Neglect is a clinical feature observed in cases of brain injury, particularly those affecting the right hemisphere, in which individuals have difficulty recognizing the contralesional side of space ^[1][2][1,2]. According to previous studies, patients with neglect tend to have poorer rehabilitation outcomes [3], longer hospital stays [4], and are less likely to be discharged to their homes [5] than patients without neglect. Consequently, neglect has been identified as a significant predictor of decreased levels of functional independence [6], increased utilization of healthcare resources [7], and increased burden on the patient’s family [8]. These findings highlight the detrimental effects of neglect on various aspects of patient outcomes.

Neglect has been divided into various subtypes according to classification criteria because of its heterogeneous characteristics ^[9][10][9,10]. For example, neglect is classified as sensory, motor, and representational (imaginal) neglect. In addition, in terms of the spatial domain, neglect is classified as personal neglect, peri-personal neglect, and extra-personal neglect. In terms of the frame of reference, neglect is categorized as egocentric neglect and allocentric neglect (Table 1) ^[9][10][9,10].

Criteria	Aspect of Neglect	Range of Space	Frame of Reference
Subtype	Sensory neglect	Personal neglect	Egocentric neglect
	Motor neglect	Peri-personal neglect	Allocentric neglect
	Representational neglect	Extra-personal neglect

The frame of reference is defined as the process in which the brain accepts information and identifies the up–down, left–right, front–back directions, and the relative scale of each axis by specifying the central location [11]. The brain receives information differently depending on whether the central criterion is a viewer or an object. Therefore, the symptoms of neglect are also very different [12]. Egocentric neglect is characterized by responses missing on the contralateral side with respect to the viewer. In contrast, allocentric neglect is characterized by responses missing on the contralateral side with respect to the object [13]. Because of these different characteristics of the two types, allocentric neglect may have a more negative impact on participation in activities of daily living than egocentric neglect [14]. A differential diagnosis between allocentric neglect and egocentric neglect has become possible with the development of evaluation tools, such as the Apple Cancellation Test and the Broken Heart Test ^[15][16][15,16]. In general, these evaluations are performed by counting the number of correct (target of complete shape) and incorrect (one of the left and right targets is incomplete) responses after the patient has marked the target of a complete shape ^[15][16][15,16]. Egocentric neglect is determined by the asymmetry score (commission error) between the correct answers on the right and the left based on the center of the evaluation paper. On the other hand, allocentric neglect is discriminated by the asymmetry score (omission error) between the incorrect answer on the right and the wrong answer on the left based on the target in all incorrect answers ^[15][16][15,16]. The criteria for diagnosing egocentric (cut-off ≥ 3) and allocentric neglect (cut-off ≥ 2) are the same in the Apple Cancellation Test and the Broken Heart Test ^[15][16][15,16]. Egocentric and allocentric neglect, which have been reported on relatively recently, have been mentioned by many scholars as requiring differential diagnosis and intervention according to the type [17].

2. Therapy for Patients with Different Subtypes Neglect

Turgut et al. ^[18][32] examined the effects of digital therapy (eye-tracking training based on a computer program) in 32 stroke patients with neglect. They classified the patients into two groups (experimental group, 16 patients, and control group, 16 subjects). The control group received only basic rehabilitation training, and the experimental group received rehabilitation training and eye tracking (digital therapeutics, DTx) at the same time. The eye tracking training consisted of two programs: (1) when a blue square on the computer screen moved in the direction of the neglect, and (2) when the square changed to red, training to press the switch was performed (20 min/one session, total of eight sessions over two weeks). The differential evaluation of egocentric and allocentric neglect was performed using the Apple Cancellation Test. This eye-tracking task was performed simultaneously with transcranial direct current stimulation (tDCS, 1.5 and 2.0 mA) of the right posterior parietal lobe (P4). The assessments were conducted as two pre (T1, T2)/two post (T3, T4) comparisons, and there were no significant changes in the Apple Cancellation Test and the three horizontal line bisection test in both groups at T1 and T2. However, the neuropsychological test (body orientation) and the clock drawing test conducted at the last assessment of the experimental group (T4, 6 days after the end of training) showed effective significant changes in both egocentric and allocentric neglect ^[18][32]. This researchtudy had several limitations. This researchstudy was not designed to observe the long-term effects and only the effects of DTx because digital therapy was applied with tDCS. In addition, there were two pre-training assessments for both groups, but there was no statistical analysis, and both groups underwent separate rehabilitation during the intervention period. Hagiwara et al. ^[19][23] reported the effects of the training to provide visual cues using virtual reality (VR) for treating allocentric neglect in four stroke patients. The visual cues were provided using a head-mount display (HMD) type VR, Oculus Rift (Oculus VR., Inc. 19800 MacArthur Blvd Irvine, CA 92612 United States). The training consisted of a four-digit reading task followed by four actions: (1) displaying the clue stimulation, (2) obscuring the surrounding environment, (3) moving the clue stimulation, and (4) removing the obscuration. This action drew the patient’s attention to the neglected side, and then a four-digit number was read again. This training was repeated 10 times. Evaluation was performed using the apple cancellation test and line bisection with three horizontal lines, twice before and after the intervention. Patients showed fewer errors in the Apple Cancellation and Line Bisection Tests after training than those before training ^[19][23]. On the other hand, this researchtudy was conducted only once. Therefore, the author could only demonstrate the immediate effects of the treatment. Thus, the long-term effects of the treatment were not reported. The small number of subjects was another limitation. Furthermore, although researcherswe used the Apple Cancellation Test, which can be used for differential diagnosis, researcherswe cannot understand egocentric neglect by analyzing only the results for commission. Huygelier et al. ^[20][33] investigated the effect of therapy using a VR game on stroke patients with neglect. They classified the subjects into two groups (experimental group, seven stroke patients with neglect; control group, 15 normal subjects). Both groups underwent training via VR using an Oculus Rift head-mounted display (HMD) VR device (Oculus VR., Inc.). The VR training included a variety of contents for treating neglect, including stimulation and cueing on the neglected side. Six sessions (45 min/1 session/day) were performed over seven days. The assessments were the heart cancellation test in the psychometric properties of the Dutch Oxford Cognitive Screen (OCS-NL) and the letter cancellation test and figure copy task in the Behavioral Inattention Test (BIT) were administered before and after the intervention. The authors reported that presenting various stimuli and cues in the neglected area using VR was effective in accurate target recognition in egocentric neglect patients ^[20][33]. On the other hand, their study was limited by the short period (seven days) and the use of distinguishable evaluation tools (to determine whether neglect is present). In addition, although this researchtudy used the heart cancellation test of OCS-NL, which can distinguish between types of neglect, researcherswe could only indirectly check the effect on egocentric neglect by counting only omissions (the number of commissions was used to determine allocentric neglect). Schenke et al. ^[21][35] reported on the effects of digital therapy (auditory cueing and computer-based eye-tracking training) in stroke patients with egocentric neglect. This researchtudy consisted of Study 1 and Study 2. In Study 1, they classified the patients with neglect into two groups (experimental group, 11 patients; control group, 14 patients). The experimental group underwent auditory cueing with basic rehabilitation. The control received only basic rehabilitation. This researchstudy involved training detect to the movement of sound from the right to left headphones while wearing headphones (30 min/session, five times/week for three weeks). In Study 2, another eight patients with neglect (one group) were subjected to computer-based training (digital therapeutics), which provided patients with auditory cueing and eye-tracking training simultaneously. In this researchstudy, the patients had to press a button in the following situations: (1) when the target moved from right to left on the computer screen and (2) when they heard the word “here” through the headphones (30 min/session, five times/week for three weeks). In Study 1, the line bisection test (3 lines) was administered (twice) before and (twice) after training, and the Apple Cancellation Test was administered (twice) before and (once) after training. In Study 2, the visual scanning test was administered twice before and after training. After three weeks of treatment, both studies reported that the severity of egocentric neglect decreased on the line bisection, the Apple Cancellation Test, and the visual scanning test. In addition, Study 2 confirmed that digital therapeutics allowed for the simultaneous provision of auditory cues and eye-tracking training and that convenient treatment was possible ^[21][35]. On the other hand, they did not report the long-term effects of this DTx and control group in Study 2. In addition, there was no mention of allocentric neglect despite the use of the discriminative apple cancellation test. In 2021, Vestito et al. ^[22][24] examined the effects of computer-assisted cognitive training (CCT) on egocentric neglect in one patient with posterior cortical atrophy. The four-week training consisted of CCT only for the first two weeks (T1) using the Remote Monitoring Validation Engineering System (ReMoVES: University of Genova). In this researchtudy, three of the most appropriate games for the treatment of neglect in ReMoVES, including ShelfCans, OwlNest, and ChinaLanterns, which did not provide detailed information about the games, were applied (20 min/session, five times/week for two weeks). During the next two weeks (T2), CCT and tDCS (P4, 1.5 mA on the right posterior parietal lobe (P4) were delivered simultaneously (20 min/session, five times/week, for two weeks). Evaluations were carried out using Albert’s Test, line bisection test (22 lines), and Apple Cancellation Test installed in ReMoVES. Three tests were performed: one before training (T0, once) and one each at the end of T1 and T2 training. ResearchersWe only report significant results for all tests performed after T2. Therefore, they reported that the simultaneous application of both modalities (CCT and tDCS) was more effective than only CCT alone ^[22][24]. This researchtudy used the Apple Cancellation Test for a differential diagnosis. On the other hand, there was no direct relationship between the changes in egocentric and allocentric neglect. Another limitation of this researchstudy was that the improvements in egocentric neglect could be identified indirectly by the difference in omission on the Apple Cancellation Test. Turgut et al. ^[23][34] compared the effect of two DTx in sixty patients with neglect. In Study 1, 30 stroke patients (right hemisphere injury) were divided into an experimental group (20 patients with neglect) and a control group (10 patients without neglect). In Study 2, another 30 stroke patients (right hemisphere injury) were divided into an experimental group (17 patients with neglect) and a control group (13 patients without neglect). In both studies, computerized “three blocks of Posner’s covert shift of attention task (Posner task)” were used for DTx. The Posner task was performed by pressing a square on the 24-inch monitor, where the stimulus image was presented between the two blank squares on the monitor. The cue elicits a rapid response by predicting where the stimulus picture will appear. The task was performed in two sessions consisting of three blocks of 100 trials each, with two days between the sessions. Response time to a stimulus was measured. In Study 1, endogenous cueing (centrally presented arrows as cues) was provided. In Study 2, exogenous cueing (peripherally presented crosses were used as cues) was provided and compared. In both studies, response time was measured twice, three days apart. The tools used for evaluation are the Apple Cancellation Test and the line bisection test. They reported that response time and omissions decreased with endogenous cueing but not with exogenous cueing. Therefore, endogenous cueing was more effective for egocentric neglect ^[23][34]. This researchtudy had a short study period of 4 days to explore the feasibility of training with the Posner task. This suggests that a longer-term training study is needed. On the other hand, this researchstudy presented the average unilateral neglect type between groups using the Apple Cancellation Test, but the results only compared omission error and response time to identify egocentric neglect. Therefore, their study was limited by the lack of a detailed comparison of the rehabilitation effects according to neglect type. Finally, there is no detailed description of the evaluation tool used for line bisection, making it difficult to understand the exact tool.