Category: Nevin Manimala

Neural Netw. 2026 Aug 3;205(Pt B):109461. doi: 10.1016/j.neunet.2026.109461. Online ahead of print.

ABSTRACT

This work revisits diffusion Transformers for relative-depth-conditioned and joint image-depth synthesis, focusing on two bottlenecks: (1) cross-modal attention degrades around edges and structural regions, causing geometric distortions; (2) the depth branch can be overly influenced by the semantic branch, leading to over-coupling and instability. To address these issues, we propose JointConn-v2 with GCM-WFM (Gated Cross-Modal Weighted Flow Matching), a unified framework for joint and depth-conditioned image synthesis. JointConn-v2 strengthens bidirectional guidance via Swap-Q cross-attention, explicitly injects 2D relative positions and edge energy into the attention logits through a Geometric Mask Bias, performs token-level spatial selection of where and in which direction to fuse cross-modal features via Regional Routing, and controls whether and how strongly cross-branch injection occurs through a sample-level Content Gate with residual fusion. On the training side, we introduce GCM-WFM, which regresses a joint vector field in the packed sequence space and incorporates temporal, geometric, gating, and routing terms into the objective with diagonal weights. Our approach achieves a better balance between edge controllability and cross-modal consistency. The current formulation treats depth as a normalized relative geometric signal and is not intended to preserve absolute metric scale. Code is available at https://github.com/haizhu12/JointConn-v2.

PMID:42574824 | DOI:10.1016/j.neunet.2026.109461

Oral Oncol. 2026 Aug 10;181:108102. doi: 10.1016/j.oraloncology.2026.108102. Online ahead of print.

ABSTRACT

OBJECTIVE: This study aimed to evaluate the predictive value of a quantitative index measuring the extent of somatic copy-number alteration (SCNA) for malignant transformation (MT) in oral leukoplakia (OLK).

METHODS: A prospective cohort of 122 patients with OLK was followed for a median of 74 months. Whole-exome sequencing (WES) was performed on fresh-frozen tissue specimens. SCNA-L was defined as the total autosomal length of segments exceeding predefined copy-number and minimum-width thresholds. The cohort was stratified into high- and low-SCNA-L groups using a median cutoff of 3.3. The primary outcome was MT to oral squamous cell carcinoma (OSCC). Statistical analyses included Kaplan-Meier survival analysis, Cox proportional hazards models, time-dependent ROC curves, a 1,000-resample bootstrap optimism-corrected C-index, and bootstrap internal validation using calibration curves.

RESULTS: The MT rate was significantly higher in the high-SCNA-L group (26.7%) than in the low-SCNA-L group (6.5%; P < 0.01). SCNA-L remained an independent predictor of MT in the multivariable analysis (hazard ratio = 4.684, P = 0.006). The predictive model, incorporating SCNA-L and lesion type, demonstrated adequate discrimination and satisfactory calibration for mid-term prediction.

CONCLUSIONS: As an independent predictor of MT in OLK, the quantitative index SCNA-L can serve as a molecular supplement to conventional pathological grading.

PMID:42574790 | DOI:10.1016/j.oraloncology.2026.108102

Child Abuse Negl. 2026 Aug 10;180:108257. doi: 10.1016/j.chiabu.2026.108257. Online ahead of print.

ABSTRACT

BACKGROUND: Children’s increasing use of digital platforms elevates their risk for online sexual exploitation (OSE), especially among those with sexual abuse histories. Despite known risk, data on prevalence and related characteristics remain limited. Digital safety screening (DSS) offers child abuse professionals a tool to assess risk or prior exploitation but is not standardized.

OBJECTIVE: To examine age-related variation in OSE disclosure within a clinically relevant population and characterize how DSS is applied in this clinical setting.

PARTICIPANTS AND SETTING: A retrospective chart review was conducted of 307 patients aged 10-18 years who received forensic interviews (FIs) for evaluation for suspected sexual abuse by a single social worker between 2022 and 2024 at a Child Advocacy Center in the Midwestern United States.

METHODS: FI notes were analyzed through content analysis to characterize the utilization of DSS and patient disclosure or denial of OSE. Descriptive statistics revealed prevalence: logistic regression examined associations between patient age and OSE.

RESULTS: Of 307 patients, 96.1% received DSS. Among 299 who discussed OSE, 41.5% reported at least one experience. Odds of reporting increased with age, beginning at 12.0% at age 10, 27.3% at 11, and increasing to 60.0% at age 16.

CONCLUSIONS: OSE exposure likely increases with age and may vary by sex assigned at birth. Findings underscore the need for routine, standardized DSS validated across developmental stages and languages, as well as further research on prevalence and risk and prevention and response strategies.

PMID:42574786 | DOI:10.1016/j.chiabu.2026.108257

Immunobiology. 2026 Aug 6;231(5):153224. doi: 10.1016/j.imbio.2026.153224. Online ahead of print.

ABSTRACT

The effects of long-term parental microplastic exposure on offspring immunity remain unclear. This study investigated how different parental exposure patterns affect immune status in F1 offspring. Parental rats were divided into four groups: paternal, maternal, dual-parental, and control. Treated groups received polystyrene microplastics (5 mg/L) in drinking water for 90 days. F1 offspring were raised under standard conditions until 8 weeks of age, after which fecal samples, thymus, spleen, and serum were collected for immune evaluation. Maternal microplastic exposure significantly disrupted gut microbiota α-diversity, dysbiosis index, and composition in F1 offspring (all P < 0.05). These changes were accompanied by decreased RBCs and PLT counts, elevated serum TNF-α, reduced thymic CD3⁺ and CD4⁺ T cells, and downregulated IL-10 mRNA and NF-kB protein expression. Histological examination revealed blurred corticomedullary boundaries, sparse cellularity, and lymphocyte vacuolization in the thymus, along with thinning of the periarteriolar lymphatic sheaths in the spleen, further indicating immune imbalance. Paternal and dual-parental exposure also induced gut microbiota dysbiosis and reduced thymic CD4⁺ T cells. Paternal exposure upregulated Th17-related RORγt and TNF-α mRNA in the thymus, whereas dual-parental exposure decreased thymic Nrf2 protein, splenic CD4⁺/CD8⁺ ratios, and white pulp area. Thymic pathology was observed in both groups. In conclusion, our findings suggest that prolonged parental exposure to microplastics may disrupt immune homeostasis in F1 offspring by altering gut microbiota composition and modulating oxidative stress or inflammatory responses. Moreover, the extent of these effects varies with the different parental exposure patterns.

PMID:42574783 | DOI:10.1016/j.imbio.2026.153224

Eur J Radiol. 2026 Aug 5;204:113141. doi: 10.1016/j.ejrad.2026.113141. Online ahead of print.

ABSTRACT

BACKGROUND: To systematically compare the diagnostic accuracy of abbreviated breast MRI (AB-MRI) and full diagnostic protocol MRI (FDP-MRI) for breast cancer detection.

METHODS: Ovid-MEDLINE, Embase, Cochrane Library, and Web of Science were searched up to December 1, 2025. Risk of bias was assessed using the modified Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool.

RESULTS: Forty studies (14,266 patients) were included. In screening settings, the first postcontrast subtracted (FAST) protocol achieved sensitivity comparable to FDP‑MRI but significantly lower specificity; adding T2‑weighted (T2W) imaging raised specificity to a level not statistically different from FDP‑MRI. In enriched‑screening settings, FAST with or without T2W performed comparably. In diagnostic settings, FAST combined with ultrafast (UF) and T2W demonstrated significantly higher specificity than FDP‑MRI (ratio 1.61, p < 0.001), while FAST and rapid abridged multiphase (RAMP) with or without T2W or diffusion‑weighted imaging (DWI) were comparable to FDP‑MRI. T2W + DWI alone showed significantly reduced sensitivity in diagnostic settings.

CONCLUSION: The diagnostic performance of AB‑MRI protocols relative to FDP‑MRI varies by protocol and clinical setting; findings apply only to the specific regimens evaluated. FAST maintains high sensitivity, with T2W offsetting specificity limitations in screening, while FAST + UF + T2W offers superior specificity in diagnostic settings. However, most subgroup analyses included only 3-4 studies, and others fewer than three; all results therefore require cautious interpretation and validation in larger, prospective studies before clinical adoption.

PMID:42574771 | DOI:10.1016/j.ejrad.2026.113141

Annu Rev Psychol. 2026 Aug 10. doi: 10.1146/annurev-psych-100925-034807. Online ahead of print.

ABSTRACT

Large language models (LLMs) are entering psychological research both as tools and as objects of inquiry. Yet many studies apply human instruments to LLMs without establishing that the outputs are reliable or interpretable, raising the risk of measurement phantoms-statistical regularities mistaken for genuine psychological phenomena. This review argues that robust AI psychological research requires integrating two methodological traditions: psychometric validation of what a score means and causal inference standards for what the results warrant. It develops a dual-validity framework in which evidentiary demands scale with scientific ambition: from tool use through behavioral characterization and human simulation to cognitive modeling. Classifying text may require only accuracy and reliability; claiming that an LLM simulates anxiety or illuminates cognitive mechanisms requires additional evidence, including construct validity evidence and experimental controls. Progress depends on developing computational analogs of psychological constructs rather than assuming human measures automatically apply to language models.

PMID:42574761 | DOI:10.1146/annurev-psych-100925-034807

JMIR Nurs. 2026 Aug 10;9:e100775. doi: 10.2196/100775.

ABSTRACT

BACKGROUND: Patient safety investigation reports support organizational learning only when they are complete, usable, and sufficiently detailed. Conventional free-text reports are often inconsistent and may omit information needed for review and learning. Project NARRATE (Nursing AI-Refined for Accurate Transcription of Events) is a nursing-led ambient artificial intelligence workflow that uses prompts aligned with the World Health Organization Minimal Information Model for Patient Safety Incident Reporting and Learning Systems, Situation-Background-Assessment-Recommendation output, and visible missing-information cues to support structured supervisor reporting.

OBJECTIVE: This study aimed to compare the completeness and narrative quality of conventional and NARRATE-period supervisor investigation reports for falls and medication administration-related incidents.

METHODS: We conducted a retrospective pre-post document-quality study at a tertiary academic medical center in Singapore. We reviewed 150 deidentified completed supervisor investigation reports: 75 conventional reports from June to August 2025 and 75 confirmed NARRATE reports from January to March 2026. NARRATE use was voluntary, and recorded use represented approximately 40% of eligible postimplementation reports. Two blinded reviewers rated reports using a World Health Organization (WHO)-aligned completeness checklist and an adapted 8-domain Physician Documentation Quality Instrument (PDQI). Report-level comparisons were adjusted for repeated reports by the same supervisor using random-intercept linear mixed-effects models. A stratified 60-report plain-paragraph rerating examined whether visible structure influenced ratings.

RESULTS: All 150 reports were analyzed. Unadjusted mean WHO total completeness was 11.81 (SD 3.39) for conventional reports and 13.61 (SD 2.54) for NARRATE reports; the unadjusted difference was 1.80 points, and the cluster-adjusted mean difference was 1.95 (95% CI 0.91-3.00; P<.001). The adapted PDQI mean was 3.61 (SD 0.52) and 4.13 (SD 0.34), respectively; the unadjusted difference was 0.52 points, and the cluster-adjusted mean difference was 0.53 (95% CI 0.37-0.69; P<.001). In the plain-paragraph sensitivity analysis, the completeness advantage remained (adjusted mean difference 1.70, 95% CI 0.27-3.14; P=.02), as did the adapted PDQI mean advantage (adjusted mean difference 0.25, 95% CI 0.06-0.43; P=.009). Explanation, organization, and comprehensibility remained significantly higher after deformatting; actions were borderline (P=.05), and synthesis, internal consistency, and fairness/balance were not statistically significant.

CONCLUSIONS: Among voluntary early adopters, NARRATE use was associated with more complete reports and higher adapted PDQI mean scores after accounting for supervisor clustering. Because recorded use represented approximately 40% of eligible postimplementation reports and users self-selected, findings may reflect adopter and supervisor characteristics. Results support the structured workflow as a whole, not any single AI component, and do not demonstrate downstream patient-safety effects. Confirmatory evaluation under broader adoption with a concurrent, reliably classified comparison group is needed.

PMID:42574744 | DOI:10.2196/100775

J Med Internet Res. 2026 Aug 10;28:e89124. doi: 10.2196/89124.

ABSTRACT

BACKGROUND: Lifestyle modification delivered through digital self-management is central to metabolic dysfunction-associated steatotic liver disease (MASLD) care, yet long-term engagement remains the threshold beyond which clinical benefit is realized. Understanding attrition requires examining both retention (dropout) and adherence (usage quality), which are often evaluated in isolation. Existing systematic reviews of digital interventions for MASLD have focused predominantly on clinical effectiveness, leaving less attention on attrition.

OBJECTIVE: This study aimed to integrate quantitative retention metrics with qualitative adherence insights and characterize the determinants of attrition in digital MASLD self-management interventions.

METHODS: Following PRISMA-S (Preferred Reporting Items for Systematic Reviews and Meta-Analyses literature search extension) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines, a comprehensive search of five databases (PubMed, Web of Science, Embase, Cochrane Library, and CINAHL) was conducted. The initial search was conducted in June 2025, and a subsequent update was made on April 17, 2026. Eligible studies enrolled adults with MASLD or nonalcoholic fatty liver disease in structured digital self-management interventions reporting retention or adherence data. Methodological quality was assessed using the Mixed Methods Appraisal Tool. A convergent segregated design was adopted. Retention proportions were pooled using random-effects meta-analysis with logit transformation, restricted maximum likelihood estimation, and Hartung-Knapp-Sidik-Jonkman adjustment. Adherence data were synthesized through inductive framework synthesis. Findings were subsequently integrated narratively.

RESULTS: In total, 21 studies met the eligibility criteria, of which 15 (n=1,032) contributed to the quantitative synthesis. The pooled retention proportion was 80% (95% CI 72%-87%) with substantial between-study heterogeneity (I²=73.6%). App-based platforms showed the highest point estimate and the lowest within-group heterogeneity, although no subgroup difference reached statistical significance. Adherence varied widely and was not amenable to meta-analytic pooling. Thematic synthesis identified 4 interacting domains shaping adherence, namely platform and design, human support and professional integration, motivational and behavioral strategies, and patient-level characteristics. Access friction at entry, gated coaching architecture, the absence of proximal biological feedback, and psychological comorbidity recurred as attenuators of long-term engagement.

CONCLUSIONS: This review innovatively integrates retention and adherence to provide a comprehensive framework of attrition dynamics specific to MASLD. While retention compared favorably with adjacent fields, long-term adherence depended less on platform type than on accessible human support, alignment of feedback with the disease’s silent course, and psychological screening. Although evidence certainty was rated very low under Grading of Recommendations Assessment, Development and Evaluation, reflecting blinding constraints intrinsic to digital interventions and a predominance of pilot or feasibility designs, these findings carry clear real-world implications. Future interventions would benefit from establishing standardized, component-level reporting that distinguishes retention from adherence, to reliably evaluate the true therapeutic potential of digital MASLD interventions.

PMID:42574740 | DOI:10.2196/89124

J Med Internet Res. 2026 Aug 10;28:e58233. doi: 10.2196/58233.

ABSTRACT

BACKGROUND: Electronic early warning/track-and-trigger systems (EW/TTS) are crucial for patient monitoring, detecting clinical deterioration (CD), and activating rapid response teams. Understanding the current level of automation in EW/TTS is essential.

OBJECTIVE: This study aimed to provide a comprehensive overview and critical assessment of electronic EW/TTS, including automated features, algorithms, and technologies, following a published registered study protocol.

METHODS: Based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, we included studies from PubMed, Web of Science, and Scopus published between January 2010 and December 2025 describing EW/TTS applied in real-world settings, and electronic systems for CD detection. We excluded studies outside the clinical context or those that used manual scoring charts. We applied a descriptive narrative approach and a methodological quality assessment according to the Joanna Briggs Institute Critical Appraisal Checklist.

RESULTS: After removing outliers and duplicates, the query returned 1181 studies. The selected studies (n=43) reported CD as the primary objective in 24 of 44 (54.5%) reported primary objectives, with ICU transfer in 16 of 68 (23.5%) reported secondary objectives, and mortality prediction in 10 of 68 (14.7%) reported secondary objectives. EW/TTS primarily relied on vital signs and assessment scores, accounting for 42 of 67 (62.7%) reported clinical indexes to detect and predict CD effectively. Among the included systems, 18 of 43 (41.9%) had a measured automation level, 11 of 43 (25.6%) had a managed automation level, and 7 of 43 (16.3%) had a defined automation level. The studies focused on several technological domains, with a strong emphasis on data analytics (24/43, 55.8%) and hardware technologies (7/43, 16.3%). Predictive algorithms, including statistical and machine learning approaches, were used in 11 of 43 (25.6%) systems. Interoperable connectivity was reported in 30 of 43 (69.8%) systems, including connectivity with electronic health records, wearable devices, and communication platforms such as Ascom Unite, as well as integrations using standards such as Health Level Seven Fast Healthcare Interoperability Resource and Health Level Seven. Evaluations of the systems showed earlier warning (14/70, 20%), higher accuracy (12/70, 17.1%), and lower specificity (9/70, 12.9%) as the main reported outcomes. Electronic EW/TTS were most prevalent in the United States (15/43, 34.9%), the United Kingdom (6/43, 14%), and the Netherlands (6/43, 14%).

CONCLUSIONS: Current EW/TTS systems implemented a measured level of automation and primarily focused on patient monitoring in hospital surgery wards. More than half of EW/TTS featured data exchange capabilities and connectivity with other systems. Reported outcomes of EW/TTS included early warning, high accuracy, and lower specificity. However, the included evidence was limited by heterogeneous prediction targets, inconsistent performance metrics and time horizons, and poor reporting of development history and system failure. Using clinically validated wearable devices and establishing a standardized data collection framework may further improve system accuracy and reliability.

PMID:42574739 | DOI:10.2196/58233

J Psychoactive Drugs. 2026 Aug 10:1-9. doi: 10.1080/02791072.2026.2710057. Online ahead of print.

ABSTRACT

The therapeutic effects of psychedelic-assisted treatments covary with acute mystical experiences as well as enhancements in psychological flexibility. Psychological flexibility, a key construct in acceptance and commitment therapy (ACT), has broad benefits for adaptive functioning and well-being, making it a vital focus of psychedelic-assisted therapy. More than 200 participants (54.9% female, 70% Caucasian, 72% with a college degree) completed the Mystical Experiences Questionnaire (MEQ-30) addressing their most profound mystical experience as well as the Multidimensional Psychological Flexibility Index (MPFI) and their psychedelic use. Analyses revealed that psychedelic use had an indirect effect on psychological flexibility through mystical experiences, underscoring the potential role of mystical experiences in fostering psychological growth. Participants who had used a psychedelic reported significantly higher scores on the MEQ-30 and MPFI compared to those with only non-drug mystical experiences. The flipped model did not achieve statistical significance. These findings suggest that mystical experiences might act as a catalyst for increases in psychological flexibility, and they underscore the need for further research on the mechanisms linking mystical experiences and psychological flexibility, particularly in clinical contexts. Psychedelic-assisted therapy, informed by frameworks like ACT, shows promise for enhancing psychological flexibility and fostering long-term mental health improvements. Preparatory and integration sessions that emphasize flexibility may improve treatment outcomes.

PMID:42574736 | DOI:10.1080/02791072.2026.2710057