AI and Cancer Research ── Cancer biology ── 2026-10-13
Cancer molecular biology and AI
- Abstract
Cancer classification based on molecular information is a more complex multiclass learning problem, as tumors from different tissues can present with different expression profiles and cancers with similar molecular characteristics can occur in different tissues. The Cancer Genome Atlas (TCGA) gene-expression data is used to create a practical pan-cancer classification framework that is described in this paper. The working dataset consisted of 11,062 labelled samples representing 33 cancer types, after preparation and labelling. Representation of the input was reduced to 500 gene-expression features using a feature-selection stage. A single XGBoost multiclass classifier was then trained to understand the relationship between the chosen expression profile and the 33 cancer classes. The implementation performed with an accuracy of 94.49%, precision of 94.66%, recall of 94.49%, and F1 score of 94.27% on held-out test data. To make the model easier to inspect, a feature-importance analysis based on SHAP was added. The trained model was deployed into a Streamlit dashboard, with a FastAPI backend to upload a gene-expression sample and display the predicted cancer type, probabilities of cancer classes and explanation information. The system is a research and educational prototype, not a clinically validated diagnostic system; an independent validation should be required before clinical interpretation.
Journal IF-equivalent: 0.5 (OpenAlex 2-year mean citedness, value as of 2026-10-12, retrieved 2026-10-13; not the official Clarivate IF)Reference: Lokesh N, Thara L, Kumuthaveni R. A Comparative Study of Machine Learning Models for Pan-Cancer Prediction and Explainable Feature Analysis. Int J Sci Res Sci Eng Technol. 2026 Oct 11;13(5):296-308. doi:10.32628/ijsrset2613529.Checked: Abstract only - Abstract
Histopathological growth patterns (HGPs) in colorectal liver metastases (CRLM) significantly influence patient prognosis and response to systemic therapy. While the desmoplastic HGP is associated with better outcomes, the replacement HGP often indicates a more aggressive phenotype. Current preoperative imaging cannot distinguish these patterns, and biopsies are often unreliable due to limited sampling of the tumor–liver interface. This systematic review and meta-analysis evaluates the diagnostic performance of radiomics/deep learning models for the non-invasive prediction of HGPs based on imaging data. This study was conducted following PRISMA guidelines, searching 4 databases up to August 30, 2026. Studies developing radiomics or deep learning models to predict HGPs in CRLM patients using preoperative CT or MRI were included. Methodological quality and risk of bias were assessed using the RQS 2 and QUADAS-2 tools. Pooled sensitivity and specificity were calculated using a random-effects model. Eight retrospective reports representing seven independent study populations were included in the systematic review; seven independent populations comprising 724 unique patients were included in the quantitative synthesis. The pooled sensitivity and specificity were 0.86 (95% CI: 0.81–0.90) and 0.77 (95% CI: 0.64–0.87), respectively. The SROC curve yielded an AUC of 0.845. Subgroup analysis did not reach statistical significance in most analyses, and its findings should be interpreted with caution given the limited number of studies available. The mean RQS 2.0 score was 17.1 (31%), indicating that while technical development is standardized, prospective validation and clinical application measures are currently lacking. Imaging-based AI models showed encouraging discriminatory performance for non-invasive HGP classification. However, the available evidence does not demonstrate clinical utility or improvement in patient management or outcomes. Prospective multicenter validation and decision-impact studies are required before clinical implementation can be considered.
Journal IF-equivalent: 2.7 (OpenAlex 2-year mean citedness, value as of 2026-10-12, retrieved 2026-10-13; not the official Clarivate IF)Reference: Rahnama Y, Eghbali S, Pishraft-Sabet H, Eghbali S, Kolahi S, Ahmadi-Tafti SM, et al. Imaging radiomics for predicting histopathological growth patterns in colorectal liver metastases: a systematic review and diagnostic meta-analysis. World J Surg Onc. 2026 Oct 11 [Epub ahead of print]. doi:10.1186/s12957-026-04613-y.Checked: Abstract only