AI News Archive: June 3, 2026 — Part 19
Sourced from 500+ daily AI sources, scored by relevance.
- MathMentorAI
A personalized AI coach for learning mathematics
- WHYCRYPTO
Explainable Crypto Market Intelligence
- AI Prompt Library
Free AI prompt library with 200+ categorized prompts
- ICTDesk — Open Source Live Support
AI-powered live chat with real-time visitor intelligence
- IntelPredict
Regime-aware swing trade setups with defined risk
- PromptFlow
150 AI prompts to create 30 days of content in 2 hours
- ChefQuest: Learn to Coook
Import any tiktok cooking video and we generate a guide game
- Chromix-Reflex Game
My first mobile game, built entirely with AI co-pilots
- Digital Crew — Luminous Protocol
Autonomous agents that analyze, architect, and pitch.
- Intorducing Devilukeai
the app millions have been waiting for
- AEO GEO AI — Free AI Visibility Checker
See if ChatGPT, Claude & Gemini recommend your brand. FREE
- itel S24 – 108MP AI Camera Smartphone
108MP AI Camera, Helio G91 Power & 5000mAh Battery
- Cahoni Builds Quotes, Schedules by Voice
Voice AI that builds quotes in pdf, excel & schedules crews
- AI Photo Generator - LaFoto
ai headshot generator, ai headshots, professional headshots
- UniMind
AI that turns your lecture PDFs into quiz & exam predictions
- faffno
The AI-native operating system for modern life
- Remy
The AI fitness coach that remembers your injuries.
- Tripoh
Your AI travel planner for smarter journeys
- Study Cabinet
World's First AI-Powered Education OS
- Agent Relay
Switch AI coding agents without losing your work
- Bernini AI
Bernini AI — Generate & Edit Video with ByteDance Bernini
- AiYesNo
AI agent
- Structyn
Generate production-ready apps in 15 languages
- AskJobs.ai
AI Recruitment Platform
- Lettera — AI Cover Letter Generator
ai
- Architecture Diagram AI
Generate beautiful architecture diagrams from text
- Spanish AI APIs
AI APIs for Spanish text — sentiment & summarizer
- yourcvsucks.com —Vera will roast your CV
Your complete AI career toolkit — roast to interview-ready
- Plide
Monitor AI usage, costs, and compliance in one place
- Csong AI
Create original songs from text and lyrics with AI.
- AIsa
Capability layer for the agentic economy.
- VoiceOS
Use your voice to control apps 10 faster.
- Anomaly AI
AI data analysis workspace for business datasets and reports.
- Markty AI
Hire AI workers for marketing tasks.
- AskSary
Every AI model. One workspace.
- Pilea AI
Agents turn customer signals into prioritised tasks.
- WriteABookAI
The AI-Native Book Writing Platform
- Easy MCP AI - Claude ChatGPT Connector for Wordpress
The Most Complete End-to-End MCP for WordPress
- leania.ai
MRI scan for Businesses: Find What to Cut, Replace, Automate with AI
- Daivenci AI
Predict future babies and generate professional AI art.
- Leania.ai for AI consultants
Stand out and win more clients will personalised AI recommendations
- Integrating Histology with Spatial Molecular Programs Using a Multimodal Foundation Model
Histopathological assessment remains central to cancer diagnosis and stratification, yet its mechanistic interpretation remains limited without molecular context. To address this, we developed SQUALL, a multimodal foundation model integrating histology with spatial molecular programs. For pretraining, we assembled histMol, a large-scale corpus of 1.76 billion paired histology-spatial transcriptomics spots/bins across 33 tissues and 12 platforms from 3,446 tissue sections. Following pretraining, SQUALL enables transcriptome-wide virtual biomarker profiling, prognostically relevant spatial niches discovery, and integrative disease progression modeling. Leveraging its multimodal embeddings, SQUALL identifies niches associated with tertiary lymphoid structure (TLS) maturation and ovarian cancer relapse, reconstructs molecular trajectories of breast cancer invasion across 325,112 spots, and uncovers underlying transcriptional programs. Applied to whole-slide images from 898 patients, SQUALL outperforms existing pathology foundation models in outcome prediction while enabling interpretable risk stratification. Together, these results establish spatially aligned multimodal pretraining as a new paradigm for extending molecular insights into pathology images.
- ViTAMIn-O: Democratizing computer vision-based machine learning for stem cell research
Deep Learning (DL) holds exciting potential in automating the prediction of organoid differentiation results. Nevertheless, current models lack adaptability, openness, and robustness in performance. Additionally, broad employments of predictive models in wet-lab settings necessitate machine learning expertise, often not readily available in biologically oriented laboratories. To offer an intuitive solution, we present ColabViTAMIn-O, a code-free platform together with ViTAMIn-O. ViTAMIn-O is a fully open organoid-specific DL model trained and tested on a total of 34 organoid categories, incorporating annotated images across transmitted light microscopy (TLM) modalities at single-organoid resolution. It is adaptable to downstream prediction tasks of varying dataset sizes and outperforms established models even with linear-probing. It performs reliably within a few-shot framework and is even extensible to human embryo TLM imaging data at single specimen level. By releasing our platform, centralized model hub, and datasets, we hope to encourage broader deployments of specialized DL models in stem-cell laboratories.
- Automated assessment of neonatal internal capsule maturation on T2-weighted MRI across 7T and 3T
Motivation: Quantitative assessment of neonatal internal capsule (IC) maturation remains largely reliant on qual- itative visual evaluation, limiting objectivity and scalability. Approach: We developed a fully automated 3D deep learning framework for anatomically detailed segmentation of IC subregions and PLIC myelin-related signal from structural T2-weighted MRI, trained on both high-resolution 7T and conventional 3T neonatal datasets. Volumetric and intensity-based metrics were derived, and developmental trajectories were modelled using postmenstrual age (PMA) and postnatal age (PNA), with normative modelling used to quantify individual deviations. Results: The pipeline achieved high segmentation accuracy across field strengths (Dice > 0.95, relative volume difference < 5%). IC metrics showed robust age-related changes, with volumetric measures increasing and intensity- based measures decreasing with PMA. PNA effects indicated prematurity-related modulation at equivalent maturational age. These patterns generalized to 3T, where normative modelling revealed significant deviations in preterm infants, particularly for myelin-related intensity measures. Conclusion: Structural T2-weighted MRI, combined with anatomically informed segmentation, enables quantitative and biologically meaningful assessment of neonatal IC maturation. This provides a scalable framework for studying early white matter development and supports potential clinical translation.
- Simple cumulative weighting of routine surveillance data identifies epidemic wave origins more accurately than a large language model: evidence from eight COVID-19 waves in Japan
Identifying the origin of an emerging epidemic wave within days of onset could enable targeted response before national spread, yet current methods rely on genomic sequencing that lags clinical detection by 2-4 weeks. We analysed daily COVID-19 cases from Japan's 47 prefectures across eight waves (2020-2023), aggregated into 11 regional blocks. Wave onset was defined by the first difference of the K-value (K'). Six surveillance indicators were evaluated with and without cumulative historical weighting ({lambda} = 0.75) and benchmarked against a large language model (Claude Haiku), scored by F1 against genomically confirmed origins. At 14 days after onset, cumulative weighting of peak and cumulative incidence (B1+prior, B3+prior) reached mean F1 = 0.622, exceeding the model (0.524); the gap was largest in Wave 7 (1.000 vs 0.333). Simple cumulative weighting of routine surveillance data identified wave origins more accurately than a language model, without proprietary tools or sequencing.
- Leveraging Digitization, Archiving and Artificial Intelligence to Re-examine Predictors of Sustained Mental Health Care Engagement in Ugandan First-Episode Psychosis Patients: A Study Protocol
Background: We previously examined the burden and predictors of sustained mental health care engagement in Ugandan first episode psychosis patients by retrospective chart review methods. However, the extensive requirements of chart reviews meant that we could only extract data from a random 10% sample of 1677 newly enrolled Ugandan first-episode psychosis patients at Butabika National Referral Mental Hospital in 2018. The Hekima Platform has been designed to transform handwritten files into datasets for analysis. Objectives: This study aims to: (1) utilize the Hekima Platform to transform paper-based clinical charts of all 1677 Ugandan psychosis patients enrolled at Butabika Hospital for the first time in 2018 into a standardized, anonymized longitudinal database and (2) re-examine predictors of sustained MHC engagement in this cohort. Methods: We will digitize and archive all patient charts. We will then use the Hekima Platform to extract handwritten clinical data into machine-readable text using user-trained machine learning and deep learning models and natural language processing (NLP) techniques to generate a structured, anonymized database. A minimum 10% random sample of extracted data will be manually validated using Cohen's kappa. For the analytical aim descriptive statistics bivariate analysis and multivariable logistic regression will model predictors of sustained engagement. Exploratory machine learning approaches are used as a complementary analytical strategy. Ethical approval has been obtained from the Uganda National Council for Science and Technology and Butabika Hospital's Research Ethics Committee. Expected outcomes: Patient clinical charts are a rich data source but there are extensive requirements to be able to use them for research. This study will generate the first AI-assisted standardized longitudinal database from handwritten psychiatric records in Uganda enabling well-powered analyses of predictors of MHC engagement. Findings will inform targeted interventions to improve retention in care and will offer a scalable model for mental health research in low- and middle-income countries.
- Interpretable machine learning for coeliac disease diagnosis: quantitative morphometry of duodenal biopsies
Background Coeliac disease affects approximately 1% of the global population and remains substantially underdiagnosed. Histopathological assessment of duodenal biopsies is the diagnostic gold standard but is subject to approximately 20% inter-observer disagreement. While machine learning approaches show promise, most prior work relies on black-box models with limited interpretability, restricting clinical adoption. Methods We present an interpretable pipeline that follows established histopathological criteria by extracting clinically meaningful morphological features from H&E-stained whole-slide images. Five sequential stages perform pre-processing, semantic segmentation of villi, crypts, intraepithelial lymphocytes (IELs) and enterocytes, crypt morphometry, villus length estimation via a novel polyline-based keypoint model, and coeliac disease classification using three quantitative features: IEL-to-enterocyte ratio, villus-to-crypt area ratio, and villus-length-to-crypt-depth ratio. Training and validation used data from four institutions; independent testing used 1,357 WSIs from two further institutions including one with a previously unseen scanner manufacturer, spanning five diagnostic categories: coeliac disease, normal mucosa, chronic inflammation, gastric metaplasia, and gastric heterotopia. Results Semantic segmentation achieved villus and crypt precision and recall of 87-90%. Villus length estimation correlated strongly with expert annotations (Pearson's r=0.85, mean relative error 13.5% post-calibration). All three morphological features significantly separated coeliac disease from all non-coeliac diagnostic groups across internal and external datasets (p<0.01 in all comparisons). On the test set the diagnostic classifier achieved accuracy 94.5%, PPV 92.9%, NPV 94.7%, and AUC 0.982. Conclusions This interpretable framework achieves strong multi-centre diagnostic performance while producing quantitative morphological outputs, villus length, crypt depth, and IEL-to-enterocyte ratios, that directly reflect established histopathological criteria, representing a meaningful step towards standardised AI-assisted coeliac disease diagnosis.
- Audited large language model triage for systematic review screening in national clinical guideline production: validation and prospective deployment
Title and abstract screening limit the timeliness of systematic reviews used for clinical guidelines. We evaluated audited large language model (LLM) triage at Sweden's National Board of Health and Welfare. Ten LLMs from five model families were tested on 419 Cochrane reviews comprising 26,892 records, and the selected ensemble was externally validated on 133 reviews including 8,501 records matched to planned guideline topics. The same locked model pair was then used prospectively across 24 systematic reviews in two national guideline programmes. On the 419-review selection benchmark, the selected Gemini-3-flash plus GPT-5.1 ensemble achieved 98.0% (95% CI, 97.3-98.7) mean review-level sensitivity, while topic-matched validation yielded 96.7% sensitivity (95% CI, 93.7-98.9). Prospective deployment screened 74,679 records, placed 63,858 (85.5%) in the AI-excluded pool and reduced estimated first-pass screening effort from 415 to 34 person-days. Across 600 randomly sampled AI-excluded records from the migraine and dementia programmes, none was confirmed as a final false negative after post-unblinding adjudication; across the completed 680-record audit, all 38 final retained records had been AI flagged, whereas locked blinded human consensus missed seven. These findings support locked, audited LLM triage, with human oversight and programme-specific monitoring, for systematic reviews used in national guidelines.
- Comfort with AI for HIV Prevention Among Cisgender Women in New York City
Background: Long-acting pre-exposure prophylaxis (PrEP) expands HIV prevention options for women. However, PrEP impact depends on addressing persistent gaps in awareness, access, and use. Artificial intelligence (AI) tools, including conversational agents, are being explored to advance PrEP uptake, but comfort with AI may influence their impact. Thus, we examined women's comfort with AI and its association with PrEP awareness. Methods: We analyzed self-reported data from women aged [≥]18 years in a cross-sectional survey conducted in New York City from August 2023 to August 2024. We performed descriptive analyses, applied latent class analysis to identify AI knowledge/comfort profiles, and estimated unadjusted and adjusted odds ratios to assess associations between profile membership and PrEP awareness. Results: Among 306 respondents without a diagnosis of HIV who completed AI-related survey items, the median age was 36. Most women identified as Hispanic/Latina (60%) or Non-Hispanic Black (18%), had not completed college (53%), and spoke only English or were bilingual (81%). Latent class analysis identified four AI knowledge/comfort profiles that differed by PrEP awareness, race/ethnicity, borough, prior drug use, and technology utilization. Women with varied AI knowledge, broad AI discomfort, and comfort with clinicians maintaining privacy had lower odds of PrEP awareness (OR: 0.35, 95% CI: 0.16-0.75), but this association did not persist after statistical adjustment. Conclusions: PrEP awareness and AI knowledge were limited, yet many women expressed openness to AI-enabled tools when privacy was assured. AI-enabled HIV prevention tools should prioritize trust, transparency, confidentiality, and the lived contexts of the women they intend to serve.
- To RAG, or Not to RAG? A Comparative Evaluation of Retrieval-Augmented Generation for ICD Coding of German Tumor Diagnoses
Introduction Coding tumor diagnoses from free-text clinical documentation currently requires substantial manual effort. Promising approaches for automating this process include large language mod-els (LLMs), embedding models, and retrieval-augmented generation (RAG). While previous studies often focus on a single method, we directly compare these approaches on a real-world dataset of tumor diagnosis descriptions to assess their strengths and limitations. Methods We evaluated nine different embedding models using similarity search and embedding-based classification, as well as LLM-based coding, with and without RAG, on a real-world dataset of 2,024 unique German tumor diagnosis descriptions labeled with ICD-10 and ICD-O topography codes. The retrieval knowledge base was constructed exclusively from stand-ardized Alpha-ID, ICD-10-GM, and ICD-O-3 classifications. Performance was assessed for exact (full-code) and partial (three-character) code prediction. For RAG, we evaluated base and fine-tuned versions of Llama 3.1 8B and Llama 3.3 70B. Results Qwen3-Embedding-8B, the largest embedding model, yielded the best results. It achieved 47.8% exact-match and 72.1% partial-match accuracy for ICD-10 coding with classification, and 42.7% exact-match and 73.5% partial-match accuracy for ICD-O coding with similarity search. The other embedding models, including medically specialized ones, showed varied but lower performance. RAG improved base LLM perfor-mance and outperformed embedding-based approaches on partial-match accura-cy (80.6% partial-match accuracy for ICD-10 and 75.0% for ICD-O with Llama 3.3 70B), but not on exact-match accuracy. Conclusion A direct comparison with embedding-based approaches is essential to determine whether the additional effort of RAG is justified. The strong variation in performance also highlights the importance of model selection. Further advances in embedding-based methods, potential-ly supported by larger and more diverse training data, may offer a promising direction for future work.