Ghazal Mehri-Kakavand

Ghazal Mehri-Kakavand

Auto-generated profile, based on publicly available information

Ghazal Mehri-Kakavand

Department of Anatomy and Medical Imaging, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
4 May 2026 in Neurorehabilitation and Neural Repair
Verified
Journal Article
Open access
SLShaojun Li
Shaojun Li
BCBenjamin Chong
Benjamin Chong
GMGhazal Mehri-Kakavand
Ghazal Mehri-Kakavand
CSCatherine Shi
Catherine Shi
DTDenise Taylor
Denise Taylor
Alan WangAlan Wang
Alan Wang
Background and Purpose: Augmented reality (AR) and artificial intelligence (AI) have been applied to the assessment and rehabilitation of post-stroke spatial neglect (SN). This study aims to evaluate the feasibility, effectiveness, degree of personalization, and ecological validity of AR, AI, and hybrid methods for SN assessment and rehabilitation. Methods: PubMed, Scopus, Web of Science, Embase, CINAHL, IEEE Xplore, and the ACM Digital Library were searched up to August 2025. Two reviewers independently screened articles, extracted data, and assessed risk of bias and outcome-level certainty. Results: Of 268 screened studies published between 2000 and 2025, 15 met the inclusion criteria, including 11 assessment studies (8 AI, 1 AR, and 2 hybrid) and 4 AR rehabilitation studies, involving 567 participants. AI assessment methods demonstrated high diagnostic accuracy (area under the curve (AUC) up to 0.95), and 1 AR assessment showed strong diagnostic accuracy (AUC = 0.89). Four AR rehabilitation studies reported acceptable feasibility, with 1 randomized controlled trial (RCT) showing improvements in several neglect outcomes. Ecological validity and personalization were generally very low, and the overall certainty of evidence ranged from low to very low. Conclusion: Current evidence for AR and AI SN assessment and rehabilitation methods remains insufficient to determine their feasibility, effectiveness, ecological validity, and degree of personalization, largely due to small sample sizes, methodological heterogeneity, and the limited number of RCTs. Future research should focus on developing standardized, scalable frameworks that integrate AR with adaptive AI models, and multicenter RCTs are required to confirm clinical efficacy and long-term functional outcomes.

22 September 2026 in Journal of Imaging Informatics in Medicine
Verified
Journal Article
Open access
PDF file
GMGhazal Mehri-Kakavand
Ghazal Mehri-Kakavand
SMSibusiso Mdletshe
Sibusiso Mdletshe
MPMohamad Pursamimi
Mohamad Pursamimi
AWAlan Wang
Alan Wang
The objective of the study is to assess whether deep-learning fusion of positron emission tomography (PET) and computed tomography (CT) improves performance compared with matched unimodal PET and CT and whether different fusion architectures are associated with model performance. Five databases were searched (January 2018–8 May 2025) following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020. Studies were classified by clinical task, fusion level (early, intermediate, late, or hybrid), and fusion mechanism (fixed or learnable). Incremental and absolute performance were synthesised using robust variance estimation with publication-level clustering. Of 892 records, 103 studies (133 analytical units) were included; 30 units entered incremental synthesis, and 61 entered absolute synthesis (60 weighted). PET/CT fusion showed positive effects for Dice similarity coefficient (mean difference (MD) = 0.044; 95% confidence interval (CI), 0.027–0.060), area under the receiver operating characteristic curve (AUC; MD = 0.090; 95% CI, 0.036–0.144), and concordance index (C-index; MD = 0.038; 95% CI, 0.037–0.038); Dice and C-index were exploratory. Incremental I 2 was 0% for Dice and C-index and 94.6% for AUC. Absolute performance was 0.758, 0.820, and 0.728, respectively, with substantial heterogeneity. Intermediate and hybrid fusion had higher absolute Dice than early fusion. Learnable fusion operation did not significantly outperform fixed fusion operation. Deep-learning PET/CT fusion shows potential value in oncologic imaging, but heterogeneous evidence and limited external validation constrain its interpretation. A clearer distinction between fusion level and mechanism, supported by more standardised and externally validated studies, is needed to identify clinically meaningful fusion strategies.