Breast cancer treatment increasingly depends on precision medicine approaches, where understanding tumor response to preoperative therapies profoundly impacts clinical decision-making. Neoadjuvant therapy, administered before surgery to shrink tumors, poses a significant challenge owing to the difficulty in predicting which patients will achieve complete eradication of cancer cells, defined as pCR. Existing assessment tools often lack sufficient accuracy or require invasive procedures, calling for sophisticated imaging biomarkers that can reliably forecast treatment success.

Utilizing a large retrospective cohort of 234 patients from two different medical institutions, the study meticulously integrated diverse data sources to construct and validate predictive models. The primary dataset, consisting of 204 cases, facilitated model training, while an independent external dataset of 30 patients served as a rigorous test bed to evaluate generalizability. This dual-cohort design strengthens the findings and minimizes bias often encountered in single-center studies.

Traditional radiomics involves extracting quantitative features from medical images that characterize tumor heterogeneity, shape, and texture. However, these features alone can capture only a fraction of the full complexity inherent in tumor biology. To transcend this limitation, the researchers incorporated three-dimensional deep learning features derived from both the entire DCE-MRI volume and focused tumor regions. This fusion enabled the model to harness subtle imaging cues and spatial relationships invisible to conventional methods.

Critically, the investigation focused on two distinct temporal phases of the DCE-MRI scans—the early enhancement phase and the peak enhancement phase. These intervals reflect different physiological and vascular properties of the tumor microenvironment, providing complementary insights into tumor perfusion and permeability. By merging features from these phases, the team hypothesized that predictive accuracy could be significantly bolstered.

Feature selection methodology was rigorous and multi-tiered. The researchers first applied independent sample t-tests to reduce feature redundancy and eliminate statistically insignificant variables. Subsequently, least absolute shrinkage and selection operator (LASSO) regression refined the feature set further by penalizing less informative predictors. The final model utilized the top ten most discriminative features, balancing complexity and overfitting risks effectively.

Logistic regression models were constructed to combine the chosen features and compute the probability of a patient achieving pCR. Model performance was quantified using receiver operating characteristic (ROC) curves and the corresponding area under the curve (AUC) metric, which captures overall discriminative ability. The DeLong test provided a statistical framework to compare AUC values across different models, confirming the superiority of integrated feature approaches.

The results revealed that models based solely on traditional radiomics features from combined early and peak DCE-MRI phases already demonstrated promising predictive power. However, the performance was markedly enhanced by augmenting these with deep learning features, culminating in the RD_EP model. This integrated model achieved impressive AUCs of 0.892 on the training dataset and 0.825 on the external validation cohort, indicating robust generalization and clinical utility potential.

Further elucidation of the model’s decision process was provided by SHapley Additive exPlanations (SHAP) analysis. This state-of-the-art interpretability tool highlighted that two specific radiomics texture features were predominant contributors to prediction accuracy. Understanding which features drive model outputs is pivotal for clinical acceptance, enabling oncologists to trust and integrate AI-driven tools into routine practice.

The implications of this study are profound. By accurately identifying patients who will respond favorably to NAT, clinicians can avoid overtreatment in responders and tailor intensified regimens or alternative therapies for non-responders. This personalized approach could translate into reduced surgical morbidity, optimized use of medical resources, and improved survival rates.

Moreover, the incorporation of multi-phase imaging and diverse feature extraction methodologies exemplifies the future trajectory of oncologic imaging. It underscores the significance of temporal dynamics in tumor physiology, encouraging further research into temporal radiomics and deep learning applications across cancer types and imaging modalities.

Despite its strengths, the study has areas requiring future exploration. Expanding sample sizes, including more diverse patient populations, and integrating additional molecular or genomic data could further enhance model accuracy. Prospective validation in clinical trial settings would also solidify the model’s applicability and impact on treatment decision algorithms.

This integration of advanced computational techniques with sophisticated imaging datasets represents a paradigm shift in breast cancer management, paving the way for more intelligent, less invasive, and highly personalized therapeutic strategies. As imaging technology and artificial intelligence rapidly evolve, such interdisciplinary approaches will likely redefine the biomarker landscape in oncology.

In summary, the research led by Zhang, Cai, Cui, and colleagues exemplifies cutting-edge efforts to harness the synergy between radiomics and deep learning applied to DCE-MRI at multiple temporal phases. Their work offers a promising avenue toward real-time, non-invasive prediction of neoadjuvant therapy outcomes in breast cancer—a critical step towards genuinely personalized medicine.

The future of breast cancer care lies in precision diagnostics harnessed through multidisciplinary innovation, and this study constitutes a significant leap forward. It sets a compelling precedent for the continued fusion of imaging science, artificial intelligence, and clinical oncology for the benefit of patients worldwide.

As these technologies become increasingly accessible, patients will benefit from more tailored treatment plans, ultimately leading to better prognoses and quality of life. This model could soon become a staple in breast cancer centers globally, informing decisions and improving outcomes with unprecedented confidence.

Such research not only advances our understanding but also showcases the transformative potential of deep learning-driven radiomics as a cornerstone in the era of personalized cancer therapy.

Subject of Research: Predicting pathological complete response (pCR) to neoadjuvant therapy in breast cancer patients by integrating radiomics and deep learning features from early and peak phases of dynamic contrast-enhanced MRI.

Article Title: Predicting breast cancer response to neoadjuvant therapy by integrating radiomic and deep-learning features from early-and-peak phases of DCE-MRI.

Article References:
Zhang, Y., Cai, J., Cui, C. et al. Predicting breast cancer response to neoadjuvant therapy by integrating radiomic and deep-learning features from early-and-peak phases of DCE-MRI. BMC Cancer 25, 1747 (2025). https://doi.org/10.1186/s12885-025-15095-8

Image Credits: Scienmag.com

DOI: 11 November 2025

Historically, predicting HCC development in cirrhotic patients has relied heavily on clinical parameters such as age, sex, albumin-bilirubin (ALBI) grade, and platelet counts. These factors are encapsulated in the age-male-ALBI-platelet (aMAP) score, a clinical model that, while valuable, has shown limitations in identifying patients at the highest risk with adequate sensitivity and specificity. Recognizing this gap, the researchers embarked on an ambitious study to enhance risk stratification by harnessing the latent information embedded in radiological imaging through advanced computational techniques.

The study cohort comprised 2,411 patients enrolled over five years in a prospective, multicenter cirrhosis observational dataset in China. All participants underwent rigorous three-phase contrast-enhanced abdominal CT imaging at the point of enrollment, providing a rich repository of high-resolution liver and spleen images. These volumes were processed to extract radiomic features—a class of quantitative descriptors capturing intensity, texture, shape, and wavelet characteristics—from anatomically delineated liver and spleen regions using PyRadiomics. Complementing this approach, deep learning features were extracted through a ResNet-18 convolutional neural network architecture, fine-tuned to capture abstract representations of tissue heterogeneity and microenvironment patterns often imperceptible to the human eye.

Feature selection was a critical step given the high dimensionality of radiomic and deep learning-derived data. The investigators employed the least absolute shrinkage and selection operator (LASSO) regression technique, an advanced regularization method adept at reducing overfitting and isolating the most prognostically informative variables. This rigorous selection was essential to distill a robust set of image-derived biomarkers, enabling integration with the traditional aMAP clinical score to form the novel aMAP-CT model.

Remarkably, this composite model demonstrated superior predictive capability across multiple validation cohorts, with area under the receiver-operating characteristic curve (AUC) values ranging from 0.809 to 0.869. Such performance surpasses existing risk assessment tools, underlining the transformative potential of combining structural imaging signatures with clinical parameters. Notably, the model successfully stratified patients into distinct high-risk and low-risk groups, with the three-year cumulative incidence of HCC soaring to 26.3% within the high-risk stratum while remaining a low 1.7% in the low-risk group.

Delving deeper, the authors showcased the efficacy of a stepwise stratification approach. Initiating with the classic aMAP score to broadly categorize cirrhosis patients, the subsequent incorporation of CT-based image features refined risk categorization, pinpointing a subset of highly vulnerable individuals constituting 7% of the entire cohort. Within this nuanced subgroup, the three-year risk of HCC development was a startling 27.2%, emphasizing the precision enabled by this hybrid model. This hierarchical methodology not only streamlines patient surveillance efforts but also allocates medical resources more judiciously, focusing intensive monitoring on those most likely to benefit.

From a pathophysiological standpoint, extracting information from the spleen alongside the liver is a novel and compelling aspect of this research. The spleen often undergoes morphological and functional changes secondary to portal hypertension and systemic inflammation in cirrhosis. By incorporating spleen radiomics and deep learning signatures, the model captures an integrative view of the hepatic and extrahepatic factors contributing to tumorigenesis, thus fostering a more holistic risk assessment framework.

The use of three-phase contrast-enhanced CT imaging in this study underscores the clinical practicality of the approach. This imaging modality is widely available and routinely utilized in liver disease management, facilitating the potential rapid adoption of the aMAP-CT model in diverse healthcare settings. Furthermore, the automated nature of radiomics and deep learning feature extraction minimizes operator dependency and subjective interpretation errors, enhancing reproducibility and consistency.

In terms of clinical impact, the aMAP-CT model holds promise for revolutionizing HCC surveillance protocols. Traditional blanket screening approaches often suffer from low yield and considerable cost, not to mention the psychological and logistical burdens imposed on patients. A stratification tool of this caliber can enable personalized surveillance intervals and intensity, potentially enabling earlier tumor detection when therapeutic interventions are more effective and survival rates significantly improved.

Ethical considerations and rigorous peer review have underpinned this study’s publication in the Journal of Clinical and Translational Hepatology, ensuring adherence to robust scientific standards. The research represents a shining example of translational medicine where cutting-edge AI methodologies converge with clinical hepatology, highlighting the rapidly evolving role of computational tools in enhancing patient care.

Looking ahead, validation in other ethnic cohorts and integration with additional biomarkers such as genomics and serum molecular signatures may further refine and generalize the aMAP-CT model. Prospective trials assessing the real-world impact of AI-driven risk prediction on HCC surveillance outcomes and health economics will be essential to fully realize the potential benefits illuminated by this study.

In conclusion, this innovative study compellingly demonstrates that combining liver and spleen imaging signatures with clinical indices via AI-powered computational models vastly improves hepatocellular carcinoma risk stratification in cirrhosis patients. The aMAP-CT model heralds a new era of precision hepatology, offering a beacon of hope for earlier detection and better prognosis in a disease long shrouded in clinical uncertainty.

Subject of Research: Hepatocellular carcinoma risk stratification in cirrhosis patients through integration of radiomics and deep learning CT signatures with clinical models.

Article Title: Hepatocellular Carcinoma Risk Stratification for Cirrhosis Patients: Integrating Radiomics and Deep Learning Computed Tomography Signatures of the Liver and Spleen into a Clinical Model

News Publication Date: 1-Aug-2025

Web References:
Journal of Clinical and Translational Hepatology – https://www.xiahepublishing.com/journal/jcth
DOI – http://dx.doi.org/10.14218/JCTH.2025.00091

Image Credits: Jin-Lin Hou, Hong-Yang Wang, Rong Fan

Keywords: Hepatocellular carcinoma, Machine learning, Radiomics, Deep learning, Computed tomography, Cirrhosis, Risk stratification, AI in medicine