Clear cell renal cell carcinoma, abbreviated ccRCC, accounts for the majority of kidney cancer cases worldwide, and its clinical management hinges on a pathological grading system known as WHO/ISUP grading. Under this system, pathologists examine tumor tissue under a microscope and assign a grade, typically ranging from 1 to 4, based on the prominence of nucleoli and other cellular features. The grade carries real consequences for patients: higher-grade tumors behave more aggressively, are more likely to metastasize, and influence decisions about surgery, surveillance, and adjuvant therapy. The problem is that this grading traditionally requires a biopsy or a resected specimen, meaning that oncologists and urological surgeons must often make treatment plans without knowing how aggressive the tumor actually is.

The new research, led by Dingyang Lv, Xiaomei Yao, and Jinshuai Li, with corresponding authors Ying Qiao and Weibing Shuang, set out to close that information gap. Their approach rests on a simple but powerful premise: the biological aggressiveness of a tumor leaves fingerprints in the way it enhances with contrast dye across different phases of a CT scan. Rapidly dividing, poorly differentiated tumor cells tend to have different vascular architectures, different cell densities, and different patterns of contrast wash-in and wash-out compared with indolent, well-differentiated ones. While a human radiologist can perceive some of these differences qualitatively, radiomics quantifies them at a scale and sensitivity far beyond visual assessment.

To build and test their model, the researchers assembled one of the largest cohorts yet used for this task: 884 patients with confirmed ccRCC drawn from three separate hospitals in Shanxi Province, China—the First Hospital of Shanxi Medical University, Shanxi Provincial People’s Hospital, and Shanxi Bethune Hospital. The multicenter design matters. Models trained on data from a single institution often perform impressively on that institution’s patients but collapse when confronted with scans from different CT scanners, different imaging protocols, and different patient populations. By splitting the cohort into a training set of 459 patients, an internal validation set of 198, and an external validation set of 227 drawn from other centers, the team forced their model to prove that it could generalize.

The technical pipeline began with the extraction of radiomics features from two phases of contrast-enhanced CT: the arterial phase, when contrast dye floods into the renal arteries and highlights a tumor’s blood supply, and the venous phase, when the dye has distributed more broadly through the tissue. Regions of interest were delineated around each tumor, and from these volumes the researchers computed hundreds of quantitative descriptors capturing tumor shape, first-order intensity statistics, and higher-order textures describing how pixel intensities are spatially arranged. Because raw radiomics data are notoriously noisy and redundant, the team applied feature dimensionality reduction and selection procedures, including LASSO—least absolute shrinkage and selection operator regression—to distill the raw feature set down to a stable, informative core.

What followed was an unusually systematic model-building exercise. Rather than committing to a single algorithm, the researchers fed their selected arterial-phase features, venous-phase features, and combined two-phase features into six different machine learning classifiers, generating eighteen candidate radiomics models in total. The classifier roster spanned the standard arsenal of modern machine learning: logistic regression, support vector machines, random forests, extremely randomized trees (ExtraTrees), eXtreme gradient boosting (XGBoost), and LightGBM, a gradient boosting framework known for its speed and efficiency with large feature sets. Each candidate was evaluated with receiver operating characteristic analysis, and the winner emerged clearly: an ExtraTrees model built on the combined arterial and venous phase features, denoted CTA_CTV, which achieved an area under the curve of 0.691 in the external validation set. The ExtraTrees algorithm, which injects additional randomness into how decision trees split and sample the data, has a well-earned reputation for resisting overfitting—a critical property when the number of features is large relative to the number of patients.

But the team did not stop with imaging data alone. In parallel, they used univariate and multivariate logistic regression to identify clinical and radiological characteristics—variables such as patient demographics and conventional imaging findings—that independently contributed to grading prediction, and used the ExtraTrees classifier again to construct a clinical model from these features. The final and perhaps most consequential step was fusion. Using stacking ensemble learning, a technique in which a meta-learner is trained to optimally combine the outputs of multiple base models, the researchers integrated their radiomics model and clinical model into a single decision framework, with LightGBM serving as the combiner.

The results validated the strategy. The combined model achieved an AUC of 0.703 (95% confidence interval: 0.628–0.778) in the external validation cohort, outperforming both the clinical model and the radiomics model on their own. In other words, neither the patient’s clinical picture nor the tumor’s quantitative imaging texture told the whole story alone, but stacked together they carried complementary information. Beyond raw discrimination, the team evaluated clinical usefulness using decision curve analysis, a method that quantifies the net benefit of acting on a model’s predictions across a range of decision thresholds. The combined model delivered the highest net benefit, suggesting that in practical terms it would help more patients than it would mislead.

The implications for clinical practice are substantial. Preoperative knowledge of WHO/ISUP grade could reshape how urologists counsel patients about partial versus radical nephrectomy, how they select candidates for active surveillance of small renal masses, and how they stratify risk in trials of neoadjuvant and adjuvant therapies. For patients with small kidney tumors—which are increasingly detected incidentally as abdominal imaging becomes more common—a reliable non-invasive grade estimate could mean the difference between immediate surgery and careful monitoring. The model’s modest but meaningful performance in a fully external cohort is notable in a field where reported accuracies often shrink dramatically outside the institution where the model was born.

The authors are careful to frame their findings appropriately. An AUC in the low seventies, while clinically useful, is not a substitute for pathology, and the study is retrospective in design, drawing on patients already diagnosed and treated at three centers within a single Chinese province. Broader validation across international populations, diverse scanner vendors, and prospectively collected cohorts will be needed before such models can be deployed at the bedside. The written informed consent of patients was waived owing to the retrospective nature of the work, but the study complied with the Declaration of Helsinki and received approval from the ethics boards of all three participating hospitals.

Nevertheless, the study adds to a rapidly growing body of evidence that the information needed to characterize a tumor’s biology is already embedded in images clinicians acquire every day. By demonstrating that multi-phase CT radiomics, carefully selected and intelligently fused with clinical variables, can approximate pathological grading non-invasively, the researchers have taken a concrete step toward a future in which risk stratification for kidney cancer begins the moment a scan is performed—before a single cell has been examined under a microscope. The work was published open access, and the authors declare no competing interests and received no external funding for the study.

Amidst these challenges, neuroscientists and clinicians have been on the quest for predictive tools that can identify patients at risk of needing secondary DC early enough to enable preventive measures. A pioneering study led by Dr. Zhongyi Sun at Central South University, China, unveils promising advances in this domain through the integration of radiomics and machine learning. Radiomics, a frontier technique, entails the extraction of vast quantitative data from medical images—data that describe the subtle textural, morphological, and intensity-based features of brain tissue and lesions that lie beyond the perceivable scope of the naked eye.

The research team analyzed computed tomography (CT) scans obtained before surgical hematoma evacuation in TBI patients who had initially undergone craniotomy with bone flap replacement. The cohort consisted of 65 adult individuals, some of whom later required secondary DC due to unmanageable intracranial hypertension. From each CT scan, more than a hundred distinct radiomic features were meticulously extracted, encompassing nuanced characteristics of both hemorrhagic lesions and surrounding cerebral edema. These features included shape descriptors, intensity heterogeneity metrics, and texture attributes that reflect microenvironmental heterogeneity linked to pathological processes.

To harness the predictive potential of these data, the team deployed sophisticated machine learning algorithms to construct models assessing the likelihood of secondary DC necessity. Traditional models grounded on demographic and clinical variables alone exhibited limited predictive capacity, underscoring their inadequacy in forecasting such complex outcomes. However, the models rooted in radiomic signatures demonstrated robust predictive accuracy, effectively differentiating patients destined for secondary surgical intervention. When these imaging-derived features were synergistically combined with select clinical data, model performance further improved, highlighting the complementary nature of radiomics alongside standard clinical assessments.

These findings evoke a transformative perspective in the management of TBI. As Dr. Sun elucidates, the aspiration is to transition from a reactive to a proactive treatment paradigm, enabling clinicians to identify patients at imminent risk for critical ICP elevation well before clinical deterioration ensues. This shift could allow for preemptive adjustments in monitoring strategies, optimization of pharmacological therapies, and judicious timing of surgical interventions, ultimately mitigating secondary brain injury and improving patient prognoses.

Beyond the immediate clinical advantages, the fusion of radiomics and machine learning aligns with the broader evolution of neurosurgery and critical care into data-driven disciplines. By converting routine neuroimaging into quantitative biomarkers of disease progression, this approach could standardize risk stratification in trauma centers worldwide. It holds the potential to streamline neurosurgical decision-making, foster interdisciplinary collaboration between radiologists, neurosurgeons, and data scientists, and spearhead the development of personalized therapeutic strategies tailored to the unique radiomic profile of each patient.

Moreover, the radiomics-based model underscores the critical role of artificial intelligence (AI) in reshaping future neurosurgical practice. As AI continues to penetrate medical imaging, it promises faster, more accurate interpretations that can anticipate adverse clinical trajectories. In TBI, where time is brain, such predictive analytics could mean the difference between irreversible damage and functional recovery. Automated pipelines for radiomic feature extraction and real-time risk scoring could soon be integrated into hospital workflows, augmenting clinician expertise with cutting-edge computational insights.

The implications also extend to healthcare systems and resource allocation. Earlier identification of high-risk patients may facilitate more efficient deployment of intensive care resources, optimize surgical scheduling, and reduce healthcare costs associated with emergency reoperations and prolonged ICU stays. Furthermore, by improving survival and neurofunctional outcomes, these advances have the potential to ameliorate the long-term socio-economic burden TBI imposes on patients, families, and societies.

Dr. Sun’s team acknowledges current limitations, including the relatively small sample size and single-center nature of the study. They advocate for future multicenter investigations involving larger, diverse patient populations that will validate and refine the model’s predictive accuracy. Additionally, technological advancements in automated image analysis and cross-platform compatibility will be vital in translating these research findings into real-world clinical tools seamlessly integrated with existing neuroimaging systems.

The journey to transform the paradigm of traumatic brain injury management through predictive modeling is emblematic of the broader revolution occurring at the intersection of medicine, computational science, and engineering. This pioneering study exemplifies how harnessing high-dimensional imaging data with AI can illuminate hidden biological signals, enabling clinicians to foresee and forestall life-threatening complications. As these techniques mature, they promise to usher in an era of precision neurosurgery, where individualized care plans informed by quantitative imaging could dramatically enhance patient survival and quality of life.

Dr. Sun remarks poignantly on the societal impact of this work: “Traumatic brain injury disproportionately affects young populations and carries lifelong repercussions. Developing anticipatory clinical tools is imperative not only for saving lives but also for preserving the dignity and potential of countless individuals altered by brain trauma.” This vision reflects an inspiring commitment to leveraging technological innovation to serve humanity’s most vulnerable.

In sum, the integration of radiomics and machine learning represents a groundbreaking stride in neuroscience, offering a powerful predictive lens through which the perilous course of secondary intracranial hypertension following TBI can be foreseen. This approach moves beyond conventional clinical indicators, empowering practitioners with data-driven foresight and enhancing the delicate art of neurosurgical decision-making. As the field advances, such innovations hold profound promise to fundamentally improve outcomes in traumatic brain injury – a testament to human ingenuity in the face of nature’s most daunting challenges.

Subject of Research: People

Article Title: Radiomics-based machine learning model for predicting secondary decompressive craniectomy in TBI patients after emergent craniotomy with bone flap replacement

News Publication Date: January 8, 2026

References: DOI: 10.1186/s41016-025-00423-5

Image Credits: Dr. Zhongyi Sun from Central South University, China

Keywords: Neuroscience, Traumatic injury, Brain injuries, Medical imaging, Artificial intelligence, Machine learning, Radiology, Neurosurgery, Biomarkers