SEARCH · PubMed Health

Results for “Multimodal learning”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3Linked to original sources

PubMed1979

A follow-up of alcoholics treated by multimodal therapy.

A general hospital sponsored for the psychotherapeutic treatment of alcoholism is described. In this context a multimodal approach, emphasizing methods derived from principles of learning, is applied to training the individual in new life-style skills for the management of alcoholism. Demographic characteristics of the population served by the program are of a predominantly blue collar clientele, mostly employed (72%), married (61%), and from urban centers (95%). Attrition as a major problem in evaluating results at the follow-up stage is identified and a method of reporting follow-up results taking this factor into account is presented. This method showed that under the most stringent conditions for reporting results, 36.64% of a sample of 131 alcoholics were showing improvement at 12 months, while under the least stringent condition 84% were showing some improvement over the same period.

Adult↗

PubMed2026

The role of circulating tumor DNA (ctDNA) to detect minimal residual disease in locally advanced gastroesophageal carcinoma: the BUTTERFLY study.

BACKGROUND: Despite advances in perioperative and neoadjuvant strategies, patients with locally advanced gastroesophageal cancers remain at high risk of recurrence after curative intent treatment. No validated biomarkers are available to detect minimal residual disease (MRD) or to guide post-operative risk-adapted management. Circulating tumor DNA (ctDNA) has emerged as a noninvasive tool for disease monitoring; single-parameter or tumor-informed assays, however, may lack sensitivity in low-tumor burden settings. Multimodal, tumor-agnostic approaches may overcome these limitations. METHODS: The BUTTERFLY study is a prospective, multicenter observational study enrolling patients with stage II-III gastric, gastroesophageal junction, or esophageal cancer treated with perioperative chemotherapy or neoadjuvant chemoradiotherapy followed by surgery. It evaluates the diagnostic performance and prognostic value of an academic, tumor-agnostic, multimodal ctDNA assay for MRD detection and prognostic stratification. Serial plasma samples are collected from baseline through post-operative follow-up and at relapse. Cell-free DNA is analyzed using the Agnostic Liquid Biopsy Multimodal Advancement (ALMA) platform, integrating tumor fraction estimation, somatic copy number alterations, fragmentomic features, single-nucleotide variants, and whole-genome methylation profiling. Multimodal features are combined with clinical variables using machine learning-based models to enhance MRD detection and relapse risk stratification. The primary endpoint includes sensitivity and specificity of ALMA-defined ctDNA/MRD status at the 4-8 weeks after surgery landmark, whereas secondary endpoints assess diagnostic performance at other time points and associations between ctDNA status and dynamics with disease-free survival, overall survival, treatment response, and lead time to recurrence. FUTURE PERSPECTIVES: If validated, this tumor-agnostic, multimodal ctDNA approach may enable earlier molecular relapse detection and support personalized post-operative management strategies.

circulating tumor DNA (ctDNA)↗

PubMed2026

Beyond predictive performance: A systematic review and critical methodological appraisal of AI/ML and conventional modelling strategies in breast, colorectal, and pancreatic Cancer.

BACKGROUND: Predictive modelling for cancer risk, treatment-related complications, and survival is central to precision oncology. Conventional logistic regression (LR) and Cox proportional hazards (CoxPH) regression remain widely used but are limited when modelling nonlinear interactions, high-dimensional imaging features, and multimodal clinical-metabolic predictors. Artificial intelligence (AI) and machine learning (ML) methods offer expanded capability through automated feature extraction, ensemble learning, and flexible survival modelling, but the evidence on when AI/ML adds value over conventional models across cancer sites and predictive tasks remains fragmented. OBJECTIVE: To systematically evaluate the methodological performance, validation strategies, and translational limitations of AI/ML models compared with conventional statistical models in published predictive-modelling studies for breast, colorectal, or pancreatic cancer. METHODS: PubMed, Scopus, and Web of Science were searched for studies published between January 2019 and March 2025. Two reviewers independently conducted title-and-abstract screening, full-text eligibility assessment, and PROBAST risk-of-bias assessment. Sixty-five studies (n = 907,567 participants) were narratively synthesised by cancer site, predictive task, model family, comparator, validation strategy, predictor modality, and calibration or explainability reporting. RESULTS: The 65 studies comprised breast cancer (n = 35), colorectal cancer (n = 21), and pancreatic cancer (n = 9). AI/ML superiority over LR and CoxPH was task- and data-dependent. CNN- and U-Net-based models predominated in imaging and body-composition tasks, tree-based ensembles consistently outperformed LR for tabular perioperative complication prediction, and CoxPH remained competitive, and in the largest pancreatic risk study, superior to XGBoost (C-index 0.802 vs 0.723) in well-structured datasets. PROBAST analysis-domain risk was moderate in 54 of 65 studies (83%), driven by limited external validation, sparse calibration reporting (11/65), and few decision-curve analyses (7/65). CONCLUSION: AI/ML adds the most methodological value in imaging-derived feature extraction and nonlinear perioperative prediction, while conventional regression remains preferable in large, structured datasets with linear predictors. Clinical translation requires standardised body-composition definitions, external validation, calibration assessment, decision-curve analysis, and explainability, in line with TRIPOD+AI and CLAIM standards.

Humans↗

PubMed2026

Research agenda to advance anhedonia assessment, understanding and treatment: an ECNP-GALENOS expert meeting report.

Anhedonia, broadly defined as a reduced ability to experience interest or pleasure, represents an important transdiagnostic neuropsychiatric symptom dimension which may benefit from targeted diagnostics and treatments. Different lines of research have proposed that it comprises multiple facets, including deficits in anticipatory ('wanting') and consummatory ('liking') reward processing as well as reward learning and affects different aspects of life (eg, social, physical, cognitive). Certain facets-more specifically anticipation, motivation and reward learning-likely involve blunted phasic dopaminergic signalling. However, recent meta-analytical evidence of human depression studies indicates that prodopaminergic antidepressants produce relatively small improvements in anhedonia symptoms and suggest that mechanisms beyond dopamine likely contribute to anhedonia. This stimulated an expert meeting to review the literature and define priorities for future research in anhedonia. A central key priority is developing a translational biologically-informed nomenclature and consensus that solves the current mismatch between constructs, paradigms and measures, and mechanisms, which separates discrete reward-related processes such as effort allocation, reward learning and anticipatory interest versus consummatory pleasure. Clinical research priorities are improved multimodal measurement tools, integrating neurobiological frameworks (eg, neuroimaging, electrophysiology and liquid biomarkers capturing dopaminergic, glutamatergic, opioid and immunometabolic pathways) and transdiagnostic studies across neuropsychiatric disorders and developmental stages. Innovative trial designs that explicitly target anhedonic phenotypes as a primary outcome and test mechanism-based interventions are also needed. Translational research recommendations include back-translation strategies that begin with patient-relevant phenotypes followed by the development of comparable human and animal tasks that target reward-related processes, such as effort allocation, reward learning and anticipatory interest versus consummatory pleasure, improve cross-species behavioural paradigms and enhance methodological rigour and reproducibility. Collectively, these recommendations will help refine the conceptualisation of anhedonia and advance its role within precision psychiatry as a mechanistically grounded target across multiple disorders.

Humans↗

PubMed2026

Efforts towards a precision medicine approach in juvenile idiopathic arthritis.

Juvenile idiopathic arthritis (JIA) is the commonest group of childhood arthritides. Despite the availability of advanced therapeutics, many children and young people (CYP) with JIA experience disease flares, and in some, chronic joint damage. Tailoring treatment based on unique biological profiles would benefit CYP with JIA given their variable clinical presentation and disease course. To date, biomarkers to predict treatment response are lacking. With advances in single cell technologies, we are now able to profile the genes and proteins of target tissues at unprecedented resolution to define the biological basis of disease and guide novel treatment approaches. The complex analyses and combination of biological and clinical outcome data from large datasets across disease phenotypes have become possible with the development of computational and machine learning methods. Here, we summarize the strategies to integrate data through multimodal based approaches to maximize precision medicine and research priorities for CYP with JIA.

Humans↗

PubMed2025

Q RadFusion: Hybrid Quantum Classical Radiogenomic Framework for Breast Cancer Diagnosis.

BACKGROUND AND PURPOSE: Breast cancer remains the most common cancer in women worldwide, with early and accurate diagnosis critical for patient survival. Radiogenomics integrates imaging phenotypes with genomic profiles, offering a pathway to precision diagnostics. However, existing classical machine learning models often struggle with the high dimensionality and heterogeneity of multimodal data, leading to issues in calibration and reproducibility. This study presents Q RadFusion, a hybrid quantum-classical framework designed to enhance breast cancer diagnosis by fusing mammography and genomics data. METHODS: Q RadFusion was implemented on two publicly available datasets: CBIS-DDSM (2,600 curated mammography cases, TCIA) and TCGA-BRCA (1,000 genomic profiles, GDC). Imaging preprocessing included bias-field correction, segmentation, and harmonization, while genomic data underwent normalization and imputation. Feature selection was performed using the Quantum Approximate Optimization Algorithm (QAOA), and features were mapped into a quantum Hilbert space using Variational Quantum Circuits (VQC). For multimodal fusion, ResNet encoded mammography features, and a Transformer encoded genomic features. Patient-level and site-held-out splits were used for evaluation. RESULTS: Q RadFusion achieved an AUC of 0.96 and accuracy of 94%, outperforming baselines including CNN-LSTM, ResNet + XGBoost, and multimodal Transformers. Ablation studies confirmed the contribution of quantum components, with optimal performance observed at circuit depth, qubits, and QAOA layers. The model also demonstrated improved calibration and ~ 80% fewer parameters compared to deep fusion networks. CONCLUSION: Q RadFusion demonstrates that hybrid quantum-classical radiogenomic integration can deliver accurate, reproducible, and clinically meaningful diagnostic support for breast cancer, with strong potential for future clinical translation.

Breast Cancer↗

PubMed2026

Scalable, generalizable and uncertainty-aware integration of spatial multiomics across diverse modalities and platforms with SCIGMA.

Recent advances in spatial omics technologies have enabled simultaneous profiling of transcriptomic, proteomic, epigenomic, metabolomic and imaging data at high spatial resolution, offering unprecedented opportunities to dissect tissue complexity. However, integrating these diverse and large-scale spatial multimodal datasets remains a major computational challenge. We present SCIGMA, a scalable and generalizable deep learning framework for spatial multiomics integration. SCIGMA introduces an uncertainty-aware contrastive learning objective and multiview graph neural networks to preserve modality-specific signals while learning biologically meaningful joint representations. Unlike previous methods, SCIGMA provides spatially resolved uncertainty estimates, interpretably identifying regions of biological or technical heterogeneity. SCIGMA supports integration of up to five modalities, and its modular framework is extensible to future technologies with even more modalities. It also scales to more than 1 million spatial locations, enabling analysis of high-resolution datasets such as Visium HD and Xenium Prime. We evaluated SCIGMA across 19 datasets spanning 8 modalities, 10 tissues and 9 platforms. On benchmarkable datasets, SCIGMA outperformed other methods in spatial domain detection, modality preservation, feature reconstruction and reproducibility. SCIGMA identifies biologically meaningful structures, refined spatial domains and modality-specific regulatory programs, providing a robust, flexible and future-ready solution for scalable spatial multimodal integration.

Multiomics↗

PubMed2025

MedImg: An Integrated Database for Public Medical Images.

The advancements in deep learning algorithms for medical image analysis have garnered significant attention in recent years. While several studies have shown promising results, with models achieving or even surpassing human performance, translating these advancements into clinical practice is still accompanied by various challenges. A primary obstacle lies in the availability of large-scale, well-characterized datasets for validating the generalization of approaches. To address this challenge, we curated a diverse collection of medical image datasets from multiple public sources, containing 105 datasets and a total of 1,995,671 images. These images span 14 modalities, including X-ray, computed tomography, magnetic resonance imaging, optical coherence tomography, ultrasound, and endoscopy, and originate from 13 organs, such as the lung, brain, eye, and heart. Subsequently, we constructed an online database, MedImg, which incorporates and systematically organizes these medical images to facilitate data accessibility. MedImg serves as an intuitive and open-access platform for facilitating research in deep learning-based medical image analysis, accessible at https://www.cuilab.cn/medimg/.

Humans↗

PubMed2025

A Graph Contrastive Learning Method for Enhancing Genome Recovery in Complex Microbial Communities.

Accurate genome binning is essential for resolving microbial community structure and functional potential from metagenomic data. However, existing approaches-primarily reliant on tetranucleotide frequency (TNF) and abundance profiles-often perform sub-optimally in the face of complex community compositions, low-abundance taxa, and long-read sequencing datasets. To address these limitations, we present MBGCCA, a novel metagenomic binning framework that synergistically integrates graph neural networks (GNNs), contrastive learning, and information-theoretic regularization to enhance binning accuracy, robustness, and biological coherence. MBGCCA operates in two stages: (1) multimodal information integration, where TNF and abundance profiles are fused via a deep neural network trained using a multi-view contrastive loss, and (2) self-supervised graph representation learning, which leverages assembly graph topology to refine contig embeddings. The contrastive learning objective follows the InfoMax principle by maximizing mutual information across augmented views and modalities, encouraging the model to extract globally consistent and high-information representations. By aligning perturbed graph views while preserving topological structure, MBGCCA effectively captures both global genomic characteristics and local contig relationships. Comprehensive evaluations using both synthetic and real-world datasets-including wastewater and soil microbiomes-demonstrate that MBGCCA consistently outperforms state-of-the-art binning methods, particularly in challenging scenarios marked by sparse data and high community complexity. These results highlight the value of entropy-aware, topology-preserving learning for advancing metagenomic genome reconstruction.

canonical correlation analysis↗

PubMed2025

CrossAttOmics: multiomics data integration with cross-attention.

MOTIVATION: Advances in high throughput technologies enabled large access to various types of omics. Each omics provides a partial view of the underlying biological process. Integrating multiple omics layers would help have a more accurate diagnosis. However, the complexity of omics data requires approaches that can capture complex relationships. One way to accomplish this is by exploiting the known regulatory links between the different omics, which could help in constructing a better multimodal representation. RESULTS: In this article, we propose CrossAttOmics, a new deep-learning architecture based on the cross-attention mechanism for multiomics integration. Each modality is projected in a lower dimensional space with its specific encoder. Interactions between modalities with known regulatory links are computed in the feature representation space with cross-attention. The results of different experiments carried out in this article show that our model can accurately predict the types of cancer by exploiting the interactions between multiple modalities. CrossAttOmics outperforms other methods when there are few paired training examples. Our approach can be combined with attribution methods like LRP to identify which interactions are the most important. AVAILABILITY AND IMPLEMENTATION: The code is available at https://github.com/Sanofi-Public/CrossAttOmics and https://doi.org/10.5281/zenodo.15065928. TCGA data can be downloaded from the Genomic Data Commons Data Portal. CCLE data can be downloaded from the depmap portal.

Humans↗

PubMed1990

Addiction in adolescents.

Some symptoms seen in adolescents with the disease of chemical dependence are similar to those seen in adults. Because of their age, lack of personality development, dependent family role, immaturity, and acting out of age-related behavioral tendencies, however, symptoms specific to this population occur. These may become exacerbated and telescope--intensify and shorten--the progression of the disease. A plan to solve the problem of adolescent chemical dependence must focus on education, demonstration, cooperation, prevention, intervention, habilitation, treatment, and recovery. The phenomenon of denial in a chemically dependent adolescent yields a more complex delusional system that dictates age-specific intervention approaches. Habilitation is necessary for successful adolescent treatment and recovery because what is needed is an initial process of learning, not relearning or rehabilitation. If specific adolescent issues are addressed through comprehensive, multimodality treatment approaches, then treatment and recovery outcomes for chemically dependent adolescents and their families are substantially improved. Primary care physicians must be alert to the possibility of drug use in their young patients and aware of treatment options.

Adolescent↗

PubMed1991

Biofeedback-assisted relaxation training for the aging chronic pain patient.

The older segments of the U.S. population are expanding rapidly and account for a disproportionate amount of health care, including treatment for pain-related musculoskeletal disorders. In a prospective study with objective measures and one-year follow-up, Middaugh et al. (1988) found that older patients (55-78 yr; N = 17, 76% success) treated in a multidisciplinary chronic pain rehabilitation program enjoyed a success rate equal to that of younger patients (29-48 yr, N = 20, 70% success). The current study presents additional data on these two groups of patients to compare their ability to learn the physiological self-regulation skills taught in the biofeedback/relaxation component of the multimodal program. This component included progressive muscle relaxation training, diaphragmatic breathing instruction, and EMG biofeedback. Repeated measures ANOVA showed significant increases in digital skin temperature (peripheral vasodilation) and decreases in respiration rate both within and across training sessions (p values = .04 to .0001) with no differences between age groups (p greater than .05). EMG measures for the upper trapezius ms in patients with cervical pain showed similar deficits in muscle control at evaluation and similar improvements with biofeedback training for the two age groups. These findings indicate that older pain patients responded well to the biofeedback/relaxation training component of the multimodal pain program.

Adult↗

PubMed1989

On cross-modal similarity: the perceptual structure of pitch, loudness, and brightness.

Examined how pitch and loudness correspond to brightness. In the Experiment 1, 16 Ss identified which of 2 lights more resembled each of 16 tones; in Experiment 2, 8 of the same 16 Ss rated the similarity of lights to lights, tones to tones, and lights to tones. (1) Pitch and loudness both contributed to cross-modal similarity, but for most Ss pitch contributed more. (2) Individuals differed as to whether pitch or loudness contributed more; these differences were consistent across matching and similarity scaling. (3) Cross-modal similarity depended largely on relative stimulus values. (4) Multidimensional scaling revealed 2 perceptual dimensions, loudness and pitch, with brightness common to both. A simple quantitative model can describe the cross-modal comparisons, compatible with the view that perceptual similarity may be characterized through a malleable spatial representation that is multimodal as well as multidimensional.

Attention↗

PubMed1977

Unit responses in the frog's caudal thalamus.

Single unit microelectrode recordings were made in the caudal thalamic region of the frog, Rana pipiens, from a sample of multimodal sensory units which included monocular, binocular, tactile, and spontaneous types, and which showed a wide variety of receptive field sizes, response rates, preferred stimulus sizes habituation rates, and velocity and tactile sensitivities. A number of special response properties were occasionally observed, including directional sensitivity, stationary object sensitivity, afterdischarges, and visual and tactile inhibitory fields. Computer analysis of spontaneously active units revealed four types: regular, exponential, bursting, and multimodal.

Animals↗

PubMed2026

Deep generative models in biological sequence and structure analysis and design.

Deep generative models have transformed biological sequence modeling from predictive analysis toward increasingly controllable design. Early biological applications of Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs) established latent representation learning and sequence synthesis, while recent advances in transformer-based language models, discrete diffusion, flow-matching, and multimodal generative frameworks have substantially expanded the scope of biological design. This review examines generative models for DNA, RNA, and protein sequence design, emphasizing how different model classes represent biological constraints, operate over discrete and continuous spaces, and integrate sequence, structure, and function. We compare VAEs, GANs, autoregressive and masked language models, diffusion models, and flow-based approaches across genomics, transcriptomics, and proteomics, with particular attention to controllability, long-range dependency modeling, structural grounding, generalization, and experimental utility. We further examine evaluation strategies, out-of-distribution generalization, and closed-loop design-build-test-learn workflows that connect in silico generation with empirical validation. We distinguish fundamental modality-dependent constraints including sequence discreteness, context length, structural coupling, and physical or thermodynamic requirements from architecture-dependent advantages that reflect the current state of the field. Current studies suggest that long-context models are particularly useful for genome-scale representation and sequence modeling, whereas structure-aware diffusion, flow-based, and inverse-folding approaches provide better frameworks for geometry-constrained RNA and protein design. This perspective provides a critical framework for understanding the present capabilities, limitations, and convergence of generative approaches toward reliable and experimentally grounded biological design.

Biological sequence analysis↗

PubMed2025

Generative AI Models in Time-Varying Biomedical Data: Scoping Review.

BACKGROUND: Trajectory modeling is a long-standing challenge in the application of computational methods to health care. In the age of big data, traditional statistical and machine learning methods do not achieve satisfactory results as they often fail to capture the complex underlying distributions of multimodal health data and long-term dependencies throughout medical histories. Recent advances in generative artificial intelligence (AI) have provided powerful tools to represent complex distributions and patterns with minimal underlying assumptions, with major impact in fields such as finance and environmental sciences, prompting researchers to apply these methods for disease modeling in health care. OBJECTIVE: While AI methods have proven powerful, their application in clinical practice remains limited due to their highly complex nature. The proliferation of AI algorithms also poses a significant challenge for nondevelopers to track and incorporate these advances into clinical research and application. In this paper, we introduce basic concepts in generative AI and discuss current algorithms and how they can be applied to health care for practitioners with little background in computer science. METHODS: We surveyed peer-reviewed papers on generative AI models with specific applications to time-series health data. Our search included single- and multimodal generative AI models that operated over structured and unstructured data, physiological waveforms, medical imaging, and multi-omics data. We introduce current generative AI methods, review their applications, and discuss their limitations and future directions in each data modality. RESULTS: We followed the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines and reviewed 155 articles on generative AI applications to time-series health care data across modalities. Furthermore, we offer a systematic framework for clinicians to easily identify suitable AI methods for their data and task at hand. CONCLUSIONS: We reviewed and critiqued existing applications of generative AI to time-series health data with the aim of bridging the gap between computational methods and clinical application. We also identified the shortcomings of existing approaches and highlighted recent advances in generative AI that represent promising directions for health care modeling.

Artificial Intelligence↗

PubMedDate not supplied

Attention-deficit hyperactivity disorder in adults.

It has been estimated that 30% to 70% of children who are diagnosed as having attention-deficit hyperactivity disorder (ADHD) will continue to show symptoms of the condition as adults. Since the prevalence of ADHD among school children may be 3% or more, its prevalence among adults may be 1% or 2%. The third revised edition of the Diagnostic and Statistical Manual (1987) of the American Psychiatric Association lists three essential features for the diagnosis of ADHD: "developmentally inappropriate inattention, impulsiveness, and hyperactivity." Other conditions associated with ADHD in adults include learning disabilities (or their sequelae), general anxiety disorder, drug and alcohol abuse, and dysthymic and cyclothymic disorders. Strong correlations have been found between ADHD and oppositional defiant and conduct disorders in children and an increased risk for antisocial disorders in adults. A combination of genetic, biologic, and environmental factors appears to be implicated in the etiology of ADHD. The management of adult ADHD requires a multimodal approach. The patient needs to be informed of the cause of his or her impulsive and often self-destructive behavior. Many patients will have learning difficulties that require evaluation and remediation by specialists in learning disabilities. Psychotherapy can help the patients resolve disturbances in perceptions of self and others and family therapy can address difficulties in the adult's relationships with family members. Pharmacotherapy of adult ADHD includes the use of central nervous system stimulants, such as methylphenidate, dextroamphetamine, and pemoline, of the tricyclic antidepressants imipramine and desimipramine, and of other antihypertensive, analgesic, and antimanic drugs.

Adult↗

PubMed2026

Beyond Morphology: Reframing Lymph-Node Metastasis Prediction Through Clonal Ecology-Decades-Long Genomic Instability and Polyclonal-to-Monoclonal Transitions as the Missing Dimension in Cancer.

Recent whole-genome, lineage-tracing, single-cell, and spatial studies have reshaped our understanding of tumor evolution, revealing that cancers can arise from polyclonal populations, undergo decades-long genomic instability before clinical detection, and progress through dynamic changes in subclonal composition, cellular state, and ecological organization. These findings challenge the assumption underlying morphology-based prediction models that metastatic risk can be inferred from static histological features alone. Here, we revisit lymph-node metastasis prediction in colorectal cancer through clonal ecology, integrating computational pathology with evolutionary oncology. Drawing on the subclonal switchboard model proposed in 2012 and subsequent artificial intelligence (AI)-enabled approaches for tracking dominant and dormant subclones, we synthesize evidence that metastatic potential reflects clonal ancestry, evolutionary timing, spatial niche architecture, cellular plasticity, intercellular interactions, dormancy, and treatment-driven shifts in subclonal fitness. We define five complementary methodological pillars for operationalizing clonal ecology: single-cell transcriptomics for resolving rare subclones, evolutionary trajectories, and adaptive cell states; lineage tracing and phylogenetics for reconstructing clonal ancestry and divergence; spatial transcriptomics and genomics for mapping subclonal geography and tumor-stromal-immune interactions; longitudinal liquid biopsy surveillance for monitoring residual disease, clonal turnover, and emerging resistance; and AI-enabled multimodal integration for connecting histopathology, genomics, spatial biology, and longitudinal data into predictive ecological-state models. Multiple-instance learning and pathology foundation models provide scalable computational foundations for evolution-aware prediction. Translationally, dormant subclones represent actionable reservoirs of recurrence. A longitudinal clinical and experimental study of KMT2A-rearranged acute myeloid leukemia further supports central predictions of the subclonal switchboard framework by demonstrating treatment-associated shifts in subclonal dominance, persistence of cryptic adaptive programs, and ecological rewiring during resistance and relapse. We propose clonal ecology as a measurable dimension for extending morphology-driven prediction toward integrative models that anticipate evolutionary transitions, identify therapeutic windows, and proactively constrain adaptive tumor ecosystems before resistant or metastatic subclones achieve clinical dominance.

Humans↗