| Study | Language | Inclusion criteria | N aphasia | N control | Data reuse | Notes |
|---|---|---|---|---|---|---|
| Weiller et al. (1995) | German | Lesion including L pSTG; moderate-to-severe Wernicke's aphasia in the subacute period; now recovered and not aphasic per formal testing; able to perform verb generation task | 6 | 6 | N | 6 patients were selected from a database of 600 carefully documented cases |
| Belin et al. (1996) | French | MCA; persistent severe non-fluent aphasia followed by marked improvement with MIT | 7 | 0 | N | |
| Ohyama et al. (1996) | Japanese | Able to repeat single words | 16 | 6 | N | |
| Heiss et al. (1997) | German | — | 6 | 6 | N | |
| Karbe et al. (1998) | German | MCA; able to repeat single words | 12 | 10 | N | Only 7 of the 12 patients took part at T2 |
| Cao et al. (1999) | US English | Aphasia with significant recovery over months to years (ADPASS > 70th percentile) | 6 | 37 | N | 2 additional patients excluded: 1 unable to reliably describe performance post-scan; 1 due to head motion |
| Heiss et al. (1999) | German | AAT repetition ≥ 50 | 23 | 11 | N | |
| Kessler et al. (2000) | German | Mild to moderate aphasia on TT; at least 50 out of 150 on AAT repetition | 24 | 0 | N | |
| Rosen et al. (2000) | US English | L IFG, possibly extending to neighboring regions | 6 | 14 | Y | 1 participant was reported in a previous case study; of the 14 controls, 6 were studied with PET and 8 with fMRI |
| Blasi et al. (2002) | US English | L IFG, possibly extending to neighboring regions | 8 | 14 | N | |
| Leff et al. (2002) | UK English | — | 15 | 8 | N | |
| Blank et al. (2003) | UK English | Initial non-fluent aphasia due to anterior perisylvian lesion; subsequently recovered the ability to speak in sentences; patients were divided into those with and without damage to the IFG pars opercularis (POp+: n = 7; POp-: n = 7) | 14 | 12 | N | 8 of 12 controls included in Blank et al. (2002) |
| Cardebat et al. (2003) | French | No severe aphasia; no leukoaraiosis | 8 | 6 | N | |
| Sharp et al. (2004) | UK English | Lesion in vicinity of L STG; no extensive frontal damage; no inferior temporal damage; able to perform tasks | 9 | 18 | N | |
| Zahn et al. (2004) | German | Global aphasia in the first three months; some improvement of comprehension within 6-12 months | 7 | 14 | N | |
| Crinion & Price (2005) | UK English | — | 17 | 18 | N | |
| de Boissezon et al. (2005) | French | Subcortical stroke; no severe aphasia | 7 | 0 | N | |
| Connor et al. (2006) | US English | L IFG, possibly extending to neighboring regions | 8 | 14 | Y | Re-analysis of data from Blasi et al. (2002) |
| Crinion et al. (2006) | UK English | — | 24 | 11 | N | Results of control participants previously reported in Crinion et al. (2003) |
| Saur et al. (2006) | German | MCA; age < 70 years; able to distinguish forward vs backward speech outside the scanner; no pronounced small vessel disease | 14 | 14 | N | 4 additional patients excluded: 1 health problems; 1 scanner noise; 2 did not tolerate fMRI; 198 patients with aphasia were screened |
| Meinzer et al. (2008) | German | — | 11 | 0 | N | |
| Raboyeau et al. (2008) | French | Naming deficit; good comprehension | 10 | 20 | N | |
| Richter et al. (2008) | German | Main deficits in production rather than comprehension | 16 | 8 | N | 8 additional patients excluded: 5 completed only one of the two sessions; 3 unable to perform the tasks |
| de Boissezon et al. (2009) | French | Only part of L MCA; able to perform word generation; no severe aphasia | 13 | 0 | Y | 7 out of 13 patients appear to represent the same data reported in de Boissezon et al. (2005) |
| Fridriksson et al. (2009) | US English | — | 11 | 10 | N | |
| Menke et al. (2009) | German | Moderate to severe anomia | 8 | 9 | N | |
| Specht et al. (2009) | German | — | 12 | 12 | N | 15 controls were scanned but 3 were randomly excluded to match group sizes for jICA. |
| Warren et al. (2009) | UK English | Comprehension deficit per CAT and TROG (1 patient did not meet this criterion); anterolateral superior temporal cortex spared | 16 | 11 | Y | 8 additional patients excluded: lesions involved L anterolateral superior temporal cortex; reanalysis of subset of dataset from Crinion et al. (2006) |
| Chau et al. (2010) | Cantonese | — | 7 | 0 | N | |
| Fridriksson (2010) | US English | — | 19 | 0 | Y | 7 additional patients excluded: 6 for making fewer than 5 correct responses in one or more sessions; 1 for excessive head motion; "several" patients overlapped with those reported by Fridriksson et al. (2009, 2010); demographic data includes excluded patients |
| Fridriksson et al. (2010) | US English | — | 15 | 9 | N | |
| Sharp et al. (2010) | UK English | Lesion in vicinity of L STG; no extensive frontal damage; no inferior temporal damage; able to perform tasks | 9 | 18 | Y | Additional analysis of same dataset as Sharp et al. (2004) |
| Thompson et al. (2010) | US English | Agrammatic | 6 | 12 | N | |
| Tyler et al. (2010) | UK English | — | 14 | 10 | N | 2 of the 14 patients were not stroke, but were post resective surgery |
| van Oers et al. (2010) | Dutch | MCA; mRS < 3; able to perform at least 2 out of the 3 tasks | 13 | 13 | N | |
| Papoutsi et al. (2011) | UK English | — | 14 | 15 | Y | Reanalysis of same dataset from Tyler et al. (2011); 1 patient had post-surgical haematoma rather than stroke (per Tyler et al., 2011) |
| Sebastian & Kiran (2011) | US English | — | 8 | 8 | N | |
| Szaflarski et al. (2011) | US English | Moderate aphasia, L MCA | 8 | 0 | N | 3 additional patients excluded: 2 metallic artifact; 1 seizure at time of stroke |
| Tyler et al. (2011) | UK English | — | 14 | 15 | Y | Not stated, but it seems like most of the patients also participated in Tyler et al. (2010); 1 patient had post-surgical haematoma rather than stroke |
| Weiduschat et al. (2011) | German | Age 55-85 | 10 | 0 | N | 4 additional patients excluded: 3 malfunction of TMS device or claustrophobia; 1 recovered nearly completely prior to intervention |
| Allendorfer et al. (2012) | US English | MCA; moderate-severe aphasia; mRS ≤ 3 | 16 | 32 | Y | "Part of a larger ongoing study", may overlap with other studies from this group |
| Fridriksson, Hubbard, et al. (2012) | US English | Broca's aphasia | 10 | 20 | N | 3 additional patients excluded: 1 due to a metal implant; 2 for severely non-fluent speech; demographic data includes excluded patients |
| Fridriksson, Richardson, et al. (2012) | US English | — | 29 | 14 | Y | 1 additional patient excluded: contraindications to MRI; 26 of 30 patients were included in Fridriksson (2010); demographic data includes excluded patient |
| Marcotte et al. (2012) | Canadian French | Moderate-severe aphasia; anomia | 9 | 0 | N | |
| Schofield et al. (2012) | UK English | Comprehension deficit | 20 | 26 | Y | 1 additional patient excluded: excessive head motion; patients recruited from database so may have participated in prior studies from this group, but not stated explicitly; demographic data includes excluded patient |
| Wright et al. (2012) | UK English | — | 21 | 21 | Y | Unclear how many, if any, patients were included in previous studies from this group; design is identical to Tyler et al. (2010); 3 of the 21 patients were not stroke, but were post resective surgery |
| Szaflarski et al. (2013) | US English | — | 27 | 0 | N | |
| Thiel et al. (2013) | German | — | 24 | 0 | N | 6 additional patients excluded: 4 did not tolerate MRI or PET scans; 2 TMS device was defective |
| Abel et al. (2014) | German | Anomia; no severe AoS or dysarthria | 14 | 0 | N | 9 additional patients excluded: 4 for ceiling performance; 5 for technical problems |
| Benjamin et al. (2014) | US English | "at least minimal evidence of non-fluent output"; lesion including precentral gyrus or underlying white matter | 14 | 0 | N | |
| Brownsett et al. (2014) | UK English | No involvement of ACA territory | 16 | 17 | N | 3 additional patients excluded: 2 withdrew after attempting first scan; 1 had severe dysarthria |
| Mattioli et al. (2014) | Italian | L MCA; comprehension mildly impaired | 12 | 10 | N | Treated and untreated groups differed in severity at baseline, albeit not significantly |
| Mohr et al. (2014) | UK English | MCA; mild-moderate non-fluent aphasia; no severe comprehension deficit | 6 | 0 | N | 6 additional patients excluded: 4 for health risks; 2 for technical problems and data loss; patient numbers in tables 1 and 2 appear not to correspond with patient numbers later in the paper |
| Robson et al. (2014) | UK English | Wernicke's aphasia (impaired spoken single word comprehension, impaired single word repetition, fluent, sentence-like speech with phonological/neologistic errors) | 12 | 12 | N | |
| Szaflarski et al. (2014) | US English | — | 32 | 32 | Y | Some participants included in Allendorfer et al. (2012); one participant was < 18 years old at time of stroke; there was also a perinatal stroke group, not relevant for this review; 3 participants were excluded but it is not stated whether they were adult or perinatal patients. |
| van Hees et al. (2014) | Australian English | — | 8 | 14 | N | |
| Abel et al. (2015) | German | Anomia; no severe AoS or dysarthria | 14 | 14 | Y | 9 additional patients excluded: 4 for ceiling performance; 5 for technical problems; same dataset as Abel et al. (2014) |
| Kiran et al. (2015) | US English | Impaired naming | 8 | 8 | N | |
| Sandberg et al. (2015) | US English | — | 10 | 0 | N | |
| Geranmayeh et al. (2016) | UK English | No severe receptive aphasia | 53 | 24 | N | Prior strokes were allowed only if no aphasia resulted |
| Griffis et al. (2016) | US English | Moderate aphasia, L MCA | 8 | 0 | Y | 3 additional patients excluded: 2 metallic artifact; 1 seizure at time of stroke; same patients as Szaflarski et al. (2011); different fMRI paradigm acquired in the same sessions |
| Sims et al. (2016) | US English | Some spared tissue in L IFG | 14 | 8 | Y | 2 additional patients excluded: 1 had no spared tissue in the L IFG; 1 had a R hemisphere stroke; although not stated, it is apparent that many of the patients were included in Sandberg et al. (2015) |
| Darkow et al. (2017) | German | L hand motor area spared; mild aphasia | 16 | 16 | N | |
| Geranmayeh et al. (2017) | UK English | — | 27 | 0 | Y | Patients are a subset of those in Geranmayeh et al. (2016); 24 control participants are described, but no imaging data from the controls are analyzed in this paper |
| Griffis, Nenert, Allendorfer, & Szaflarski (2017) | US English | — | 43 | 43 | Y | Same dataset as Griffis et al. (2017) Hum Brain Mapp |
| Griffis, Nenert, Allendorfer, Vannest, et al. (2017) | US English | — | 43 | 43 | Y | Data were collected as part of "several separate studies" |
| Harvey et al. (2017) | US English | Mild-moderate non-fluent aphasia; relatively intact comprehension; able to produce meaningful words and phrases | 6 | 0 | N | |
| Nardo et al. (2017) | UK English | Anomia; good single word comprehension; relatively spared word and nonword repetition; no AoS; spared or partially spared L IFG | 18 | 0 | N | |
| Nenert et al. (2017) | US English | At least mild aphasia per TT | 19 | 38 | Y | Patients are a subset of the 24 participants in Szaflarski et al. (2015), a clinical trial on CIAT |
| Qiu et al. (2017) | Mandarin | Broca's aphasia | 10 | 10 | N | |
| Skipper-Kallal et al. (2017a) | US English | Able to name 20% of pictures correctly in the scanner | 32 | 25 | Y | 14 additional patients excluded: < 20% accuracy in scanner; 29 of the participants overlap with the other Skipper-Kallal et al. (2017) paper |
| Skipper-Kallal et al. (2017b) | US English | 10% accuracy on scanner task | 39 | 37 | Y | 10 additional patients excluded: < 10% accuracy in scanner; 29 of the participants overlap with the other Skipper-Kallal et al. (2017) paper |
| Dietz et al. (2018) | US English | — | 12 | 0 | Y | 2 additional patients excluded: 1 for illness; 1 for MRI contraindication or personal conflict (inconsistent information provided); same data as Dietz et al. (2016), which is a methodological paper |
| Hallam et al. (2018) | UK English | Semantic aphasia; left frontal damage (+ other regions, typically) | 14 | 16 | N | |
| Nenert et al. (2018) | US English | Aphasia at acute screening (not necessarily at first study time point) | 17 | 85 | N | 1 additional patient excluded: significant signal artifacts; presence and severity of aphasia assessed on hospital admission, not at first study time point, so it is not clear that all participants actually had aphasia at first study time point |
| Pillay et al. (2018) | US English | Residual phonologic retrieval deficit; intact semantic processing | 21 | 0 | N | |
| Szaflarski et al. (2018) | US English | — | 12 | 0 | N | 1 additional patient excluded: scanned at only 2 out of 3 time points |
| van de Sandt-Koenderman et al. (2018) | Dutch | Severe non-fluent aphasia (< 50 words/minute); articulation deficits; repetition severely affected; moderate-good auditory comprehension | 9 | 0 | N | |
| van Oers et al. (2018) | Dutch | MRS ≤ 3; ability to perform tasks | 12 | 8 | N | |
| Barbieri et al. (2019) | US English | — | 18 | 23 | N | 1 additional patient excluded: developed a hematoma between baseline and post-testing; one patient had two strokes within one day, but we would consider that essentially a single stroke |
| Johnson et al. (2019) | US English | Anomia | 30 | 17 | N | 5 additional patients excluded: 2 withdrew from non-treatment arm; 3 fMRI acquisition errors; 1 did not complete treatment and post-treatment scanning (but of these latter 4, one must have at least completed the non-treatment arm); there were 26 patients in the treated group and 10 in the untreated group, but 6 patients overlapped between the two groups (they joined the treated group after completing the untreated phase) |
| Kristinsson et al. (2019) | US English | < 80% on PNT; able to name at least 5 out of 40 items during fMRI; WAB-R spontaneous speech ≥ 2; WAB-R auditory comprehension ≥ 2 | 87 | 0 | Y | 65 were previously included in Fridriksson et al. (2018), a tDCS study |
| Purcell et al. (2019) | US English | Chronic dysgraphia (acquired impairment in spelling) | 21 | 0 | N | 4 additional patients excluded: 3 health reasons; 1 data acquisition error |
| Sreedharan, Chandran, et al. (2019) | Malayalam | Broca's aphasia or anomic aphasia; comprehension relatively preserved; "motivated for speech therapy" | 8 | 4 | N | 3 additional patients excluded: 2 for claustrophobia; 1 for transportation issues |
| Hartwigsen et al. (2020) | German | Lesion involving left temporo-parietal cortex and sparing left frontal cortex; relatively well-recovered | 12 | 0 | N | 2 additional patients excluded: 1 lost to follow-up; 1 did not show any sound-related neural activation in auditory cortex after sham cTBS |
| Stockert et al. (2020) | German | Lesion localized to frontal or temporal cortex | 34 | 17 | Y | 50 additional patients excluded: 19 lesions spanned frontal and temporal, or were subcortical, or had persisting large vessel occlusions; 31 not all three timepoints were acquired; 8 patients were included in Saur et al. (2006); there may also be overlap with Saur et al. (2010), a study that did not meet our inclusion criteria; 1630 patients screened for inclusion; frontal patients scanned later than temporal patients at T1 and T2 |
N aphasia = Number of individuals with aphasia; N control = Number of control participants; Data reuse = Were any of the participants included in any previous studies?; AAT = Aachen Aphasia Test; ACA = anterior cerebral artery; ADPASS = Aphasia Diagnostic Profiles Aphasia Severity Score; AoS = apraxia of speech; CAT = Comprehensive Aphasia Test; CIAT = constraint-induced aphasia therapy; fMRI = functional magnetic resonance imaging; IFG = inferior frontal gyrus; jICA = joint independent components analysis; L = left; MCA = middle cerebral artery; MIT = melodic intonation therapy; mRS = modified Rankin Scale; N = No; PET = positron emission tomography; POp+ = pars opercularis damaged; POp- = pars opercularis spared; pSTG = posterior superior temporal gyrus; R = right; STG = superior temporal gyrus; T1, T2, etc. = first time point, second time point, etc.; TMS = transcranial magnetic stimulation; TROG = Test for Reception of Grammar; TT = Token Test; Y = Yes; Yellow underline = minor limitation; Orange underline = moderate limitation.
