Table 4 displays the most cited references from the 2,325 retrieved publications. Speech and language therapy for aphasia following stroke occupied the first place in terms of citations (102), with a mediated centrality of 0.14, from the Cochrane Database of Systematic Reviews (IF12.008, Q1). Using the clustering function of CiteSpace to group co-cited references as shown in Figure 9A, the modularity Q was 0.7896 (>0.5), indicating that the clustering network was reasonable, the S value was 0.9127 (>0.7), indicating that the homogeneity of the clustering network was acceptable. The gradation of the circles from purple to yellow depicts the temporal dimension and can represent a shift in the focus and direction of the study. “Non-invasive brain stimulation” was the largest cluster (cluster #0), followed by “transcranial direct current stimulation (tDCS)” (cluster #1), “fMRI” (cluster #2), and “functional connectivity” (cluster #3); other significant clusters included “positron emission tomography,” “constraint-induced aphasia therapy,” “technology,” “therapeutic relationships,” “genetics,” “health-related quality of life” and others.

The most frequently cited literature with the strongest citation bursts is considered the basis for future cutting-edge research, as shown in Figure 9B, which shows the top 20 most cited bursts and is arranged chronologically. The blue line indicates the observed time interval from 1998 to 2022, and the red color represents the duration of the burst, thus illustrating research hotspots and durations. The Western Aphasia Battery-Revised had the highest outbreak citation strength (strength = 36.42) and was an assessment of language function in patients with acquired neurological disorders (e.g., stroke, head injury) with aphasia (Kertesz, 2007). There is a continuing outbreak of citations for seven articles, four of which deal with rehabilitation methods for post-stroke aphasia, including various means of speech and language rehabilitation and transcranial direct current stimulation (tDCS) therapy (Brady et al., 2016; Breitenstein et al., 2017; Fridriksson et al., 2018; Palmer et al., 2019). A previous meta-analysis thoroughly examined the incidence, rehabilitation trends, and outcomes of post-stroke aphasia (Flowers et al., 2016), thus serving as an invaluable resource for understanding post-stroke aphasia for researchers. Two other papers provide insights into the mechanisms of functional recovery and neuroplasticity of language neural networks in post-stroke aphasia (Hartwigsen and Saur, 2019; Kiran and Thompson, 2019).

First, keywords with similar meaning were combined to determine the frequency of keyword publication, and the top 20 keywords are summarized as shown in Table 5. The keyword network can show the main research objects and research hotspots, and we use VOSviewer to implement the co-occurrence network of keywords and cluster the keywords into groups. The minimum number of keyword occurrences was set to 20; 172 out of 8,016 keywords met the criteria and were thus included in the analysis and divided into 4 color clusters. The keyword clustering visualization is shown in Figure 10A. In addition to “stroke, aphasia, rehabilitation”, the main keywords in the red cluster included “language, plasticity, fMRI, brain, activation”; the main keywords in the green cluster included “disability, clinical trials, depression, prognosis”; the main keywords in the blue cluster included “people, quality of life, communication, participation”; and the keywords in the yellow cluster included “transcranial magnetic stimulation, tDCS, non-invasive brain stimulation.” In addition, we performed a visualization of the temporal overlap of keywords, as shown in Figure 10B, with the earlier keywords shown in purple and the yellow ones indicating the most recent keywords. A change from purple to yellow represents a shift in research trends and hotspots.

Figure 10C displays the top 25 keywords from 1998 to 2022 with the highest citation bursts. In addition to the search term-related terms, the top five keywords with the strongest citation explosion were transcranial magnetic stimulation (bust strength = 14.93), positron emission tomography (bust strength = 14.37), plasticity (bust strength = 10.59), activation (bust strength = 9.63), and motor recovery (bust strength = 8.92). Over time, positron emission tomography (1999–2011), unilateral neglect (1999–2010), activation (2000–2010), amphetamine (2000–2010), and induced movement therapy (2005–2014) have received the most sustained attention for these keywords. Participation (2019–2020), experience (2020–2022), and connectivity (2020–2022) have been used more recently and may become hot research topics in the future.

Current research suggests that after stroke-induced brain lesions, the neurofunctional network of the brain can be reorganized to help compensate for and partially recover areas of the damaged neural network (Joy and Carmichael, 2021). Brain function is reorganized in three consecutive phases: acute, subacute, and chronic, the first two being spontaneous neuroplastic changes, while the chronic phase requires the provision of various specific interventions to promote further functional improvement (Rossini and Forno, 2004; Kerr et al., 2011). Factors affecting neuroplasticity in post-stroke aphasia include intrinsic factors (e.g., the age of onset, blood perfusion, stroke lesion characteristics and language impairment patterns) and extrinsic environmental factors (e.g., treatment and psychosocial variables) (Hartwigsen and Saur, 2019). Functional neuroimaging can reveal the neural mechanisms of language network reorganization by mapping neural activity and functional and structural connectivity changes to predict outcomes and improve treatment efficiency (Hartwigsen and Saur, 2019). Among numerous methods, functional magnetic resonance imaging (fMRI) has been most often used to examine functional connectivity changes in networks associated with language recovery. By recording changes in connectivity over time in the absence of a task (e.g., resting state), fMRI can explore the natural processes and patterns of language function recovery after brain injury; comparison with task-based connectivity changes allows the construction of a model for treatment-induced language function recovery (Stockert et al., 2020; Wilson and Schneck, 2020). Electrophysiological methods are also frequently used to study neural changes in patients with post-stroke aphasia. Magnetoencephalography (MEG) and electroencephalography (EEG) are the most commonly used measurement tools (Johnson and Fridriksson, 2022). By tracing the spatial topography of specific oscillatory bands and frequencies associated with preserved language abilities after stroke, electrophysiology can be used to explore the electromagnetic reorganization activity of communication connections between electrophysiological regions in post-stroke aphasia (Shah-Basak et al., 2022). In addition, research on stroke recovery biomarkers (SRBs) is booming, while SRBs that promote language reorganization are still poorly understood and could be incorporated into post-stroke aphasia rehabilitation in the future (Boyd et al., 2017; Ulanov and Shtyrov, 2022). In the future, multivariate analysis, using multiple functional neuroimaging and non-invasive brain stimulation approaches, can be adopted to integrate and analyze the collected data to more acutely capture the neural network changes in the brain that drive recovery from post-stroke aphasia (Hartwigsen and Saur, 2019). In addition, further linking speech therapy to neural changes in specific lesion sites of language activity and connectivity, and actively explore and develop treatment-induced neural reorganization models that can provide the basis for personalized targeted rehabilitation of post-stroke aphasia (Mattioli, 2019).

A large stroke study found that it is practically feasible to follow a scientific framework for selecting and implementing standardized language assessment procedures in acute care and inpatient rehabilitation settings to assess early changes in language performance at multiple time points (Stipancic et al., 2019; Schliep et al., 2020). In practice, Language function assessment provides clinicians with information for the differential diagnosis of aphasia, measures the effectiveness of various treatment modalities, and provides clues for investigating the anatomical and biological correlates of language function, as well as the location and function of damaged brain structures (Kertesz, 2022). Numerous previous studies have also extensively investigated the physical and psychological influences on post-stroke aphasia using language function assessment tools and have longitudinally followed the long-term recovery of language function in patients with post-stroke aphasic patients to monitor their language improvement over time and to guide individualized speech therapy (Wang et al., 2018; Stipancic et al., 2019). Since the scoring of language function and sign manipulation usually require professional guidance, meeting the needs of aphasic patients for teleconsultation has become a hot topic. A videoconferencing management approach for WAB-R that allows language and communication assessments to be operated remotely through videoconferencing technologies (VCT) and can serve as a viable alternative to face-to-face assessments (Dekhtyar et al., 2020). In addition, various language function screening tools such as the Language Screening Test (LAST) (Flamand-Roze et al., 2011), ScreeLing (Doesborgh et al., 2003), etc., have been developed to enable treatment-oriented assessment of language function in a short period of time in a clinical setting (El Hachioui et al., 2017). Among them, The Short and Tailored Evaluation of Language Ability (STELA), a newly developed computer-based Japanese language proficiency assessment system for patients with aphasia, simplifies the administration of test entries by computer, automates the recording of scores and supplements them with additional objective measurements (e.g., reaction time), greatly reducing the testing time for patients (Inamoto et al., 2023). Since most current tools for assessing language function follow neural models rather than psycholinguistic theories, the assessment validity of post-stroke aphasia patients may be affected if they also have cognitive impairment or psycho-behavioral disorders, and other coexisting neurological deficits, such as apraxia or visual field deficits, may hinder the prognostic outcome of post-stroke aphasic patients and need to be evaluated together (Kertesz, 2022). At the same time, there are also limitations in current language function assessment methods, such as linguistic differences between cultures and geographic regions, and whether the localized scales have reliability and validity, which can affect the assessment of language function and its efficacy (Sun et al., 2020). It follows that more language function assessment tools need to be developed and validated in the future to explore the feasibility of their measurement in different countries and populations. In addition, the use of intelligent medical assessment methods will be further expanded in the future for different populations and in a wider range of clinical settings, providing the prerequisites for conducting extensive clinical practice with quantitative or automated instruments for post-stroke aphasia assessment by increasing clinical practice based on this aspect.

Speech and language therapy (SLT) is by far the most popular method for treating post-stroke aphasia. “Speech and language therapy for aphasia following stroke” is the most frequently cited reference for this article (102 articles) (Brady et al., 2016). This systematic review reports that SLT improves reading, writing, communication function, and speech expression in post-stroke aphasic patients and suggests that high-intensity training improves language skills, but there is often a high rate of disengagement. A multicenter, single-blinded, randomized controlled trial found that self-managed, computerized speech and language therapy (CSLT) plus usual care significantly improved word finding in patients with chronic stroke aphasia and that patients improved their participation in recovery through the use of self-managed CSLT (Palmer et al., 2019). Although there is evidence that people with post-stroke aphasia can benefit from SLT, transportation closures, time conflicts, or lack of appropriate rehabilitation services in hospitals make long-term recovery very difficult (Morris et al., 2017). Therefore, it is important to ensure proper treatment dosage and to maintain long-term recovery. Research indicates that tele-rehabilitation for aphasia is not only has equal validity compared to traditional face-to-face language rehabilitation, but also allows patients to enjoy high-quality rehabilitation in a familiar environment that is more convenient (Cacciante et al., 2021). Currently, remote rehabilitation via videoconferencing is considered a viable rehabilitation model for post-stroke aphasia (Øra et al., 2020), through web-based oral reading of language for aphasia (Web ORLA), people with chronic aphasia are encouraged to perform repetitive choral and independent readings with a virtual therapist, which can lead to improvements in language function (Cherney et al., 2021). In addition, it has been suggested that the virtual context provided by VR technology can promote the ecological effectiveness of speech therapy (Bu et al., 2022). It is thus clear that providing precise rehabilitation goals for post-stroke aphasic patients and exploring rehabilitation tools that are motivating and not easily limited by time and place will become a hot topic in this field. In terms of physiotherapy, non-invasive brain stimulation, such as transcranial direct current stimulation (tDCS), can improve post-stroke aphasia by modulating cortical excitability as an adjunct to synergistic SLT (Lefaucheur et al., 2017; Hartwigsen and Saur, 2019). However, most of the current studies have small sample sizes, inconsistent intervention durations, and significant differences between interventions and participants across studies. Large, multicenter randomized controlled trials are still needed in the future to clarify the optimal time to start recovery from post-stroke aphasia, the type of aphasia, and the stimulation parameters of physical therapy (Picano et al., 2021; REhabilitation and recovery of peopLE with Aphasia after StrokE (Release) Collaborators, 2022). When assessing outcomes, the impact of the intervention on social communication or quality of life should be included in addition to attention to outcome indicators at the site of injury.

Communication barriers often limit the ability of people with post-stroke aphasia to communicate with others, preventing them from accessing appropriate rehabilitation information and services (Worrall et al., 2011). Family support, acquaintance with others, and a supportive social environment can improve the participation in rehabilitation and quality of life of people with post-stroke aphasia, and accessible information and collaborative interaction with health professionals can lead to self-rehabilitation management of people with post-stroke aphasia (Manning et al., 2019; Manning et al., 2020). A meta-analysis examined the adjustment process of family communication patterns during post-stroke aphasia and the factors that facilitate and hinder family communication, encouraging an open approach to family communication and a variety of modern technologies to support people with post-stroke aphasia and family members (Ramazanu et al., 2022). Qualitative findings from the perspective of patients with aphasia (Nichol et al., 2022) indicate that self-management of aphasia involves skills in daily conversation, social interaction, and life participation. Community groups and communication partners can provide peer support and assistance with self-management, health professionals can provide information and educational training in related areas, and people with aphasia can learn technology for communication and daily interaction. Current qualitative research approaches to explore in depth the rehabilitation needs and engagement experiences of people with post-stroke aphasia are topical. However, it is worth noting that post-stroke aphasia also has an important impact on the lives of family members (Grawburg et al., 2019), and further investigation of the rehabilitation goals of family members of patients with post-stroke aphasia is needed (Zawawi et al., 2020). In addition, further adoption of a mixed study approach will help us to better personalize rehabilitation for patients with post-stroke aphasia (Devanga et al., 2021).