Clinical Evidence

ShareGAN generated synthetic amyloid-β PET from structural MRI using paired imaging from 1,009 participants across ADNI and three Chinese hospitals. Amyloid-status accuracy was 88.5% and 89.4% in two external test sets; the study explores support for selecting patients for confirmatory PET.

Evidence scope: Retrospective translational research. Synthetic PET is not a replacement for real amyloid PET or treatment eligibility assessment. This publication does not establish that MRI-to-PET synthesis is an available or approved Dr.Brain™ product module.

European Radiology. 2026;36:5125–5137.

DOI: 10.1007/s00330-025-12251-3

This review pooled 11 randomized trials involving 525 participants with mild cognitive impairment. It reported improvements in global cognition, attention and executive function, while memory effects were not statistically significant.

Evidence scope: Independent research across different VR interventions, not a trial of QuickMind VR. Differences in study quality, interventions and populations limit generalization to a marketed device.

DOI: 10.3389/fnhum.2024.1467697

The VECA study evaluated eye tracking, VR and machine learning in 201 participants in Shenzhen. The model predicted MoCA scores and classified cognitive impairment across educational groups.

Evidence scope: Single-location study; performance is specific to the studied population and model. Correlation is not diagnostic accuracy. International validation and marketed-version equivalence require confirmation.

npj Digital Medicine. 2024;7:219.

DOI: 10.1038/s41746-024-01206-5

A study of 77 early-AD patients and 58 cognitively unimpaired participants combined Dr.Brain™ hippocampal measurements with VR eye-tracking assessment. The combined model had an AUC of 0.825, compared with 0.629 for hippocampal volume alone.

Evidence scope: Single-center, cross-sectional study with education imbalance. The combined model result is not a standalone product claim or proof of population-wide diagnostic performance.

Chinese Journal of Magnetic Resonance Imaging. 2024;15(6):49–53.

DOI: 10.12015/issn.1674-8034.2024.06.007

Among 823 adults aged 65 or older in Shenzhen, machine-learning models combined AD8 responses with health characteristics to screen for cognitive impairment. The multilayer perceptron achieved an AUC of 0.83 ± 0.04, compared with 0.65 for the AD8 reference approach.

Evidence scope: Community data with cross-validation, without independent international validation. This AD8-based model does not directly establish VR hardware performance, rehabilitation efficacy or full-pathway dementia outcomes.

BMC Public Health. 2024;24:1206.

DOI: 10.1186/s12889-024-18692-7

An automated four-subtest Chinese neurocognitive battery was evaluated in 86 volunteers aged 51–82. Scores were compared with paper-based tests, and automated scoring showed correlations of 0.904–1.0 with manual scoring.

Evidence scope: A limited Chinese-language validation sample. Scoring agreement is not diagnostic sensitivity or a treatment effect. Equivalence to the marketed QuickMind or VECA configuration has not been established.

American Journal of Alzheimer’s Disease & Other Dementias. 2024;39:1–11.

DOI: 10.1177/15333175241271910

In 174 retrospectively selected acute-stroke cases, RealNow™, RAPID and ShuKun automated ASPECTS were compared with radiologist consensus based on diffusion-weighted MRI. RealNow™’s total-score agreement with the reference was ICC 0.60, and its reported AUC was 0.67; regional agreement varied.

Evidence scope: Retrospective, single-center comparison with modest agreement and variable regional performance. The study does not establish universal superiority, equivalence in every case or improved patient outcomes; clinician image review remains essential.

European Journal of Radiology. 2024;175:111435.

DOI: 10.1016/j.ejrad.2024.111435

In 594 acute ischemic stroke cases, RealNow™ and RAPID were compared on CT or MR perfusion core and penumbra volumes and imaging-based thrombectomy eligibility. Volume ICCs ranged from 0.87 to 0.99; triage decisions agreed in 159/174 CT cases and 400/420 MR cases.

Evidence scope: A retrospective software-comparison study. Volume agreement does not ensure identical triage for every patient or demonstrate improved functional outcomes. Acquisition, thresholds and clinical review remain relevant.

Cells. 2022;11(16):2547.

DOI: 10.3390/cells11162547

In the PUMCH dementia cohort, quantitative MRI of 111 patients with probable Alzheimer’s disease examined cortical volume, thickness and cognitive performance across mild, moderate and severe disease. Regional atrophy, especially in left temporoparietal and occipital areas, was associated with cognitive impairment.

Evidence scope: A cross-sectional cohort. Associations across severity groups do not establish longitudinal prediction, standalone diagnosis or treatment-response monitoring.

Current Alzheimer Research. 2022;19(8):618–627.

DOI: 10.2174/1567205019666220905145756

A retrospective comparison of 95 Alzheimer’s disease and 106 frontotemporal lobar degeneration cases integrated clinical, neuropsychological and neuroimaging findings. Dr.Brain™ volume and white-matter measurements contributed to the assessment of differing regional patterns.

Evidence scope: Retrospective, selected clinical groups. The study supports complementary differential assessment, without establishing a standalone Dr.Brain™ diagnostic classifier or population-screening accuracy.

Frontiers in Aging Neuroscience. 2022;14:981451.

DOI: 10.3389/fnagi.2022.981451

Quantitative MRI compared 18 patients with type 2 diabetes and mild cognitive impairment, 18 with diabetes without MCI, and 18 healthy controls. Dr.Brain™ regional volumes were used to explore structural differences and classification features.

Evidence scope: A small, exploratory three-group study. Reported classification performance requires larger external validation and cannot be generalized to all patients with diabetes or cognitive symptoms.

Heliyon. 2022;8:e09390.

DOI: 10.1016/j.heliyon.2022.e09390

This review describes MRI features and artificial-intelligence approaches for Alzheimer’s disease assessment, alongside the domestic and international software-approval landscape reported in 2022.

Evidence scope: Historical technical and regulatory context, not a clinical efficacy study or evidence of current approval in any market. Institutional affiliation does not imply regulatory endorsement of this website or product line.

China Medical Device Information. 2022;28(13):1–9.

A single patient with cognitive and behavioral changes underwent detailed neuropsychological assessment and multimodal imaging. The manuscript describes Dr.Brain™ regional-volume analysis as part of distinguishing frontal-variant Alzheimer’s disease from behavioral-variant frontotemporal dementia.

Evidence scope: A single illustrative case cannot establish diagnostic sensitivity, specificity or clinical efficacy. Quantitative MRI formed one part of a broader clinical and biomarker assessment.

Supplied case-report manuscript, dated 2021.

DOI: 10.3760/cma.j.cn101721-20210602-00030

Automated measurements were compared with manual measurements in 204 healthy MRI scans. A separate comparison of 34 Alzheimer’s disease cases and 34 controls examined 47 brain regions. The reported measurement agreement and regional-volume differences support further clinical investigation.

Evidence scope: A preliminary measurement study with a small disease-comparison cohort. Agreement with manual measurement is not diagnostic accuracy and does not establish longitudinal treatment benefit.

Chinese Journal of Magnetic Resonance Imaging. 2021;12(6):27–33.

DOI: 10.12015/issn.1674-8034.2021.06.006

Across three Chinese MRI datasets, localized shape analysis identified hippocampal and amygdalar abnormalities associated with Alzheimer’s disease and mild cognitive impairment. Shape measures added information beyond whole-structure volume and correlated with cognitive scores.

Evidence scope: Research evidence for structural biomarkers; it does not establish the performance of every Dr.Brain™ module or current software version.

Current Alzheimer Research. 2020;17(13):1221–1231.

DOI: 10.2174/1567205018666210218150223

A randomized controlled trial of the AKL-T01 digital intervention in children with ADHD. It provides a reference for attention-focused digital intervention design and clinical evaluation.

Evidence scope: This study evaluates AKL-T01, a different product. Its results and authorization cannot be transferred to ForcusGenius™.

DOI: 10.1016/S2589-7500(20)30017-0

Deep-learning segmentation approaches were evaluated on two open MRI datasets, followed by regional white-matter hyperintensity analysis in 213 clinical cases with cerebral ischemia or lacunes. The study examines automated lesion delineation and anatomical localization.

Evidence scope: Algorithm and regional-imaging research, rather than a prospective clinical-outcome trial. Performance on these datasets does not establish equivalence to every current Dr.Brain™ software version.

Interdisciplinary Sciences: Computational Life Sciences. 2020;12:438–446.

DOI: 10.1007/s12539-020-00398-0

High-resolution magnetic resonance spectroscopy examined postmortem frontal-cortex tissue from 11 Alzheimer’s disease cases and 11 controls. Metabolite differences, including glucose-related signals and APOE-associated patterns, were explored as a basis for further biomarker research.

Evidence scope: An ex vivo tissue study with a small sample. It does not validate a clinical screening test, quantitative MRI software or any marketed Dr.Brain™ module.

Frontiers in Aging Neuroscience. 2019;10:424.

DOI: 10.3389/fnagi.2018.00424

DEFUSE 3 studied thrombectomy in patients selected using perfusion imaging. It explains the clinical context for core, penumbra and mismatch measurements.

Evidence scope: An independent treatment trial using its specified imaging-selection workflow. It is not a RealNow™ regulatory approval or direct validation study.

DOI: 10.1056/NEJMoa1713973

An eye-tracking study of facial-emotion recognition and visual attention in autistic and comparison participants. It provides background for social-attention tasks described in the supplied child-development material.

Evidence scope: Group-level research findings are not an individual diagnostic test. This is not a validation study of AutismShield™.

DOI: 10.1177/1362361316667830

A review of MRI-based brain-age estimation and its potential as a research biomarker. It provides context for the brain-aging section of the English Brain Doctor report.

Evidence scope: Independent background literature. It does not validate the Dr.Brain™ brain-age model, provide a lifespan forecast, or supply an individual diagnostic threshold.

DOI: 10.1016/j.tins.2017.10.001

A pilot study investigated virtual social-cognition training in 30 children aged 7–16, using ten sessions over five weeks. It provides context for VR social-skills practice.

Evidence scope: Small pilot study of a different intervention and selected population. It does not establish AutismShield™ efficacy or generalize to every autistic child.

DOI: 10.1016/j.chb.2016.04.033

This meta-analysis found gains in targeted working-memory performance, with limited effects on ADHD symptoms in blinded ratings. It helps define appropriate outcome measures for cognitive training.

Evidence scope: Background evidence across other interventions. Improvements on trained tasks should not be equated with broad clinical benefit or a ForcusGenius™ treatment effect.

DOI: 10.1016/j.jaac.2014.12.010