AI and Cancer Research ── 2026-10-08
New papers on cancer and AI, collected from journal feeds, PubMed (searched across all journals) and OpenAlex (a public index used instead of Google Scholar), judged by Jev (an AI that makes language judgements): original research, AI central, which of 13 themes, treatment or trials, and then ranked by importance up to a daily limit. Journals are not filtered by impact factor; each paper shows its journal's IF-equivalent (OpenAlex two-year mean citedness) and the date of that value. Each entry shows the full original abstract.
Read by cancer type: Cancer biology（2） · Breast（2） · Lung（0） · Pancreatic（1） · Prostate（2） · ACC（0） · Brain（1） · Ovarian（0） · Endometrial（1） · Gastric（1） · Liver（1） · Kidney（0） · Bladder（0）
Cancer molecular biology and AI
- Abstract
Accurate voxel-level segmentation of pulmonary nodules on chest CT is needed for diameter and volume measurement, growth tracking, and quantitative assessment in lung cancer screening. The task remains difficult: nodules vary in attenuation, size, and contact with vessels or pleura, and clinically relevant morphology often lies in a thin, ambiguous margin. Existing systems commonly trade contour fidelity for speed—large prompt-driven foundation models can be accurate but slow, whereas compact CNN or SSM decoders are efficient yet tend to smooth spiculated and ground-glass borders. We propose HSV-Net (Hybrid State-space–Vision Network), a dual-stream 3D network that is prompt-free after nodule localization: an SSM3D stream provides whole-patch context with linear-complexity state-space blocks, and an OrthoLoRA stream adapts frozen DINOv2 features on axial, coronal, and sagittal slices. A Hybrid Gated Mixer (HGM) fuses the two streams before FPN decoding, and Progressive Contour Learning (PCL) supervises overlap, signed-distance-field geometry, and uncertainty-weighted contour refinement. On a 10-patient QIN Lung CT/QIN-LungCT-Seg development subset ( n = 12 tumors; nested patient-level 5-fold CV), HSV-Net obtained the highest mean Dice (90.1% ± 1.0%; 95% CI [88.9, 91.3]) and Boundary F1 (86.3% ± 1.3%) among the compared methods on 128 × 128 × 64 patches, with a latency of 58 ms. A standard full-volume nnU-Net v2 baseline reached 88.6% ± 1.0% Dice on the same QIN folds, still below HSV-Net. On an independent LIDC-IDRI benchmark with expert volumetric consensus masks (186 patients/294 nodules; patient-level 5-fold CV within LIDC, not QIN→LIDC transfer), HSV-Net achieved 87.8% ± 1.1% Dice and 83.6% ± 1.3% Boundary F1, outperforming full-volume nnU-Net (87.1% ± 1.0% Dice). LUNA16 was retained only as a proxy size/localization check with diameter-matched spherical references, not as contour external validation.
Journal IF-equivalent: 2.7 (OpenAlex 2-year mean citedness, value as of 2026-10-02, retrieved 2026-10-03; not the official Clarivate IF)Reference: Cheng L, Wan X. HSV-Net: hybrid state-space and self-supervised vision fusion with progressive contour learning for CT pulmonary nodule segmentation. Front Oncol. 2026 Oct 6;16:1964406. doi:10.3389/fonc.2026.1964406.Checked: Abstract only - Abstract
This study integrates AI-guided network pharmacology with multi-scale experimental validation (in vitro and ex vivo) to elucidate the therapeutic potential of small molecules from Triticum aestivum (wheatgrass) hexane extract.AI-assisted compound screening, molecular docking, and pathway enrichment identified bioactive constituents with predicted interactions against breast cancer-associated targets, including SRC, AKT1, EGFR, TNF, and IL-6.Network analysis revealed dual modulation of oxidative stress response pathways and mitotic spindle assembly checkpoints, alongside potential interference with pro angiogenic signaling.Experimental validation confirmed these predictions: the extract exhibited significant free radical scavenging activity (DPPH assay) and reduced oxidative DNA damage, as demonstrated by the COMET assay in relevant cell models.Antimitotic activity was evident through in vitro root tip assays and ex vivo CAM (chorioallantoic membrane) assays, which also revealed suppression of angiogenesis.Ex vivo toxicity assessment using the isolated chicken eye test showed a favorable safety profile, and brine shrimp lethality assays indicated low systemic toxicity.These findings demonstrate that small molecules from a wheatgrass hexane extract exert their anticancer potential via a synergistic mechanism, attenuating oxidative stress, inducing mitotic arrest, and inhibiting angiogenesis, supported by AIdriven target prediction and validated through comprehensive biological models.This combined computational experimental framework highlights the value of artificial intelligence in accelerating the discovery and mechanistic understanding of phytochemical-based therapeutics.
Journal IF-equivalent: 0.0 (OpenAlex 2-year mean citedness, value as of 2026-10-07, retrieved 2026-10-08; not the official Clarivate IF)Reference: [No authors listed]. Identification of Anti-Oxidative, Anti-Mitotic, and Anti-Angiogenic Compounds from Triticum aestivum via AI-Based Network Pharmacology and In Vitro Assays. IJDDT. 2026 Oct 6;16(64s). doi:10.25258/ijddt.16.64s.86.Checked: Abstract only
See all 2 Cancer biology papers →
Breast cancer: treatment, trials and AI
- Abstract
BACKGROUND: Breast cancer-related lymphedema (BCRL) impairs quality of life and function, and early detection is crucial for preventive intervention. AI models, particularly machine learning and deep learning, hold promise for improving BCRL prediction and personalized risk stratification. This systematic review evaluates AI applications for predicting BCRL and, where available, summarizes how their reported performance compared with traditional diagnostic or non-AI methods. METHODS: This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. A literature search in MEDLINE, EMBASE, Cochrane Library, and Google Scholar was conducted in July 2024 using keywords related to breast cancer, lymphedema, and AI. Relevant studies were screened and selected based on predefined inclusion and exclusion criteria, with performance metrics such as accuracy, sensitivity, and specificity extracted for analysis. RESULTS: Seven studies from 2018 to 2024 were included, with 5284 patients. Among the 5 studies reporting BCRL classification performance, AI models achieved accuracies of 81%-93.75% (mean 87.2%), mean sensitivity 85.5%, and mean specificity 87.6%. Commonly reported predictors included higher body mass index, greater lymph node burden, and receipt of chemotherapy or radiotherapy. Younger age and neoadjuvant chemotherapy were associated with higher predicted risk in some studies. CONCLUSIONS: AI shows promise for early detection and risk stratification in BCRL, but current models remain exploratory rather than clinically established. Small sample sizes, data heterogeneity, absent external validation, and inconsistent calibration highlight the need for larger, multi-institutional datasets, standardized evaluation, and prospective validation before clinical integration.
Journal IF-equivalent: 1.2 (OpenAlex 2-year mean citedness, value as of 2026-10-07, retrieved 2026-10-08; not the official Clarivate IF)Reference: [No authors listed]. Artificial Intelligence for Prediction of Breast Cancer-Related Lymphedema: A Systematic Review. Plast Reconstr Surg Glob Open. [Epub ahead of print]. doi:10.1097/GOX.0000000000008132; PMCID: PMC13641666.Checked: Full text checked - Abstract
This systematic review will examine explainable and interpretable artificial intelligence models for predicting response to neoadjuvant or preoperative systemic treatment in breast cancer. Eligible studies must predict pathological complete response (pCR) or an explicitly RECIST-defined response and apply an explicit explainability or interpretability component to the prediction model. We will examine how these models are developed and validated, how their explanations are evaluated, and what evidence supports their use in clinical practice. We will search PubMed/MEDLINE, Embase.com, Scopus, Web of Science Core Collection and IEEE Xplore from inception to the final search date, without database-level language or publication-year restrictions. Backward and forward citation searching will supplement the database searches. Eligible publications will be peer-reviewed journal articles, including stable early-access and article-in-press publications. Two reviewers will independently screen records and assess full texts. Data extraction will follow the CHARMS framework, and study quality, risk of bias and applicability will be assessed using PROBAST+AI. The synthesis will describe patient populations, treatments, data modalities, model families, explanation methods, validation, calibration, leakage safeguards and evidence of clinical translation. pCR and RECIST-defined outcomes will be synthesized separately. A narrative and descriptive synthesis is planned; no meta-analysis is prespecified. This registration follows the approved protocol, version 1.0, dated 30 September 2026. The final database searches are planned within seven days after registration. Searches conducted from 12 to 21 July 2026 under an earlier scoping review protocol are superseded and will not contribute records or counts to this review or its final PRISMA flow.
Journal IF: unknown (could not be matched)Reference: [No authors listed]. Explainable and Interpretable Artificial Intelligence for Predicting Neoadjuvant Treatment Response in Breast Cancer. Open Science Framework. [Epub ahead of print]. doi:10.17605/osf.io/dzkfe.Checked: Abstract only
Lung cancer: treatment, trials and AI
No new qualifying paper for this issue.
Pancreatic cancer: treatment, trials and AI
- Abstract
Molecular subtyping of pancreatic ductal adenocarcinoma (PDAC) into basal-like and classical states is a critical prognostic determinant, yet clinical implementation remains limited by the cost and turnaround time of transcriptomic sequencing. 1 , 2 , 3 Although routine histopathology captures rich morphological features, deep learning models often lack a principled connection to gene-level molecular structure. 4 , 5 We propose a graph-constrained histology model that maps morphology-derived latent features onto a fixed, data-driven gene co-expression network for pancreatic cancer molecular subtype prediction. The gene-structured outputs are interpreted as latent features constrained by gene co-expression structure, rather than as direct estimates of patient-level gene expression or empirically recovered gene-network alignment.
Journal IF-equivalent: 50.3 (OpenAlex 2-year mean citedness, value as of 2026-10-07, retrieved 2026-10-08; not the official Clarivate IF)Reference: Leyva A, Rehman Akbar A, Niazi MKK. Gene-structured histology for deriving and predicting pancreatic cancer molecular subtypes. Signal Transduct Target Ther. 2026 Oct 6;11(1):433. doi:10.1038/s41392-026-03051-2; PMCID: PMC13639034.Checked: Abstract only
Prostate cancer: treatment, trials and AI
- Abstract
To evaluate the diagnostic performance of a noninvasive preoperative ISUP classification model for prostate cancer based on biparametric MRI (bpMRI) habitat imaging (HI). Retrospective data were collected from patients with pathologically confirmed prostate cancer in two medical centers between July 2021 and August 2024, including clinical information, pathological results, and bpMRI (T2WI + DWI) features. Patients from Center 1 were randomly assigned to the training and internal validation cohorts at a 7:3 ratio, whereas those from Center 2 were adopted as the external validation cohort. The K-means clustering algorithm was applied to segment habitat subregions, followed by radiomic feature extraction from each subregion. Predictive models were established using five machine learning algorithms, and SHapley Additive exPlanations (SHAP) analysis was performed to quantify and visualize feature importance. The predictive performance of all models for prostate cancer ISUP classification was assessed using ROC analysis. Calibration curves and decision curve analysis were further adopted to evaluate model calibration and clinical net benefit. A total of 412 patients were enrolled in this study, comprising a training cohort ( n = 196), an internal validation cohort ( n = 82), and an external validation cohort ( n = 134). Multivariable analysis identified total prostate-specific antigen (tPSA) (OR = 1.041, p < 0.001) and ADC ratio (mean ADC of tumor/normal tissue) (OR < 0.001, p < 0.001) as independent predictors for prostate cancer ISUP classification. Of the 40 established predictive models, the comprehensive model integrating clinical factors, conventional imaging features, and radiomic features derived from T2WI habitat subregion 3 yielded the optimal performance. AUC values of the comprehensive model were 0.867, 0.900, and 0.826 in the training, internal validation, and external validation cohorts, respectively. The model also achieved superior predictive efficacy and clinical net benefit compared to all other models. The bpMRI–based comprehensive model integrating clinical, imaging and habitat radiomic features enables noninvasive preoperative prediction of prostate cancer ISUP classification. It may serve as an auxiliary tool to facilitate individualized preoperative treatment decision-making.
Journal IF-equivalent: 3.8 (OpenAlex 2-year mean citedness, value as of 2026-10-02, retrieved 2026-10-03; not the official Clarivate IF)Reference: Chen S, Wang Y, Chen H, Cui J, Jiang K, Shen L, et al. MRI habitat analysis using interpretable machine learning models for ISUP grade group prediction in prostate cancer. BMC Med Imaging. 2026 Oct 6 [Epub ahead of print]. doi:10.1186/s12880-026-02870-7.Checked: Abstract only - Abstract
Efficient prostate cancer management relies on risk stratification to inform clinical decision-making, care planning, and resource allocation. However, existing approaches often fail to incorporate treatment intensity, care pathways, and downstream resource utilization needed to support healthcare systems. Using data from 4,579 men enrolled in a randomized controlled trial, this study applies STRATA-PC, a treatment-aware, survival-informed phenotyping framework that leverages longitudinal prostate-specific antigen (PSA) and digital rectal examination (DRE) trajectories along with baseline clinical characteristics to derive clinically interpretable risk phenotypes and examine how data-driven risk groups relate to real-world prostate cancer care pathways and resource utilization. STRATA-PC demonstrates strong discrimination for overall survival and prostate cancer–specific mortality and identifies three well-separated phenotypes. The low-risk phenotype (C0) is characterized by surveillance-dominated care pathways with delayed or no curative treatment, reflecting low immediate clinical urgency and sustained outpatient monitoring. The high-risk phenotype (C1) exhibits early initiation of definitive treatment, greater use of multimodal therapy, and a higher likelihood of post-treatment escalation, indicating concentrated near-term demand and increased coordination complexity. The intermediate-risk phenotype (C2) demonstrates delayed treatment initiation with a higher propensity for late multimodal escalation following extended monitoring, representing a transitional care pathway with deferred but more complex downstream resource utilization. This study examines STRATA-PC-derived phenotypes from a healthcare operations perspective by comparing survival risk, symptom burden, treatment modality selection, escalation behavior, and timing of care initiation. These findings show how patient-level risk heterogeneity translates into structured variation in care pathways and healthcare demand, providing actionable insight for resource planning.
Journal IF-equivalent: 0.7 (OpenAlex 2-year mean citedness, value as of 2026-10-07, retrieved 2026-10-08; not the official Clarivate IF)Reference: Ghavidel A, Pazos P. Linking Artificial Intelligence (AI)–derived risk phenotypes to prostate cancer care pathways: Implications for treatment intensity and healthcare operations. IISE Transactions on Healthcare Systems Engineering. 2026 Oct 5 [Epub ahead of print]. doi:10.1080/24725579.2026.2740606.Checked: Abstract only
Adenoid cystic carcinoma: treatment, trials and AI
No new qualifying paper for this issue.
Brain tumours: treatment, trials and AI
- Abstract
Glioma is the most common primary malignant brain tumor, with glioblastoma (GBM) carrying a 5-year survival rate below 10%. We analyzed 693 CGGA expression profiles, including 692 samples with available WHO grade (249 GBM and 443 lower-grade glioma [LGG]). WGCNA used all 693 profiles, whereas grade-dependent differential expression and machine-learning screening used the 692 grade-labeled samples. Screening-stage five-fold cross-validation identified ElasticNet as the numerically highest-performing reference model (AUC = 0.852), although Random Forest and Ridge performed nearly identically and yielded 12 core grade-discriminative biomarkers (ALOX15, TFRC, GCLC, DPP8, MBOAT2, NCOA4, NOX5, PIK3CB, CISD2, AMD1, TFR2, TRAF6). After panel selection, a 12-gene ElasticNet model fitted only in a reproducible stratified training set (n = 486) achieved an AUC of 0.847 (95% CI 0.790–0.896) in the internal test set (n = 206); because upstream feature discovery preceded the split, this internal estimate may remain optimistic. The fixed cross-platform subset achieved an AUC of 0.775 in the independent GSE16011 cohort. Virtual knockout (scTenifoldKnk) identified convergent perturbation of a stemness-associated program (SOX2, NES, PTPRZ1, TCF4) downstream of TFRC, GCLC, and TRAF6, and reanalysis of public CRISPR-Cas9 screening data (DepMap, 67 glioma cell lines) showed dependency signals for NCOA4 (49% of lines) and TFRC (24%). A LASSO-Cox risk score stratified patients with significant survival differences (HR = 3.75; time-dependent AUC 0.741–0.791) and independent prognostic value (multivariate HR = 2.02). These results support a biologically coherent ferroptosis-related panel while emphasizing the need for prospective validation.
Journal IF-equivalent: 7.1 (OpenAlex 2-year mean citedness, value as of 2026-10-07, retrieved 2026-10-08; not the official Clarivate IF)Reference: Wang M, Wu T, Wang D. Identification of Ferroptosis-Related Grade-Discriminative Biomarkers and a Prognostic Signature in Glioma via Integrative Machine Learning and Single-Cell Analyses. BioMedInformatics. 2026 Oct 6;6(5):87. doi:10.3390/biomedinformatics6050087.Checked: Abstract only
Ovarian cancer: treatment, trials and AI
No new qualifying paper for this issue.
Endometrial cancer: treatment, trials and AI
- Abstract
Relevance: Endometrial cancer (EC) is one of the most common malignant neoplasms of the female reproductive system in developed countries. Despite a relatively favorable prognosis when detected early, diagnosis at preclinical stages remains challenging, especially in patients with obesity, metabolic disorders, and concomitant gynecological diseases. Traditional diagnostic methods, including transvaginal ultrasound (TVUS), hysteroscopy, biopsy, and magnetic resonance imaging (MRI), have limitations in sensitivity and specificity. Recently, artificial intelligence (AI) technologies have been actively integrated into medical imaging and pathological diagnostics. AI improves EC diagnostic accuracy and reduces the impact of subjective factors.The study aimed to review the current state of AI applications in the diagnosis, preoperative staging, and risk stratification of endometrial cancer, including analyses of transvaginal ultrasound (TVUS), magnetic resonance imaging (MRI), histopathological images, and radiomic data.Materials and Methods: This systematized analysis covered scientific publications indexed in the PubMed and Cochrane Library databases over the past 10 years. The final review included 22 publications addressing the application of AI in the diagnosis, staging, and risk stratification of endometrial cancer.Results: AI model performance varied by diagnostic modality and validation approach. In individual studies, AI algorithms achieved accuracies of up to 90–95% in differentiating benign from malignant endometrial lesions using MRI data and digital histopathological images. MRI-based radiomic models assessed tumor grade, depth of myometrial invasion, lymphovascular space invasion, and lymph node metastasis. TVUS-based models achieved AUC values ranging from 0.80 to 0.90. In several studies involving hysteroscopy and histopathological analysis, model performance exceeded 90%. The integration of clinical, laboratory, and imaging data substantially improved the diagnostic and prognostic value of the models.Conclusion: Applying AI to TVUS, MRI, hysteroscopic, and digital histopathological data, as well as to preoperative assessment of endometrial cancer risk factors, may improve interpretation accuracy, reduce subjectivity, and facilitate the development of personalized therapeutic approaches. Nevertheless, most models remain at the research prototype stage. Implementing them in clinical practice requires multicenter prospective studies, standardized analytical methods, and integration of multimodal AI models into routine clinical workflows.
Journal IF-equivalent: 0.0 (OpenAlex 2-year mean citedness, value as of 2026-10-07, retrieved 2026-10-08; not the official Clarivate IF)Reference: Panshina E, Yurikova O, Toleshbayev D, Atambayeva S, Amankulov Z. APPROACHES TO THE USE OF ARTIFICIAL INTELLIGENCE IN THE DIAGNOSIS, STAGING, AND RISK STRATIFICATION OF ENDOMETRIAL CANCER: A LITERATURE REVIEW. Onkol radiol Kaz. 2026 Oct 6 [Epub ahead of print]. doi:10.52532/3135-3940-2026-3-765.Checked: Abstract only
See all 1 Endometrial papers →
Gastric cancer: treatment, trials and AI
- Abstract
Background and Study Aims The depth of early gastric cancer is an important factor in determining treatment plans, such as endoscopic resection or surgery. We developed an artificial intelligence (AI) program using deep neural networks to diagnose mucosal and submucosal cancers via endoscopic images. Performance evaluation tests were conducted for its clinical applications.
Patients and Methods: A test dataset comprising 204 early gastric cancer cases treated with endoscopic resection or surgery at our hospital between 2018 and 2021 was established. A minimum of five white-light imaging (WLI) images per case were used (median, 8; range, 5-22). The primary endpoint was the sensitivity for diagnosing mucosal cancer. The AI program was defined as useful when the lower limit of the 95% confidence interval (CI) exceeded 75%, based on previous reports on the depth diagnosis of early gastric cancer. Secondary endpoints were specificity and accuracy.
Results: In our study, the sensitivity of this program for the diagnosis of mucosal cancer was 91.2%, and the lower limit of the 95% CI of sensitivity was 85.2%, which was higher than 75%, indicating the usefulness of this program. The specificity and accuracy were 66.2% and 82.8%, respectively.
Conclusions: The AI program developed to diagnose the depth of early gastric cancer is useful and expected to be clinically applicable.
Journal IF-equivalent: 1.6 (OpenAlex 2-year mean citedness, value as of 2026-10-07, retrieved 2026-10-08; not the official Clarivate IF)Reference: Hirata S, Hamada K, Kawahara Y, Aya Y, Nishio A, Yoshikawa T, et al. Performance of an Artificial Intelligence Program for Assessing the Depth of Early Gastric Cancer Using Endoscopic Images. Endosc Int Open. 2026 Sep 17;14(CP). doi:10.1055/a-2957-0692; PMCID: PMC13641218.Checked: Full text checked
Liver cancer: treatment, trials and AI
- Abstract
Objectives: To develop and validate a CT radiomics-based machine learning model for the noninvasive prediction of HANS status in advanced HCC and to investigate the prognostic and genomic characteristics associated with HANS status. Materials and Methods: This retrospective study included patients with advanced HCC treated with TACE-based triplet therapy. Patients were followed until death or last follow-up. Tumor tissue underwent whole-exome sequencing for HANS calculation, and arterial-phase CT images were used for radiomic feature extraction. A radiomics model (HANS-Rad), a clinical model (HANS-Clin), and a combined model (HANS-RC) were developed using least absolute shrinkage and selection operator regression. Model performance was evaluated using receiver operating characteristic curve analysis. Calibration and decision curve analyses were performed to assess model reliability and clinical utility. Shapley Additive Explanations (SHAP) were used for model interpretation. Genomic analyses were conducted to explore differences between predicted HANS subgroups.
Results: A total of 129 patients were included (median age, 54 years; IQR, 44–62 years; 90% men). The HANS-High group demonstrated longer overall survival compared with the HANS-Low group (median, 23.0 vs. 10.0 months; hazard ratio, 0.60; p = 0.026). The combined HANS-RC model achieved the best performance, with an area under the curve (AUC) of 0.964 in the training set and 0.869 (95% CI, 0.716–1.000) in the test set, outperforming the radiomics-only and clinical models. The model also showed favorable calibration and net benefit. Genomic analyses revealed that HANS-High tumors exhibited higher tumor mutational burden and enrichment of immune-related and oncogenic pathways, whereas HANS-Low tumors were associated with distinct mutational patterns and pathway alterations.
Conclusions: A radiomics-based machine learning model using CT images enables the noninvasive prediction of HANS status in advanced HCC, with strong prognostic relevance and biologic interpretability supported by genomic analyses.
Journal IF-equivalent: 5.3 (OpenAlex 2-year mean citedness, value as of 2026-10-07, retrieved 2026-10-08; not the official Clarivate IF)Reference: Bian C, Qin Y, Zou H, Fan W, Li J. Interpretable CT-Based Radiomics for Prediction of High-Affinity Neoantigens Associated Immunogenic States in Advanced Hepatocellular Carcinoma Treated with TACE Plus Targeted Therapy and Immunotherapy. Cancers. 2026 Oct 6;18(19):3218. doi:10.3390/cancers18193218.Checked: Abstract only
Kidney cancer: treatment, trials and AI
No new qualifying paper for this issue.
Bladder cancer: treatment, trials and AI
No new qualifying paper for this issue.