For tDCS, the electrical current is delivered in a continuous and consistent manner (direct current). The stimulation intensity will be the same as that of tRNS (1.5 to 2 mA based on the tolerance of participants), which is safe for stroke patients.The stimulation period of tRNS and tDCS will be 20 minutes.

For the sham stimulation condition, the electrode placement will be the same as the other real eletrical stimulation conditions. The current will first turn up for 30 seconds and subsequently turn off in the next 30 seconds to provide sensory feeling without actually stimulating the brain.

For the digital MT, the participants will be encouraged to move the affected arm as symmetrically and simultaneously as the unaffected hand while looking at the real-time image on the computer screen. The activities practiced during digital MT will include intransitive movements (e.g., flexion and extension of the wrist) and transitive movements (e.g., picking up a pen). The amount of time exposed to intransitive and transitive movements will be balanced for each participant. The participants will practice 30 intransitive and 30 transitive tasks over the 20 training sessions, and the actual movements will be selected based on each individual's ability and his/her needs.

The WMFT is a reliable method to evaluate the UE motor ability in patients after stroke. The WMFT contains 15 function-based tasks and 2 strength-based tasks. The time (WMFT-time) and the functional ability (WMFT-quality) for an individual to complete the tasks will be recorded. A shorter WMFT-time and a larger WMFT-quality score indicate a faster movement and better quality of movement, respectively.

The MAL will be used to assess the amount of use (AOU) and quality of movement (QOM) of the paretic UE. The MAL is a semi-structured interview that tests object manipulation and gross motor activities of daily living. The psychometric properties of The MAL adopts a 6-point ordinal scale, although patients can attribute a half-score, resulting in 11-point Likert scales with specified anchoring definitions at 6 points.Scores range from 0 to 5.

MAL have been well-established, and a higher MAL score indicates better movement quality.

The NEADL is a self-report scale that measures instrumental activities of daily living. It evaluates 4 areas of daily living, including mobility, kitchen, domestic, and leisure activities. The total score is 0-66, and a higher score indicates better daily functional ability. The psychometric properties of NEADL have been well established.

To evaluate health-related quality of life, the SIS 3.0 will be used. The SIS consists of 59 test items grouped into 8 domains (strength, hand function, ADL/IADL, mobility, communication, emotion, memory and thinking, and participation/role function). The participants will be asked to rate each item in a 5-point Likert scale regarding to the perceived difficulty in completing the task. The total score for each domain ranges from 0 to 100. An extra question will be asked to evaluate the participant's self-perceived overall recovery from stroke. The SIS 3.0 has satisfactory psychometric properties.

The Functional Abilities Confidence Scale (FACS) was designed to measure the degree of self-efficacy or confidence a patient exhibits with various movements or postures.

The FACS consists of 15 questions which are answered and scored on a 0% (not confident at all) to 100% (completely confident) basis. The higher the percentage the higher the confidence level. The patients are asked to circle the percentage that best describes his or her level of confidence that they could perform the activity, in various situations, regardless of pain or discomfort.

The DLSES is designed to measure self-efficacy in psychosocial functioning and self-efficacy in activities of daily living, regardless of level of physical impairment.

Typically, self-efficacy scales present items depicting different task demands that are rated according to the strength of belief in the ability to perform those activities.

a Likert scale of 0 'cannot do at all' to 100 'highly certain can do' was used for this measure.

A total score is obtained by summing all items, with higher scores indicative of higher self-efficacy. Participants were instructed to rate their level of confidence in performing each of the daily living activities/behaviours listed on the DLSES.

A wireless EEG device will be used to evaluate treatment-induced changes on cortical activity by ERD of mu rhythm EEG during the movements for reaching for pressing the desk button. The mu rhythm is a specific frequency range (8-12 Hz) in the EEG signal and the amplitude decrease in mu-rhythm power (called ERD) can be used to depict the temporal pattern of cortical activity when preparing, producing, and controlling movement events.

The averaged area of the entire ERD curve under the reference level will be used as the amplitude parameter of cerebral activation. To characterize the ERD difference between the affected hemisphere and the unaffected hemisphere, a lateralization index (LI) will be used: LI = (ERDR- ERDL)/(ERDL+ERDR), where ERDR and ERDL represent the overall ERD areas (cerebral activation) of the C4 and C3 (or F4 and F3) channels. Significant increase of ERD in the damaged hemisphere and increase of LI will be indicators for a good recovery.

Participants will be instructed to perform unilateral and bilateral arm movement tasks and reach-to-grasp movement tasks in sitting position. A static calibration trial is collected to configurate the joint coordinate system. Based on the position of these reflective markers, an upper body model will be constructed in Vicon Nexus for kinematic analyses. The position data of body segments and muscle activations will be captured by a seven-camera Vicon system for each participant during the task. The investigators will use Vicon nexus software and a custom Matlab program to process the kinematic and EMG data.