings(as.numeric(sub("^(\\d+).*", "\\1", x)))
  46. }
  48. # Unique per-paper key (PMID > DOI > Authors+Year label). paste(Authors, Year)
  49. # alone is NOT unique here (3 Cantu et al. 1969 papers, 2 Kågström et al. 1983,
  50. # 2 Fischer et al. 1996), so anything grouped on that string alone silently
  51. # collapses 60 papers to 56.
  52. make_study_key <- function(authors, year, pmid = NA, doi = NA, link = NA) {
  53. n <- length(authors)
  54. pmid <- rep_len(pmid, n)
  55. doi <- rep_len(doi, n)
  56. link <- rep_len(link, n)
  58. pmid_from_link <- ifelse(
  59. !is.na(link),
  60. sub(".*pubmed\\.ncbi\\.nlm\\.nih\\.gov/(\\d+).*", "\\1", link),
  61. NA
  62. )
  63. pmid_from_link[pmid_from_link == link] <- NA
  65. doi_from_link <- ifelse(
  66. !is.na(link) & is.na(pmid_from_link),
  67. sub(".*doi\\.org/(.+)$", "\\1", link),
  68. NA
  69. )
  70. doi_from_link[doi_from_link == link] <- NA
  72. pmid_val <- ifelse(!is.na(pmid) & pmid != "Not found.", pmid, pmid_from_link)
  73. doi_val <- ifelse(!is.na(doi) & doi != "Not found.", doi, doi_from_link)
  75. ifelse(
  76. !is.na(pmid_val), paste0("PMID:", pmid_val),
  77. ifelse(!is.na(doi_val), paste0("DOI:", tolower(trimws(doi_val))),
  78. paste0("LABEL:", authors, "_", year))
  79. )
  80. }
  82. d <- ext |>
  83. mutate(
  84. pct_brain = parse_pct(`Converted % Brain Perfused`),
  85. pct_brains = parse_brains(`% Brains Adequately Perfused`),
  86. isch = parse_num(`Ischemia Duration (min)`),
  87. n_arm = parse_n(`Sample Size (n)`),
  88. blood = `Blood vs. Non-Blood`,
  89. study = make_study_key(`Study Authors`, Year, link = Link)
  90. ) |>
  91. filter(isch <= 60)
  93. n_studies_check <- ext |>
  94. mutate(study = make_study_key(`Study Authors`, Year, link = Link)) |>
  95. pull(study) |> n_distinct()
  96. stopifnot(n_studies_check == 60)
  98. # ===========================================================
  99. # 0. Descriptive statistics
  100. # ===========================================================
  102. ext <- ext %>%
  103. mutate(species_clean = case_when(
  104. str_detect(`Species / Strain`, regex("^cat", ignore_case = TRUE)) ~ "cat",
  105. str_detect(`Species / Strain`, regex("^dog", ignore_case = TRUE)) ~ "dog",
  106. str_detect(`Species / Strain`, regex("^human", ignore_case = TRUE)) ~ "human",
  107. str_detect(`Species / Strain`, regex("^(pig|piglet)", ignore_case = TRUE)) ~ "pig",
  108. str_detect(`Species / Strain`, regex("^rabbit", ignore_case = TRUE)) ~ "rabbit",
  109. str_detect(`Species / Strain`, regex("^rat", ignore_case = TRUE)) ~ "rat",
  110. TRUE ~ NA_character_
  111. ))
  113. table(ext$species_clean)
  115. dur <- as.numeric(ext$`Ischemia Duration (min)`)
  116. median(dur[dur <= 120])
  118. fam <- function(x) {
  119. x <- tolower(x)
  120. dplyr::case_when(
  121. grepl("carbon|india ink|colloidal", x) ~ "Carbon black/India ink",
  122. grepl("microsphere", x) ~ "Microsphere CBF",
  123. grepl("antipyrine|autoradiograph", x) ~ "Antipyrine autoradiography",
  124. grepl("xe-ct|xe/ct|xenon-enhanced ct", x) ~ "Stable Xe-CT",
  125. grepl("xe |133xe|133-xenon|xenon clearance", x) ~ "133Xe clearance",
  126. grepl("fitc-albumin", x) ~ "FITC-albumin microscopy",
  127. grepl("functional capillary|fitc-dextran|sdf|sidestream", x) ~ "Intravital FCD/SDF",
  128. grepl("laser doppler|laser speckle", x) ~ "Laser Doppler/speckle",
  129. grepl("multiphoton", x) ~ "Multiphoton capillary RBC",
  130. grepl("angiograph", x) ~ "Angiography",
  131. grepl("asl-mri|arterial spin", x) ~ "ASL-MRI",
  132. grepl("pet ", x) ~ "PET CBF",
  133. grepl("h2 clearance|hydrogen clearance", x) ~ "H2 clearance",
  134. grepl("rbc suspension|benzidine", x) ~ "RBC/benzidine",
  135. grepl("thermocouple", x) ~ "Thermocouple",
  136. grepl("0-3 scale|semi-quantitative|qualitative multi-modal|histological grading", x) ~ "Qualitative grading",
  137. TRUE ~ paste0("UNCLASSIFIED: ", x)
  138. )
  139. }
  141. ext$method_family <- fam(ext$`Perfusion Quality Assessment Method`)
  142. sort(table(ext$method_family), decreasing = TRUE)
  144. # ===========================================================
  145. # 1. Overall dose-response: % brain perfused vs ischemia
  146. # ===========================================================
  147. cat("All cases: "); print(cor.test(d$isch, d$pct_brain, method = "spearman", exact = FALSE))
  148. cat("All cases: "); print(cor.test(d$isch, d$pct_brains, method = "spearman", exact = FALSE))
  150. # ===========================================================
  151. # 2. Study quality: distribution and simple correlations
  152. # ===========================================================
  153. cat("\n--- Study quality ---\n")
  154. yes_like <- function(x) grepl("^Yes", x, ignore.case = TRUE)
  155. sq2 <- sq |>
  156. mutate(
  157. q_isch = as.integer(yes_like(`Ischemia Duration Clearly Defined?`)),
  158. q_n = as.integer(yes_like(`Sample Size Explicitly Reported?`)),
  159. q_temp = as.integer(yes_like(`Temperature During Ischemia Reported?`)),
  160. q_method = as.integer(yes_like(`Perfusion Outcome Reproducibly Described?`)),
  161. q_blind = as.integer(yes_like(`Blinding Used in Outcome Assessment?`)),
  162. q_incl = as.integer(yes_like(`Inclusion/Exclusion Criteria Stated?`)),
  163. quality = q_isch + q_n + q_temp + q_method + q_blind + q_incl,
  164. study = make_study_key(`Study Authors`, Year, pmid = PMID, doi = DOI),
  165. year_num = suppressWarnings(as.numeric(Year))
  166. )
  168. sq_one <- sq2 |> group_by(study) |>
  169. summarise(quality = max(quality),
  170. year_num = first(year_num),
  171. across(starts_with("q_"), max), .groups = "drop")
  173. stopifnot(nrow(sq_one) == 60)
  175. cat("Item-level 'Yes' rates:\n")
  176. sq_one |> summarise(across(starts_with("q_"), mean)) |> print()
  177. cat("\nQuality score distribution (per study):\n"); print(table(sq_one$quality))
  179. cat("\nQuality vs publication year (per study):\n")
  180. print(cor.test(sq_one$year_num, sq_one$quality, method = "spearman", exact = FALSE))
  182. per_study <- d |> filter(!is.na(pct_brain)) |>
  183. group_by(study) |>
  184. summarise(mean_pct = mean(pct_brain), mean_isch = mean(isch), .groups = "drop") |>
  185. left_join(sq_one |> select(study, quality), by = "study")
  186. cat("\nQuality vs mean ischemia duration (per study):\n")
  187. print(cor.test(per_study$quality, per_study$mean_isch, method = "spearman", exact = FALSE))
  189. # ===========================================================
  190. # 2b. Quality vs perfusate type
  191. # ===========================================================
  192. get_mode <- function(x) {
  193. x <- x[!is.na(x)]
  194. if (length(x) == 0) return(NA)
  195. names(sort(table(x), decreasing = TRUE))[1]
  196. }
  198. study_cov <- ext |>
  199. mutate(
  200. study = make_study_key(`Study Authors`, Year, link = Link),
  201. isch = parse_num(`Ischemia Duration (min)`),
  202. blood = `Blood vs. Non-Blood`
  203. ) |>
  204. filter(isch <= 60) |>
  205. group_by(study) |>
  206. summarise(blood_mode = get_mode(blood), .groups = "drop") |>
  207. left_join(sq_one |> select(study, quality), by = "study")
  209. cat("\nQuality vs blood/non-blood:\n")
  210. print(kruskal.test(quality ~ blood_mode, data = study_cov))
  211. print(study_cov |> group_by(blood_mode) |> summarise(mean_q = mean(quality), n = n()))
  213. # ===========================================================
  214. # Plots
  215. # ===========================================================
  217. classify_perf <- function(blood_raw, method) {
  218. carbon <- grepl("carbon", method, ignore.case = TRUE)
  219. case_when(
  220. blood_raw == "Blood" ~ "Pure blood",
  221. blood_raw == "Mixed" ~ "Mixed blood",
  222. blood_raw == "Non-Blood" & carbon ~ "Carbon black",
  223. blood_raw == "Non-Blood" & !carbon ~ "Other non-blood",
  224. TRUE ~ NA_character_
  225. )
  226. }
  227. d <- d |>
  228. mutate(perfusate = classify_perf(blood, `Perfusion Quality Assessment Method`),
  229. perfusate = factor(perfusate,
  230. levels = c("Pure blood","Mixed blood",
  231. "Carbon black","Other non-blood")))
  233. perf_cols <- c(
  234. "Pure blood" = "#0072B2", # blue
  235. "Mixed blood" = "#CC79A7", # magenta
  236. "Carbon black" = "#D55E00", # orange
  237. "Other non-blood" = "#666666" # neutral gray
  238. )
  241. loess_grid <- function(df, ycol, span = 2) {
  243. df |>
  244. filter(!is.na(.data[[ycol]]), !is.na(isch)) |>
  245. group_by(perfusate) |>
  246. group_modify(~{
  248. fit <- loess(
  249. formula = reformulate("log10(isch)", response = ycol),
  250. data = .x,
  251. span = span,
  252. degree = 1
  253. )
  255. xgrid <- exp(seq(
  256. log(min(.x$isch)),
  257. log(max(.x$isch)),
  258. length.out = 200
  259. ))
  261. pred <- predict(
  262. fit,
  263. newdata = data.frame(isch = xgrid)
  264. )
  266. tibble(
  267. isch = xgrid,
  268. fit = pmin(100, pmax(0, pred))
  269. )
  270. }) |>
  271. ungroup()
  272. }
  274. make_plot <- function(df, ycol, ylab, title_txt) {
  276. df_clean <- df |>
  277. filter(!is.na(.data[[ycol]]), !is.na(perfusate)) |>
  278. mutate(
  279. perfusate = droplevels(perfusate),
  281. # Unknown n is plotted at the minimum point size
  282. n_plot = ifelse(is.na(n_arm), 1, n_arm),
  284. # Only relevant to Figure 3
  285. criterion_source = case_when(
  287. # Explicitly identified as paper-defined in the extraction
  288. grepl(
  289. "paper-defined",
  290. `% Brains Adequately Perfused`,
  291. ignore.case = TRUE
  292. ) |
  293. grepl(
  294. "paper-defined",
  295. `% Brains Quote`,
  296. ignore.case = TRUE
  297. ) ~ "Study-defined criterion",
  299. # Ames bloodless-ischemia experiments:
  300. # adequacy based directly on the paper's qualitative filling description
  301. `Study Authors` == "Ames et al." &
  302. grepl(
  303. "complete filling|poor filling",
  304. `% Brains Quote`,
  305. ignore.case = TRUE
  306. ) ~ "Study-defined criterion",
  308. # Kågström: presence/absence of no-reflow was the study's outcome
  309. `Study Authors` == "Kågström et al." &
  310. grepl(
  311. "no-reflow|no no-reflow",
  312. `% Brains Quote`,
  313. ignore.case = TRUE
  314. ) ~ "Study-defined criterion",
  316. # Dietrich: study's own qualitative perfusion-deficit classification
  317. `Study Authors` == "Dietrich et al." &
  318. grepl(
  319. "severe perfusion deficits",
  320. `% Brains Quote`,
  321. ignore.case = TRUE
  322. ) ~ "Study-defined criterion",
  324. # Everything else in Figure 3 uses the review-defined adequacy rule
  325. TRUE ~ "Review-defined criterion"
  326. ),
  328. criterion_source = factor(
  329. criterion_source,
  330. levels = c(
  331. "Review-defined criterion",
  332. "Study-defined criterion"
  333. )
  334. ),
  336. source_id = dplyr::case_when(
  337. grepl("pubmed.ncbi.nlm.nih.gov", Link) ~
  338. paste0(
  339. "PubMed PMID: ",
  340. sub(".*/([0-9]+)/?$", "\\1", Link)
  341. ),
  342. grepl("doi.org", Link) ~
  343. paste0(
  344. "DOI: ",
  345. sub(".*doi\\.org/", "", Link)
  346. ),
  347. TRUE ~ "Source available by clicking point"
  348. ),
  350. outcome_label = ifelse(
  351. is.na(`Additional Outcome Specifier`),
  352. "",
  353. `Additional Outcome Specifier`
  354. ),
  356. hover = glue::glue(
  357. "<b style='color:#0b2545'>{`Study Authors`} ({Year})</b><br>",
  358. "{outcome_label}<br><br>",
  359. "<b>{isch} min · {round(.data[[ycol]], 1)}% perfused</b><br>",
  360. "n = {ifelse(is.na(n_arm), 'not reported', n_arm)} · {`Species / Strain`}<br><br>",
  361. "<b>Perfusate:</b> {perfusate}<br>",
  362. "<b>Adequacy criterion:</b> {criterion_source}<br>",
  363. "<b>Assessment:</b> {`Perfusion Quality Assessment Method`}<br><br>",
  364. "<b>{source_id}</b><br>",
  365. "<span style='color:#777'>Click point to open source paper.</span>"
  366. )
  368. )
  370. df_smooth <- df_clean |>
  371. filter(perfusate %in% c("Pure blood", "Carbon black"))
  373. smooth_dat <- loess_grid(
  374. df_smooth,
  375. ycol,
  376. span = 2
  377. )
  379. # Figure 3: shape = adequacy criterion
  380. if (ycol == "pct_brains") {
  382. ggplot(df_clean, aes(
  383. x = isch,
  384. y = .data[[ycol]],
  385. colour = perfusate,
  386. size = n_plot,
  387. shape = criterion_source,
  388. text = hover,
  389. key = Link,
  390. customdata = Link
  391. )) +
  392. geom_jitter(width = 0.02, alpha = 0.4) +
  394. geom_line(
  395. data = smooth_dat,
  396. aes(
  397. x = isch,
  398. y = fit,
  399. colour = perfusate,
  400. group = perfusate
  401. ),
  402. linewidth = 1,
  403. inherit.aes = FALSE,
  404. show.legend = FALSE
  405. ) +
  407. scale_shape_manual(
  408. values = c(
  409. "Review-defined criterion" = 16,
  410. "Study-defined criterion" = 17
  411. ),
  412. name = "Adequacy criterion"
  413. ) +
  415. scale_colour_manual(values = perf_cols, drop = TRUE) +
  416. scale_size_continuous(
  417. range = c(1.5, 7),
  418. breaks = c(2, 5, 10, 20),
  419. limits = c(1, NA)
  420. ) +
  421. scale_x_log10(
  422. breaks = c(2.5, 5, 10, 15, 20, 30, 60),
  423. minor_breaks = NULL
  424. ) +
  425. labs(
  426. x = "Ischemia duration (min, log scale)",
  427. y = ylab,
  428. colour = "Perfusate",
  429. size = "n per arm",
  430. title = title_txt
  431. ) +
  432. coord_cartesian(ylim = c(-5, 105)) +
  433. theme_bw(base_size = 13) +
  434. theme(
  435. legend.position = "top",
  436. legend.box = "vertical",
  437. legend.margin = margin(2, 2, 2, 2),
  438. legend.spacing.y = unit(2, "pt"),
  439. plot.title = element_text(size = 16),
  440. axis.title = element_text(size = 14),
  441. axis.text = element_text(size = 12),
  442. legend.text = element_text(size = 12),
  443. legend.title = element_text(size = 13)
  444. )
  446. } else {
  448. # Figure 2: no shape encoding needed
  449. ggplot(df_clean, aes(
  450. x = isch,
  451. y = .data[[ycol]],
  452. colour = perfusate,
  453. size = n_plot,
  454. text = hover,
  455. key = Link,
  456. customdata = Link
  457. )) +
  458. geom_jitter(width = 0.02, alpha = 0.4) +
  460. geom_line(
  461. data = smooth_dat,
  462. aes(
  463. x = isch,
  464. y = fit,
  465. colour = perfusate,
  466. group = perfusate
  467. ),
  468. linewidth = 1,
  469. inherit.aes = FALSE,
  470. show.legend = FALSE
  471. ) +
  473. scale_colour_manual(values = perf_cols, drop = TRUE) +
  474. scale_size_continuous(
  475. range = c(1.5, 7),
  476. breaks = c(2, 5, 10, 20),
  477. limits = c(1, NA)
  478. ) +
  479. scale_x_log10(
  480. breaks = c(2.5, 5, 10, 15, 20, 30, 60),
  481. minor_breaks = NULL
  482. ) +
  483. labs(
  484. x = "Ischemia duration (min, log scale)",
  485. y = ylab,
  486. colour = "Perfusate",
  487. size = "n per arm",
  488. title = title_txt
  489. ) +
  490. coord_cartesian(ylim = c(-5, 105)) +
  491. theme_bw(base_size = 13) +
  492. theme(
  493. legend.position = "top",
  494. legend.box = "vertical",
  495. legend.margin = margin(2, 2, 2, 2),
  496. legend.spacing.y = unit(2, "pt"),
  497. plot.title = element_text(size = 16),
  498. axis.title = element_text(size = 14),
  499. axis.text = element_text(size = 12),
  500. legend.text = element_text(size = 12),
  501. legend.title = element_text(size = 13)
  502. )
  503. }
  504. }
  506. make_interactive <- function(p, output_file, page_title) {
  507. p_interactive <- p +
  508. ggplot2::labs(title = NULL) +
  509. ggplot2::theme(plot.title = ggplot2::element_blank())
  511. fig <- plotly::ggplotly(
  512. p_interactive,
  513. tooltip = "text",
  514. width = 1500,
  515. height = 900
  516. ) |>
  517. plotly::layout(
  518. showlegend = FALSE,
  519. autosize = TRUE,
  520. hoverlabel = list(
  521. bgcolor = "white",
  522. bordercolor = "#cccccc",
  523. font = list(color = "#111111", size = 13, family = "Arial")
  524. ),
  525. margin = list(l = 90, r = 30, t = 20, b = 80)
  526. ) |>
  527. htmlwidgets::onRender("
  528. function(el, x) {
  529. el.on('plotly_click', function(d) {
  530. var pt = d.points[0];
  531. var url = null;
  533. if (pt.customdata) {
  534. url = pt.customdata;
  535. } else if (pt.data.customdata && pt.data.customdata[pt.pointNumber]) {
  536. url = pt.data.customdata[pt.pointNumber];
  537. } else if (pt.data.key && pt.data.key[pt.pointNumber]) {
  538. url = pt.data.key[pt.pointNumber];
  539. }
  541. if (url && url !== 'NA' && url !== 'Not found.') {
  542. window.open(url, '_blank');
  543. }
  544. });
  545. }
  546. ")
  548. if (page_title == "Frequency of adequate perfusion") {
  550. legend_html <- htmltools::tags$div(
  551. style = "
  552. font-family: Arial, sans-serif;
  553. font-size: 15px;
  554. color: #222;
  555. margin: 0 0 12px 0;
  556. line-height: 1.5;
  557. ",
  558. htmltools::tags$div(
  559. htmltools::tags$b("Perfusate: "),
  560. htmltools::tags$span(style = "color:#0072B2;", "● Pure blood "),
  561. htmltools::tags$span(style = "color:#D55E00;", "● Carbon black "),
  562. htmltools::tags$span(style = "color:#CC79A7;", "● Mixed blood "),
  563. htmltools::tags$span(style = "color:#666666;", "● Other non-blood")
  564. ),
  565. htmltools::tags$div(
  566. htmltools::tags$b("Shape: "),
  567. htmltools::tags$span("● Review-defined adequacy criterion "),
  568. htmltools::tags$span("▲ Study-defined adequacy criterion")
  569. ),
  570. htmltools::tags$div(
  571. htmltools::tags$b("Size: "),
  572. htmltools::tags$span(
  573. "larger points = larger n per arm; unreported n plotted at minimum size"
  574. )
  575. ),
  576. htmltools::tags$div(
  577. htmltools::tags$b("Trend: "),
  578. htmltools::tags$span(
  579. "LOESS smoothers are shown for pure blood and carbon black as descriptive trends; no inferential confidence intervals are shown."
  580. )
  581. )
  582. )
  584. } else {
  586. legend_html <- htmltools::tags$div(
  587. style = "
  588. font-family: Arial, sans-serif;
  589. font-size: 15px;
  590. color: #222;
  591. margin: 0 0 12px 0;
  592. line-height: 1.5;
  593. ",
  594. htmltools::tags$div(
  595. htmltools::tags$b("Perfusate: "),
  596. htmltools::tags$span(style = "color:#0072B2;", "● Pure blood "),
  597. htmltools::tags$span(style = "color:#D55E00;", "● Carbon black ")
  598. ),
  599. htmltools::tags$div(
  600. htmltools::tags$b("Size: "),
  601. htmltools::tags$span(
  602. "larger points = larger n per arm; unreported n plotted at minimum size"
  603. )
  604. ),
  605. htmltools::tags$div(
  606. htmltools::tags$b("Trend: "),
  607. htmltools::tags$span(
  608. "LOESS smoothers are shown for pure blood and carbon black as descriptive trends; no inferential confidence intervals are shown."
  609. )
  610. )
  611. )
  612. }
  615. fig_page <- htmlwidgets::prependContent(
  616. fig,
  617. htmltools::tags$div(
  618. style = "padding: 30px; max-width: 1600px; margin: 0 auto;",
  619. htmltools::tags$h1(page_title),
  620. legend_html
  621. )
  622. )
  624. fig_page <- htmlwidgets::appendContent(
  625. fig_page,
  626. htmltools::tags$div(
  627. style = "padding: 0 30px 30px 30px; max-width: 1600px; margin: 0 auto;",
  628. htmltools::tags$p(
  629. class = "note",
  630. "Hover over points for study details. Click a point to open the source paper."
  631. )
  632. )
  633. )
  635. htmlwidgets::saveWidget(
  636. fig_page,
  637. file = output_file,
  638. selfcontained = TRUE,
  639. title = page_title
  640. )
  641. }
  643. fig1a <- make_plot(d, "pct_brain", "% of brain perfused", "Spatial completeness of perfusion")
  644. ggsave("fig1a_pct_brain.png", fig1a, width = 7.5, height = 7, dpi = 150)
  647. fig1b <- make_plot(d, "pct_brains", "% of brains adequately perfused", "Frequency of adequate perfusion")
  648. ggsave("fig1b_pct_brains.png", fig1b, width = 7.5, height = 7, dpi = 150)
  650. make_interactive(
  651. fig1a,
  652. "fig1a_pct_brain_interactive.html",
  653. "Spatial completeness of perfusion"
  654. )
  656. make_interactive(
  657. fig1b,
  658. "fig1b_pct_brains_interactive.html",
  659. "Frequency of adequate perfusion"
  660. )
  662. # ===========================================================
  663. # 3. Arms not represented in spatial completeness figure (qualitative CBF arms)
  664. # ===========================================================
  666. d2 <- ext |>
  667. mutate(
  668. pct_brain = parse_pct(`Converted % Brain Perfused`),
  669. pct_brains = parse_brains(`% Brains Adequately Perfused`),
  670. isch = parse_num(`Ischemia Duration (min)`),
  671. cat_raw = trimws(`Perfusion Quality Category`)
  672. ) |>
  673. filter(isch <= 60)
  675. not_in_fig2 <- d2 |> filter(is.na(pct_brain))
  677. qual_cat <- function(x) {
  678. x <- tolower(x)
  679. case_when(
  680. grepl("hyperperfusion|hyperemia", x) ~ "Reactive hyperemia",
  681. grepl("no-reflow absent", x) ~ "No-reflow absent",
  682. grepl("no-reflow present", x) ~ "No-reflow present",
  683. grepl("hypoperfusion", x) ~ "Hypoperfusion (other)",
  684. grepl("qualitative flow present", x) ~ "Qualitative flow present",
  685. grepl("qualitative flow absent", x) ~ "Qualitative flow absent",
  686. TRUE ~ paste0("OTHER: ", x)
  687. )
  688. }
  690. not_in_fig2 <- not_in_fig2 |>
  691. mutate(
  692. qcat = qual_cat(cat_raw),
  693. tier = case_when(
  694. qcat %in% c("Reactive hyperemia", "No-reflow absent",
  695. "Qualitative flow present") ~ "adequate",
  696. qcat == "Hypoperfusion (other)" ~ "reduced",
  697. qcat %in% c("No-reflow present",
  698. "Qualitative flow absent") ~ "failed")
  699. )
  701. cat("\n--- Arms not in Fig 2, by qualitative category ---\n")
  702. print(table(not_in_fig2$qcat))
  704. not_in_fig2 |>
  705. group_by(tier) |>
  706. summarise(n = n(), median_isch = median(isch, na.rm = TRUE),
  707. q25 = quantile(isch, .25, na.rm = TRUE),
  708. q75 = quantile(isch, .75, na.rm = TRUE),
  709. min = min(isch, na.rm = TRUE),
  710. max = max(isch, na.rm = TRUE), .groups = "drop") |>
  711. as.data.frame() |> print()
  713. cat("\nAdequate vs failed tier, ischemia duration (Wilcoxon rank-sum):\n")
  714. print(wilcox.test(
  715. not_in_fig2 |> filter(tier == "adequate") |> pull(isch),
  716. not_in_fig2 |> filter(tier == "failed") |> pull(isch)
  717. ))

ischemia_review_code.R at commit 2d68580, no license · at the source

Overview

Authors: Jeremy Kalfus1, Devin Ward1,2, Borys Wróbel2,3, Andrew T McKenzie1
ORCID iDs: Jeremy Kalfus, Devin Ward, Borys Wróbel, Andrew T McKenzie
  1. Apex Neuroscience, Salem, OR 97317, USA; (J.K.); (D.W.)
  2. Nectome, Vancouver, WA 98682, USA
  3. European Institute for Brain Research, 1013LE Amsterdam, The Netherlands
Journal: Brain sciences, volume 16, issue 8, article 882
Dates: received 10 July 2026; accepted 13 August 2026; published online 19 August 2026
Type: Review · Language: English
License: CC BY
Identifiers: DOI 10.3390/brainsci16080882 · PMID 42651190 · PMCID PMC13510562 · OpenAlex W7169072823
Open access: gold, a free copy (OpenAlex)
Preprint: doi:10.64898/2026.07.10.737845
Status: code verified
Categories: human (organism), other (organism), stroke (population), clinical / translational (subfield)
Keywords: ischemia, cerebral perfusion, no-reflow phenomenon, circulatory arrest, reperfusion, brain banking, resuscitation
Topic: Acute Ischemic Stroke Management (Epidemiology, Medicine), according to OpenAlex
Citations: not cited yet (Europe PMC); 80 references in the paper

Abstract

Highlights: What are the main findings?

How long the brain remains perfusable after circulatory arrest still needs to be established definitely.

Perfusion quality declines with increasing duration of global cerebral ischemia, but the available evidence does not identify a consistent sharp temporal threshold.

What are the implications of the main findings?

Standardized, precise measures of perfusion quality are needed at the level of microvasculature.

The perfusability window of the human brain may be extendable through pre-mortem or post-mortem interventions. This will require consideration not only of efficacy but also of practical, ethical, and legal constraints.

Abstract: Background: Global cerebral ischemia initiates a cascade of pathophysiological changes that progressively impair subsequent perfusion of brain tissue. Some authors have proposed that adequate cerebral perfusion becomes impossible after approximately 10–30 min of global ischemia. However, the extant evidence base for this threshold and its variations across studies has not yet been systematically examined. Objective: To synthesize the literature on post-ischemic cerebral perfusion success as a function of ischemia duration. Methods: We searched PubMed (11 February 2026) for studies of global cerebral ischemia in animal or human models that reported quantitative or categorical measures of perfusion quality. Eligible studies included those assessing perfusion via restoration of blood flow, via external perfusion of non-blood solutions, and/or via tracer injection following reperfusion. Studies of focal ischemia were excluded. Data extracted included species, ischemia duration, temperature during ischemia, ischemia model, perfusate type, and perfusion quality assessment method. The perfusion quality outcome was operationalized as either the average percentage of brain tissue perfused or the percentage of brains in a group that was adequately perfused. Study quality was assessed using a custom domain-specific checklist. Results: We included 60 studies with 192 study arms reporting on the perfusion of the brains of rabbits, rats, pigs, dogs, cats, and humans. Studies differed in the model of ischemia, the perfusate, the perfusion parameters, the quality assessment methods, and other factors. Longer ischemia was associated with lower perfusion quality, but substantial heterogeneity across studies prevented identification of a consistent sharp temporal threshold. Some studies found that at least partial perfusion was possible after longer periods. Within-study dose–response curves were more consistent than the pooled cross-study pattern. Conclusions: How long the brain remains perfusable after circulatory arrest has not yet been definitively established. On average, perfusion quality clearly declines rapidly as the duration of global cerebral ischemia increases. However, some studies, often using interventions such as hypothermia or vasopressors, have reported at least partial perfusion of the brain even after 30 or 60 min of ischemia. Moreover, at least partial perfusion has been reported in human brain banking studies after postmortem intervals of several hours or days in some donors. Limitations of this review include substantial heterogeneity in study methods and outcome measures, which precluded formal meta-analysis. Future research may benefit from more thorough and precise measures of perfusion quality.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repository

Its files are read in the Code ↔ Paper reader above, with 6 matches between paragraphs and lines of code.

andymckenzie/Ischemic_perfusion_review

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: 2d685802b951125875e0ea9b9f0b591abe326829, 11 August 2026
Languages: R (1)
Size: 7 files, 1 script
Software Heritage: not archived
Found in: “Data Availability Statement”
Holds: README, documentation
Not found: license file, CITATION.cff, environment file, tests, continuous integration
Tools: ggplot2 (1 file), patchwork (1 file), Plotly (1 file), tidyverse (1 file)
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
2 files
At the source: github.com/andymckenzie/Ischemic_perfusion_review

The paper's code and data availability statement is in the Data section.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 1 script, each with its path and the digest of its content;
  • 6 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data Availability Statement

R code and data used for analysis are available at https://github.com/andymckenzie/Ischemic_perfusion_review (accessed on 12 August 2026).

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 4 authors, 7 keywords, 77 references.

Cite

This paper

Kalfus, J., Ward, D., Wróbel, B., & McKenzie, A. T. (2026). How Long Is the Brain Perfusable After Global Ischemia? A Systematic Review. Brain sciences, 16(8), 882. https://doi.org/10.3390/brainsci16080882

BibTeX

@article{kalfus2026how,
author = {Kalfus, Jeremy and Ward, Devin and Wróbel, Borys and McKenzie, Andrew T},
title = {{How Long Is the Brain Perfusable After Global Ischemia? A Systematic Review}},
journal = {Brain sciences},
year = {2026},
month = aug,
volume = {16},
number = {8},
pages = {882},
publisher = {Multidisciplinary Digital Publishing Institute (MDPI)},
issn = {2076-3425},
doi = {10.3390/brainsci16080882},
url = {https://doi.org/10.3390/brainsci16080882},
pmid = {42651190},
pmcid = {PMC13510562}
}

RIS

TY - JOUR
AU - Kalfus, Jeremy
AU - Ward, Devin
AU - Wróbel, Borys
AU - McKenzie, Andrew T
TI - How Long Is the Brain Perfusable After Global Ischemia? A Systematic Review
T2 - Brain sciences
J2 - Brain Sci
PY - 2026
DA - 2026/08/19
VL - 16
IS - 8
SP - 882
SN - 2076-3425
PB - Multidisciplinary Digital Publishing Institute (MDPI)
DO - 10.3390/brainsci16080882
UR - https://doi.org/10.3390/brainsci16080882
LA - en
ER -

CSL-JSON

{
"id": "10.3390/brainsci16080882",
"type": "article-journal",
"title": "How Long Is the Brain Perfusable After Global Ischemia? A Systematic Review",
"container-title": "Brain sciences",
"author": [
{
"family": "Kalfus",
"given": "Jeremy"
},
{
"family": "Ward",
"given": "Devin"
},
{
"family": "Wróbel",
"given": "Borys"
},
{
"family": "McKenzie",
"given": "Andrew T"
}
],
"container-title-short": "Brain Sci",
"volume": "16",
"issue": "8",
"page": "882",
"DOI": "10.3390/brainsci16080882",
"PMID": "42651190",
"PMCID": "PMC13510562",
"ISSN": "2076-3425",
"publisher": "Multidisciplinary Digital Publishing Institute (MDPI)",
"URL": "https://doi.org/10.3390/brainsci16080882",
"language": "en",
"issued": {
"date-parts": [
[
2026,
8,
19
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1371/journal.pone.0344932 [code]
Cryopreservation of aldehyde-fixed whole brains.
Journal: PloS one
In common: ggplot2, 2 references, author Andrew T McKenzie
[2] doi:10.17879/freeneuropathology-2026-9411 [code]
Brain extraction for fixed tissue banking: a technical report.
Journal: Free neuropathology
In common: ggplot2, tidyverse, 3 references
[3] doi:10.1016/j.isci.2026.115196 [code]
Transcriptional and cellular maturation of the chick spinal cord in the context of distinct neuromuscular circuits.
Journal: iScience
In common: Plotly, patchwork, ggplot2, 1 other tool, other
[4] doi:10.1016/j.stemcr.2026.102922 [code]
Spatially resolved transcriptomics identifies intercellular signaling post-ischemic stroke that controls neural stem cell proliferation.
Journal: Stem cell reports
In common: Plotly, patchwork, ggplot2, 1 other tool, stroke
[5] doi:10.1038/s41514-026-00397-3 [code]
Nasal administration of Protollin enhances monocyte phagocytosis and decreases CD8&lt;sup&gt;+&lt;/sup&gt; T cell cytotoxicity in subjects with early Alzheimer's disease: a Phase 1 clinical trial.
Journal: npj aging
In common: Plotly, patchwork, ggplot2, 1 other tool, clinical / translational
[6] doi:10.1093/braincomms/fcag176 [code]
Tau topography subtypes account for clinical heterogeneity and longitudinal trajectories in early-onset Alzheimer's disease.
Journal: Brain communications
In common: Plotly, patchwork, ggplot2, 1 other tool, clinical / translational
[7] doi:10.1038/s42003-026-10957-8 [code]
Brain defence by the extracellular matrix protein Cochlin.
Journal: Communications biology
In common: Plotly, patchwork, ggplot2, 1 other tool
[8] doi:10.1038/s43587-026-01207-x [code]
A microprotein atlas of the human frontal cortex in Alzheimer's disease.
Journal: Nature aging
In common: Plotly, patchwork, ggplot2, 1 other tool
[9] doi:10.1038/s41467-026-76762-3 [code]
PD-1 regulates CD4&lt;sup&gt;+&lt;/sup&gt; T cell-mediated CD8&lt;sup&gt;+&lt;/sup&gt; T cell responses in the brain to balance viral control and neuroinflammation.
Journal: Nature communications
In common: Plotly, patchwork, ggplot2, 1 other tool
[10] doi:10.1126/sciadv.aeg3223 [code]
The extreme diversity of retinal amacrine cells has deep evolutionary roots.
Journal: Science advances
In common: Plotly, patchwork, ggplot2, 1 other tool

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.