Post stroke aphasia (PSA) is one of the most frequently happened deficiency of stoke, affecting speaking,comprehension, writing and reading of language. Generally, PSA is commonly seen in cortical damage, but in recent years it has been found that subcortical injury is also an important cause of PSA, which is called subcortical aphasia. Using fMRI technology, the investigators aim to investigate the language function of patients with subcortical cerebral infarction at different stages of recovery , and explored the mechanism of post-injury language reorganization in the brain.

The investigators recruited 60 first-episode acute cerebral infarction patients with one-side lesion in subcortical white matter (40 with left injury and 20with right injury) and 20 health volunteers. All participants are right-handed, and screened with MMSE, HAMD and HAMA to exclude cases of psychosis, post-stroke dementia and depression. Each participant was arranged to have three test sessions at different stages after the infarction (T1:within 3 days after onset of the stroke ; T2:28 ±3days after onset; T3: 90±3days after onset), with fMRI and Western aphasia battery (WAB) in each session.

The purpose of this study is to explore the pathogenesis of subcortical aphasia, and to understand the dynamic reorganization of language network during the recovery of language function.

The participants were recruited from inpatients with acute ischemic stroke in Department of Cerebrovascular disease, The Guangzhou General Hospital of Guangzhou Military Command. The diagnose of ischemic stroke was made using the diagnostic criteria of the International Association of Neurological Diseases and Stroke Association in 1982. The classification criterias for ischemic stroke were based on the current international TOAST etiological classification method.

The study passed the approval of the ethics committee of General Hospital of Guangzhou Military Command , and all participants or their guardian signed informed consent. According to the location and diagnostic criteria, the participants were divided into three groups: left hemisphere infarction patients group ,right hemisphere infarction patients group and normal healthy control group. The participants performed 3 language functional behavior tests and functional magnetic resonance (fMRI) tests within3 days,28 ±3days ,90±3days after the onset of cerebral infarction. In the healthy control group, the above examination was performed only 1 times. Language functional behavioral assessments included the Chinese version of Western Aphasia Battery(WAB), spontaneous language frequency test(SLFT) and picture naming test(PNT). Examines of fMRI included task-state fMRI , resting-state fMRI and diffusion tensor imaging(DTI).

Task state function magnetic resonance design：Using the block design, each sequence uses the "baseline-stimulus-baseline-stimulus-baseline -stimulus-baseline-stimulus-baseline-stimulus-baseline-stimulus-baseline-stimulus-baseline" model. The first baseline duration is 24s. The following baseline duration and duration of stimulus are 18s, each sequence is 240s, a total of three sequences. Scan 12min. From the Snodgrass picture database, 54 animal pictures and 54 tool pictures can be accurately identified by all subjects. Repeat 6 animal chunk and 6 tool naming chunks, six pictures of each block for 18 seconds. Choose the abstract map of all the American skunk as the baseline map of the animal name, and select the simple arrow picture as the baseline map of the tool. In order to avoid repetition effect after practice, the pictures used in the design of MRI are not repeated in speech behavior assessment. At baseline, subjects were asked to identify the direction of the arrow image and the tail direction of the skunk by silently speaking "upright" or "inverted". The subjects received task familiarization training before testing to ensure that there was no meaningless picture naming in baseline tasks, but location judgment. The design of language task is based on the design of Damasio. In the task state of magnetic resonance scanning, the patient is required to read and name the silent images of the visual images. The visual information is written by the DMDX software, the picture is projected through the brain functional audio-visual stimulation system (SA-9900) to the screen, and the subjects are placed on the head line. The reflector on the circle is observed.

Resting-state fMRI：During the rest-state fMRI scan, no task instruction was given to the participates, and the participates were completely relaxing, closing their eyes, breathing calmly, keeping their head still, but could not fall asleep, tried to avoid any systematic thinking activities, scanning 8min.

Functional magnetic resonance data acquisition：Cranial scanning was performed using the HDX3.0Tesla superconducting magnetic resonance scanner of the US GE company. The 8 channel phased array head coil is the receiving coil, and the scanning sequence and parameters are as follows：

1. T1 structure imaging using FSPGR BRAVO sequence. The parameters included: time of repetition, 8.86 ms; time of echo, 3.52 ms; field of view, 24×24 cm2; in-plane resolution,256×256; slice thickness, 1 mm;interslice gap, 1 mm; and number of slices, 176.
2. Echo-Planar Imaging (EPI) was used to acquire task-state fMRI data.The parameters included: time of repetition, 3000ms; time of echo, 40 ms; field of view, 24×24 cm2; in-plane resolution,64×64; slice thickness, 4 mm; interslice gap, 1 mm; and number of slices, 34. Scan a sequence of 240s, a total of 12 min.
3. Echo-Planar Imaging (EPI) was used to acquire rest-state fMRI data.The parameters included: time of repetition, 3000ms; time of echo, 40 ms; field of view, 24×24 cm2; in-plane resolution,64×64; slice thickness, 4 mm; interslice gap, 1 mm; and number of slices, 34. A total of 8 min.

Inclusion Criteria:

• The first stroke of the left single subcortical areas, within 72hours.
• Primary school or higher level,aged between 18-75, native language Chinese
• According to the commonly used eye chart examination, the corrected visual acuity is more than 1.0.
• According to the Edinburgh Handedness Questionnaire (EHQ) as the right handed.
• The language function was normal before the onset. After the onset, the language function was mildly to moderately impaired with Western Aphasia Battery (WAB) ( Aphasia Quotient (AQ) between in 60 to 88).
• The patient cooperate with the examination, they and their guardian signed the informed consent

Exclusion Criteria:

• History of organic diseases of the nervous system and history of craniocerebral trauma.
• History of epilepsy and psychosis.
• History of material dependence.
• Decompensation of important organ function.
• Hamilton Depression Scale(HAMD )>8 points.
• Hamilton Anxiety Scale(HAMA )>7 points.
• The Mini-Mental State Examination (MMSE)score <20 points.
• Pregnant women and breast-feeding women.
• Contraindication of MRI scanning.

The first onset, the onset time was less than 3 days, the lesion was the left and right single subcortical infarct lesion confirmed by head Magnetic Resonance Imaging(MRI).All of them are come from Neurology/Cerebrovascular Department General Hospital of Guangzhou Military Command of PLA between 2016-2017.