The clinical dilemma at the heart of the research is well known to oncologists. Patients with locally advanced gastric cancer frequently receive neoadjuvant chemotherapy before surgery to shrink tumors and improve resection outcomes. After the operation, many are offered additional adjuvant chemotherapy, yet the evidence supporting this second round of treatment in patients who have already received preoperative therapy remains contested. Some trials suggest benefit, others show marginal gains, and clinicians are left weighing toxic side effects against uncertain rewards for each individual patient. Current decisions rely heavily on post-surgical pathological staging, a blunt instrument that captures little of the biological and clinical nuance that determines whether a particular patient will relapse.

To address this gap, a multi-institutional team led by researchers at the National Cancer Center/Cancer Hospital of the Chinese Academy of Medical Sciences, together with colleagues from Tianjin Medical University Cancer Institute and Hospital, Beijing Friendship Hospital, and the Cancer Hospital of China Medical University, assembled a retrospective cohort of 1,150 patients treated with neoadjuvant chemotherapy and radical gastrectomy across four Chinese centers. Rather than relying on a single algorithm, the team embraced a large-scale combinatorial strategy. They first employed eleven different machine learning learners, each of which identified its own optimal subset of predictive features from the clinical data. These eleven feature subsets were then crossed with the eleven learners, producing 121 candidate prediction models that competed against one another for predictive supremacy.

The evaluation process was rigorous and multidimensional. The researchers assessed each candidate using the concordance index, a standard measure of how well a survival model ranks patients by risk; time-dependent receiver operating characteristic curves, which capture discrimination at specific follow-up horizons; time-dependent calibration curves, which test whether predicted probabilities match observed outcomes over time; and decision curve analysis, which quantifies the net clinical benefit of acting on the model’s predictions. Out of the 121 contenders, one model clearly rose above the rest: the GAMB-AORSF model, an acronym that combines a Generalized Additive Models via Gradient Boosting-selected feature subset with an Accelerated Oblique Random Survival Forest learner.

The technical architecture of the winning model reflects two complementary strengths. Gradient-boosted generalized additive models are highly effective at screening large sets of candidate variables and selecting a compact, informative feature subset without imposing rigid linear assumptions. The accelerated oblique random survival forest, in turn, is a tree-based ensemble method designed specifically for censored survival data. Unlike conventional random forests, which split data on single variables at each node, oblique random survival forests consider linear combinations of variables, allowing them to capture more complex interaction structures in the data. The “accelerated” designation refers to computational optimizations that make this demanding approach feasible at scale. This pairing proved exceptionally powerful for modeling time-to-recurrence outcomes.

The performance numbers are striking. In the training cohort, the GAMB-AORSF model achieved a concordance index of 0.864, and it maintained robust discrimination in two independent validation cohorts, with C-indices of 0.813 and 0.789. A C-index of 0.5 would correspond to random guessing, while 1.0 represents perfect ranking; values above 0.8 in external validation are rarely achieved in oncology prediction models, particularly those built from routinely collected clinical variables rather than expensive molecular profiling. The model also successfully stratified patients into distinct risk groups whose survival trajectories diverged substantially, providing a practical foundation for treatment personalization.

The most clinically consequential finding, however, came from the model’s use as a treatment-selection instrument. Because decisions about adjuvant chemotherapy are not randomized in routine practice, the researchers applied inverse probability of treatment weighting, a statistical technique that simulates the balance of a randomized trial by reweighting treated and untreated patients according to their probability of receiving treatment. Within the model-defined risk strata, a clear pattern emerged. High-risk patients derived significant survival benefits from adjuvant chemotherapy: across the cohorts, the treatment extended three-year restricted mean survival time by five to seven months and produced an absolute reduction in recurrence risk of 14 to 20 percent. In contrast, low-risk patients showed no significant survival improvement from additional chemotherapy, implying that many of these patients may be enduring weeks of toxic treatment with little to show for it.

Restricted mean survival time deserves particular attention as an outcome measure. Unlike hazard ratios, which can be difficult to interpret when treatment effects vary over time, restricted mean survival time quantifies the average amount of life or disease-free time gained over a fixed horizon, expressed in familiar units of months. For a high-risk patient, gaining five to seven months of cancer-free survival is a clinically meaningful benefit that most would consider worth the side effects of chemotherapy. For a low-risk patient, an unmeasurable benefit against real toxicity argues for de-escalation. The model effectively converts a population-level debate into an individual-level decision.

The study’s use of SHAP values, a game-theoretic approach to explaining machine learning predictions, further addresses a common criticism of artificial intelligence in medicine: the black-box problem. By quantifying each feature’s contribution to individual predictions, SHAP analysis allows clinicians to see why a particular patient was classified as high or low risk, fostering the transparency needed for clinical adoption. The researchers also evaluated their model against the PROBAST framework for prediction model risk of bias and reported their work following TRIPOD guidelines, signaling attention to methodological standards that many published clinical prediction tools lack.

The implications extend beyond gastric cancer. Neoadjuvant chemotherapy followed by surgery is increasingly the standard of care for multiple solid tumors, and in each setting the same question arises: after a response to preoperative therapy, does everyone still need postoperative treatment? The Chinese team’s approach, combining exhaustive model search with causal inference methods to estimate treatment effects within risk strata, offers a template that could be adapted to esophageal, rectal, and other cancers. The framework of using machine learning not merely to predict outcomes but to guide de-escalation decisions represents a shift toward genuinely precision-guided perioperative oncology.

Caveats remain. The study is retrospective, and despite external validation across multiple centers, the findings arise from a Chinese patient population treated largely with regimens such as SOX, FLOT, and DOC, which may limit immediate generalizability to other populations and treatment protocols. Prospective validation, ideally in a randomized trial design where the model is used to stratify treatment assignment, will be needed before the GAMB-AORSF model can change guidelines. Integration with emerging biomarkers such as circulating tumor DNA, which the authors note as a future direction, could further sharpen the model’s risk distinctions.

Funding for the work came from the National Natural Science Foundation of China, the Beijing Natural Science Foundation, the Capital Health Development Research Special Fund, and Tianjin medical research programs. The corresponding authors are Quan Xu, Guoliang Zheng, and Yantao Tian, with Xu Liu, Peng Jin, Peng Wang, Xinxin Shao, and Haikuo Wang contributing equally as first authors. The study received institutional review board approval and written informed consent from all participants, and the authors declared no competing interests.

For now, the message for patients and clinicians is one of cautious optimism. A tool that reliably separates those who need continued treatment from those who do not could spare hundreds of thousands of patients worldwide unnecessary chemotherapy each year, while concentrating intensive treatment on those most likely to relapse. As machine learning models like GAMB-AORSF move toward prospective testing, the era of one-size-fits-all postoperative care in gastric cancer may finally be drawing to a close.

Cervical cancer remains a major global health challenge, with approximately 94% of deaths occurring in low- and middle-income countries, where in 2022 alone, 350,000 women succumbed to the disease. Radiotherapy constitutes the cornerstone of curative treatment for cervical cancer; however, only an estimated 10% of patients in low-income nations and 40% in middle-income countries receive this vital therapy. The scarcity of healthcare professionals skilled in radiotherapy planning has been a critical impediment in scaling up access to treatment.

The AI-based radiotherapy planning technology evaluated in this study addresses these challenges by automating the intricate process of radiotherapy plan formulation. Traditionally, radiotherapy planning is painstakingly complex, requiring oncologists to delineate tumors and surrounding at-risk tissues on computed tomography (CT) scans, alongside physicists’ meticulous calculations to determine optimal radiation beam configurations. This process, typically extending over several days or weeks, is resource-intensive and depends heavily on specialized personnel.

Known as the Radiotherapy Planning Assistant, the AI software developed by the University of Texas MD Anderson Cancer Center under Professor Laurence Court’s leadership leverages advanced machine learning algorithms to identify target structures and compute optimal radiation beam arrangements autonomously. This innovation dramatically truncates planning times to just over an hour, potentially revolutionizing radiotherapy delivery by reducing patient wait times and enabling broader access in resource-constrained settings.

The ARCHERY trial, encompassing over 1,000 cancer patients across India, South Africa, Jordan, and Malaysia, rigorously evaluated the AI’s capability in delivering gold-standard radiotherapy plans for cervical, prostate, and head and neck cancers. Notably, the trial demonstrated that the AI technology generated radiotherapy plans meeting international best-practice standards in more than 95% of cervical cancer cases, an extraordinary benchmark suggesting readiness for routine clinical implementation.

For prostate cancer, the AI-generated plans achieved a high standard in 85% of cases, an outcome considered clinically acceptable for regular use. These findings underscore the AI system’s versatility and promise for enhancing radiotherapy planning across diverse cancer types and healthcare settings. Results pertaining to head and neck cancer are forthcoming, anticipated to further validate the technology’s utility.

Professor Ajay Aggarwal, chief investigator based at LSHTM and Guy’s & St Thomas’ NHS Trust, emphasized the transformative potential of AI in meeting the World Health Organization’s cervical cancer elimination goals by improving treatment access worldwide. The abbreviated planning process not only expedites therapy initiation but also helps alleviate the global shortage of specialized radiotherapy professionals, a persistent barrier to care.

Co-investigator Professor Mahesh Parmar from UCL highlighted that radiotherapy facilitates the cure of approximately 40% of all cancer cases yet remains inaccessible to millions. He underscored the unique scale and rigor of the ARCHERY trial, contrasting it with prior technology assessments typically limited to small, high-income country cohorts, which carry risks of bias and limited generalizability. This trial represents a landmark in validating AI interventions in oncology globally.

The clinical oncology community has greeted these results with considerable enthusiasm. ESTRO President Professor Matthias Guckenberger praised the potential of AI to not only maintain but potentially enhance the precision of radiotherapy treatments, which demand accurate targeting to maximize tumor eradication while minimizing collateral damage to healthy tissues. He remarked on the economic benefits, noting that while radiotherapy machines are capital intensive, optimized utilization through AI could improve cost-effectiveness and accessibility.

The ARCHERY trial was generously funded by the National Institutes of Health (NIH) in the United States, the Rising Tide Foundation, and the UK’s Medical Research Council, with coordination by the UCL Innovative Clinical Trials Unit. The extensive international collaboration included leading centers such as Cape Town University, Stellenbosch University, Tata Medical Centre (Kolkata), Tata Memorial Hospital (Mumbai), University of Malaya Medical Centre, King Hussein Cancer Center, Ghent University, and the UK Radiotherapy Trials Quality Assurance Group, reflecting a truly global effort to advance cancer care.

Radiotherapy planning’s inherent complexity lies in both anatomical contouring and dosimetric optimization — tasks meticulously performed by oncologists and physicists who manually segment tumor volumes and organs at risk within volumetric scans before defining radiation beam parameters. The Radiotherapy Planning Assistant harnesses convolutional neural networks and sophisticated dose calculation models to replicate these functions with remarkable speed and accuracy, promising to standardize care quality regardless of geographic or economic constraints.

By enabling extensive deployment of high-quality radiotherapy plans, especially in regions where skilled personnel are scarce, this AI technology can drastically improve treatment reach and outcomes. The resultant reduction in planning time not only enhances clinical workflow efficiencies but also has profound implications for patient survival and quality of life in both high-resource and underserved settings.

As AI continues to integrate into healthcare, trials like ARCHERY affirm that technological advances must be validated through large-scale, multicenter studies to establish unequivocal benefits for patients and practitioners alike. This ensures that AI-driven innovations deliver on their promise to democratize access, improve efficacy, and optimize resources in the relentless global fight against cancer.

Subject of Research:
Artificial intelligence-driven radiotherapy planning for cervical and prostate cancers.

Article Title:
AI-Powered Radiotherapy Planning Revolutionizes Cancer Treatment Access Globally

News Publication Date:
Not specified

Web References:
Not specified

References:
Not specified

Image Credits:
Not specified

Keywords:
Radiotherapy, Cancer Treatment, Cervical Cancer, Prostate Cancer, Artificial Intelligence, Machine Learning, Medical Technology, Global Health, Oncology, Clinical Trials

Soft-tissue sarcomas, though rare, present a formidable challenge in oncological practice due to their heterogeneous nature and variable prognosis. Traditionally, predicting outcomes in these tumors has relied heavily on clinical characteristics and histopathological assessment. However, the study conducted by Michot et al. demonstrates how deploying deep learning tools can significantly refine risk stratification, thereby transforming the management of such cancers.

The researchers embarked on a comprehensive analysis that utilized large datasets encompassing digital pathology images of both tumor regions and the surrounding margin areas. By training convolutional neural networks (CNNs) on this annotated data, they sought to extract intricate features that might go unnoticed in conventional analyses. This meticulous training process highlighted not only the tumor’s intrinsic characteristics but also the critical insights offered by the margins, which can influence the likelihood of recurrence post-surgery.

One of the most impressive aspects of this research is the capacity of the deep learning models to process vast amounts of data at an unparalleled speed. Traditional diagnostic methods often involve painstaking manual analyses that can be time-consuming and prone to human error. By contrast, the application of these AI models enables rapid evaluation, thereby facilitating quicker decision-making avenues for clinicians. This efficiency could allow for timely interventions, ultimately enhancing patient care.

Furthermore, the study emphasizes the importance of multimodal data integration, combining not only histopathological images but also clinical and genomic data. By leveraging diverse data types, the researchers were able to craft a more nuanced predictive model that accounts for various facets of tumor biology. This integrative approach signifies a shift towards more holistic cancer care, where treatment can be tailored to the patient’s unique tumor profile rather than a one-size-fits-all methodology.

The predictive algorithms developed in this study were rigorously validated through a series of clinical trials, enhancing the credibility of the findings. The researchers meticulously evaluated the performance of their models against existing prognostic indicators. Remarkably, the AI-driven predictions showcased superior accuracy, demonstrating their potential to become an essential component of oncological diagnostics.

Moreover, the implications of this study extend beyond mere prognostication. The findings underscore a transformative opportunity for clinical workflows, where AI can augment the capabilities of pathologists rather than replace them. By acting as a second pair of eyes, intelligent systems can help reduce diagnostic errors, providing pathologists with data-driven insights to support their conclusions.

As we contemplate the future of cancer treatment, it’s becoming clear that incorporating technology is not just an added benefit; it is rapidly becoming a necessity. The findings of this research present a compelling case for health institutions to invest in AI technologies, not only to enhance diagnostic accuracy but also to optimize therapeutic strategies. However, to fully embrace this transformation, ongoing training and education for medical professionals will be crucial in leveraging these advanced tools effectively.

Also noteworthy is the ethical dimension of integrating AI into cancer diagnostics. Despite the allure of advanced technologies improving accuracy and efficiency, robust frameworks must be established to address potential biases inherent in AI systems. Ensuring that algorithms are trained on diverse populations will be pivotal in preventing disparities in care, thereby promoting equitable access to advanced cancer treatments for all patients.

The study by Michot and colleagues marks a critical step forward in the intersection of AI and oncology, showcasing the transformative potential of deep learning in soft-tissue sarcoma prognosis. As research in this area continues to burgeon, the prospect of deploying AI-driven tools in routine clinical practice appears ever more promising. The journey has only just begun; however, the horizon looks brighter for patients as technology and medicine converge in unprecedented ways.

This transformative research encourages a reassessment of how we view prognostic tools in oncology. Better predictions will not only help medical teams make informed decisions but will also empower patients through shared understanding of their treatment trajectories. By prioritizing patient education alongside technological advancements, we can foster a more collaborative healthcare landscape.

In summation, the integration of AI and digital pathology holds immense promise for the field of oncology, particularly concerning soft-tissue sarcomas. The study provides a glimpse into a future where predictive analytics guide treatment decisions, holding out hope for improved patient outcomes. As more research emerges and technologies advance, the healthcare community stands on the brink of a revolution that could redefine how we approach cancer treatment and management.

The robust application of these findings may take time, but the profound implications for soft-tissue sarcoma management and treatment are undeniable. With further refinement and validation, predictions derived from deep learning models can soon transition from theoretical discussions to clinical tools, fundamentally reshaping practices in oncology.

As we navigate this evolving landscape, the collaboration between technologists, clinicians, and researchers will be vital in harnessing AI’s full potential. The prospect of utilizing advanced predictive models could indeed herald a new era in precision medicine, aiming for not only longer lifespans but also improved quality of life for patients grappling with cancer.

Ultimately, as the research community continues to explore the potential of AI in healthcare, the exciting intersection of technology and medicine will undoubtedly offer new avenues for enhancing human health globally. The future of soft-tissue sarcoma management is not just about survival—it is about thriving in the face of adversity, propelled forward by innovation and a relentless pursuit of excellence in patient care.

Subject of Research: Prognostic prediction in soft-tissue sarcomas using deep learning and digital pathology.

Article Title: Prognostic prediction in soft-tissue sarcomas using deep learning and digital pathology of tumor and margin areas.

Article References:
Michot, A., Le, VL., Coindre, JM. et al. Prognostic prediction in soft-tissue sarcomas using deep learning and digital pathology of tumor and margin areas. Sci Rep 15, 38534 (2025). https://doi.org/10.1038/s41598-025-20804-1.

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41598-025-20804-1

Keywords: AI in oncology, soft-tissue sarcomas, deep learning, digital pathology, prognostic prediction, precision medicine.

Tumor segmentation fundamentally shapes the treatment trajectory in head and neck cancers, where anatomical complexities pose a significant challenge for clinicians. Traditionally, delineating tumor boundaries manually is labor-intensive and prone to interobserver variability. The advent of AI-powered image analysis heralds a paradigm shift, promising automation that could elevate both accuracy and reproducibility. PET imaging reveals the metabolic activity of tumors, while the CT component of PET/CT offers invaluable anatomical detail. The integration of metabolic and structural data provides a richer substrate for AI to operate, potentially enhancing segmentation performance.

The investigative team embarked on an exhaustive search across several major scientific databases — including Scopus, Embase, PubMed, Cochrane, Web of Science, and Google Scholar — identifying studies published up to December 2024, with a meticulous update in March 2025. Their eligibility criteria were stringent, focusing on studies that utilized AI algorithms specifically for head and neck tumor segmentation employing either PET alone or PET/CT, with quantitative performance metrics available for rigorous analysis. This methodological rigor ensures that the synthesized findings rest on robust evidence.

Upon aggregating data from eleven qualifying studies, the meta-analysis revealed a clear superiority of PET/CT over PET-only in the context of AI segmentation. Quantitatively, the Dice Similarity Coefficient (DSC), a statistical measure for gauging spatial overlap between predicted and true tumor contours, exhibited an improvement of 0.05 with PET/CT. Complementary metrics such as sensitivity and precision also showed notable enhancements, with increments of 0.04 and 0.05 respectively. The Hausdorff Distance (HD95), which quantifies the maximum spatial discrepancy between segmentation boundaries, decreased by around 3 millimeters, indicating tighter tumor border approximations.

Statistical evaluation of heterogeneity—a measure of variability between study results—revealed a generally low inconsistency, bolstering the reliability of pooled estimates. Exceptions emerged with HD95, which showed substantial heterogeneity (I² = 75%), and sensitivity, exhibiting moderate variability (I² ≈ 61%). Nevertheless, sensitivity analyses, including the exclusion of particular outlying studies and SD-imputed data, reaffirmed the steadfastness of the reported superiority of PET/CT-based AI models.

A pivotal aspect of the study was the dual focus on overall versus primary tumor segmentation tasks, reflecting the clinical necessity to discern whether AI performance differentially impacts general tumor burden delineation compared to targeting the primary lesion specifically. Subgroup analyses demonstrated a uniform advantage for PET/CT across all key performance metrics, suggesting that the integration of anatomical information in PET/CT robustly augments the AI’s capability regardless of segmentation scope.

Methodological quality appraisal, employing the CLAIM (Checklist for Artificial Intelligence in Medical Imaging) framework and QUADAS-C risk of bias tool, revealed high-quality, low-bias studies included in the review. This rigorous evaluation provides confidence that the pooled results are not artifacts of suboptimal study designs. The consistent excellence across studies also signals a maturation in AI research within oncological imaging, paving the way for clinical translation.

The clinical implications of these findings are profound. AI-assisted PET/CT segmentation could expedite and refine radiotherapy contouring, potentially improving treatment precision, reducing radiation exposure to healthy tissues, and enhancing patient outcomes. The automation introduced by AI promises to alleviate the workload on clinicians and standardize tumor delineation across institutions, a critical step toward equitable cancer care.

Furthermore, the study advocates for the creation and adoption of unified datasets. Given the diversity and complexity of medical imaging data, centralized or federated learning frameworks leveraging distributed systems may be essential for scaling AI applications. Such collaborative data environments could enhance the robustness, generalizability, and applicability of AI models across heterogeneous clinical settings.

This research critically extends the evidence base supporting AI’s integration with PET/CT imaging modalities in head and neck oncology, suggesting a recalibration of imaging protocols toward hybrid methodologies. Beyond immediate segmentation improvements, this fusion sets the stage for advanced AI-driven radiomic and radiogenomic analyses, linking imaging phenotypes to molecular profiles and personalized therapy pathways.

While the study illuminates the clear advantage of PET/CT for AI-based segmentation, it also underscores the necessity for ongoing methodological innovation. Addressing heterogeneity in metrics like HD95 may require the refinement of AI architectures or ensemble strategies. Future research should also explore prospective trials incorporating automated segmentation into clinical workflows, assessing impact on decision-making and long-term outcomes.

The synergy of AI and PET/CT imaging embodies the forefront of personalized medicine. As algorithms evolve and computational power expands, the precision and automation of tumor segmentation will only intensify. This study is a clarion call for the oncology and medical imaging communities to embrace integrated AI-augmented imaging protocols for transformative patient care in head and neck cancer.

In sum, the compelling evidence presented confirms that AI-enhanced PET/CT imaging surpasses PET-only approaches in tumor segmentation tasks within the head and neck cancer domain. This not only validates existing clinical practices but also brightens the horizon for AI’s role in the seamless integration of imaging, diagnosis, and therapy planning.

This synthesis stands as a testament to the intersection of cutting-edge technology and clinical need, emphasizing that the future of oncology resides in sophisticated, AI-driven diagnostic ecosystems that empower clinicians with unprecedented accuracy, efficiency, and insight.

Subject of Research: Application of artificial intelligence in head and neck tumor segmentation comparing PET and PET/CT imaging modalities.

Article Title: Application of artificial intelligence in head and neck tumor segmentation: a comparative systematic review and meta-analysis between PET and PET/CT modalities.

Article References:
Hajimokhtari, H., Soleymanpourshamsi, T., Rostamian, L. et al. Application of artificial intelligence in head and neck tumor segmentation: a comparative systematic review and meta-analysis between PET and PET/CT modalities. BMC Cancer 25, 1656 (2025). https://doi.org/10.1186/s12885-025-14881-8

Image Credits: Scienmag.com

DOI: https://doi.org/10.1186/s12885-025-14881-8

The challenge of predicting how breast cancer will respond to chemotherapy remains a critical bottleneck in oncology. Traditional biopsy methods, though informative, offer limited insights and suffer from spatial sampling bias due to the heterogeneous nature of tumors. Radiomics, an emerging discipline that extracts high-dimensional quantitative features from medical images, offers an unprecedented window into tumor biology beyond what is visible to the naked eye. By integrating 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) imaging with deep learning algorithms, the new approach captures complex tumor phenotypes and metabolic patterns associated with treatment efficacy.

At the heart of this research lies 18F-FDG PET/CT, a hybrid imaging modality that combines metabolic and anatomical information. 18F-FDG, a radiolabeled glucose analog, is preferentially taken up by highly metabolic tumor cells, enabling visualization of active malignancies and their aggressive phenotypes. The CT component, on the other hand, provides structural information that complements metabolic data. By employing deep radiomic modeling on this multi-dimensional dataset, the researchers developed algorithms capable of discerning subtle variations in tumor texture, intensity, and shape that correlate with chemotherapy responsiveness.

The study incorporated a robust dataset of breast cancer patients undergoing neoadjuvant chemotherapy, harnessing 18F-FDG PET/CT imaging data acquired at multiple time points. Through rigorous feature extraction and preprocessing, the team converted these images into comprehensive radiomic profiles. These profiles served as inputs for deep learning models—specifically convolutional neural networks—that were trained to identify patterns predictive of pathological complete response (pCR), a key indicator of effective chemotherapy. The models underwent stringent validation procedures to ensure generalizability and reliability.

Remarkably, the deep radiomic models demonstrated superior performance when compared to conventional clinical and imaging predictors. Metrics such as accuracy, sensitivity, and specificity in predicting chemotherapy outcomes were significantly enhanced, underscoring the efficacy of combining metabolic imaging with deep radiomics. Notably, the model’s ability to predict pCR prior to treatment initiation opens avenues for early therapeutic stratification, potentially sparing non-responders from unnecessary toxicity and guiding them toward alternative regimens.

One of the intrinsic advantages of this methodology is its non-invasive nature, relying solely on routinely acquired imaging to generate predictive insights. This feature not only reduces patient burden but also facilitates seamless integration into existing clinical workflows. Furthermore, the repeatability of PET/CT scans offers opportunities for dynamic monitoring, allowing clinicians to adjust treatment plans in response to early indications of therapy resistance or sensitivity.

The implications of this research extend beyond breast cancer. The paradigm of combining 18F-FDG PET/CT with deep radiomics could be extrapolated to other solid tumors where metabolic imaging is routinely performed, such as lung, head and neck, and gastrointestinal cancers. By unveiling intricate tumor heterogeneity and metabolic diversity, these models may serve as universal tools for personalized therapy evaluation and prognostication.

Despite the promising results, several challenges remain before widespread clinical deployment can be realized. Data standardization, including harmonization of imaging protocols and feature extraction methods, is essential to replicate results across institutions. Moreover, the interpretability of deep learning models—often criticized as “black boxes”—must be enhanced to provide clinicians with actionable insights and foster trust in automated decision-support systems. The development of hybrid models that integrate radiomics with genomic and molecular data might further bolster predictive power and elucidate underlying biological mechanisms.

Ethical considerations are also paramount as AI-driven diagnostics gain traction. Patient privacy, data security, and unbiased algorithmic design need careful stewardship to prevent disparities and ensure equitable healthcare delivery. Collaborative efforts among oncologists, radiologists, computer scientists, and ethicists will be central to navigating these complex issues.

Looking ahead, prospective clinical trials designed to evaluate the impact of radiomic-based predictions on treatment outcomes are crucial. Such studies will not only validate the clinical utility of these models but also help define standardized endpoints and regulatory pathways. Coupling radiomics with emerging imaging biomarkers, such as hypoxia or immune cell infiltration markers, could further refine response assessment, enabling a multi-dimensional view of tumor behavior.

The integration of artificial intelligence into oncological imaging heralds a new chapter wherein tailored therapies are informed by intricate data signatures invisible to traditional diagnostics. The study by Jiang and colleagues exemplifies how marrying metabolic PET/CT imaging with deep learning can transform chemotherapy response prediction in breast cancer, potentially improving survival rates and quality of life for countless patients.

In conclusion, 18F-FDG PET/CT-based deep radiomic models embody a promising convergence of technology and medicine, paving the way for a future in which cancer treatment is not just reactive but anticipatory and precisely calibrated to each patient’s unique tumor biology. As research in this domain accelerates, the prospect of realizing truly personalized oncology care becomes increasingly attainable, heralding transformative impacts on global cancer management.

Subject of Research:
Prediction of chemotherapy response in breast cancer using 18F-FDG PET/CT-based deep radiomic models.

Article Title:
18F-FDG PET/CT-based deep radiomic models for enhancing chemotherapy response prediction in breast cancer.

Article References:

Jiang, Z., Low, J., Huang, C. et al. 18F-FDG PET/CT-based deep radiomic models for enhancing chemotherapy response prediction in breast cancer.
Med Oncol 42, 425 (2025). https://doi.org/10.1007/s12032-025-02982-0

Image Credits: AI Generated

Neoadjuvant chemo-immunotherapy has emerged as a frontline approach to increase the resectability and overall prognosis of patients with early-stage or locally advanced NSCLC. Despite its promise, the evaluation of treatment efficacy conventionally relies on postoperative pathological assessments, a process that can only be performed after tumor resection. This delay limits clinicians’ ability to dynamically refine therapeutic strategies during the neoadjuvant window and extends patient exposure to potentially ineffective regimens. The development of NeoPred targets this clinical impasse by enabling real-time, non-invasive predictions of MPR to guide surgical and systemic treatment decisions more proactively.

NeoPred’s unique strength lies in its utilization of dual-phase CT imaging — capturing tumor morphology and response dynamics both before the initiation of therapy and just prior to surgery. This dual temporal perspective allows the AI model to quantify subtle morphological changes induced by chemo-immunotherapy that often escape human visual assessment. Beyond imaging, the model incorporates a multimodal framework by integrating critical clinical parameters such as age, sex, body mass index, and tumor staging. Together, these data streams are fused within advanced 3D convolutional neural networks, forming a robust predictive architecture that tailors the evaluation to each patient’s complex clinical profile.

The study underpinning NeoPred is remarkable in scale and scope. Drawing on a diverse cohort of 509 NSCLC patients across four distinguished thoracic oncology centers, the research team adopted a rigorous methodology involving retrospective model training and validation on 459 cases, supplemented by prospective real-world testing on 50 additional patients. To further validate their findings externally, 59 independent cases from collaborating institutions served as a test set, reinforcing the model’s generalizability and clinical applicability across heterogeneous populations and imaging protocols.

Critically, NeoPred demonstrated formidable performance metrics throughout its evaluations. In the external validation cohort, the AI system achieved an area under the receiver operating characteristic curve (AUC) of 0.772 using imaging data alone; this precision improved with the inclusion of clinical variables, elevating the AUC to 0.787. These results underscore the added value of integrating multimodal information beyond traditional radiomics. Moreover, the model’s predictive power shone in prospective clinical application, where NeoPred surpassed expert thoracic surgeons’ interpretation accuracy, with an AUC of 0.760 compared to the human benchmark of 0.720.

One of the most compelling breakthroughs emerged from the study’s investigation into “stable disease” (SD) cases classified according to the RECIST criteria. While these patients exhibit limited apparent tumor shrinkage post-therapy, NeoPred successfully uncovered underlying major pathological responders within this group, achieving AUCs of 0.742 in external datasets and an even more striking 0.833 in prospective cases. This capability to detect “pseudo-stable” responders highlights the model’s sensitivity to nuanced morphological and dynamic tumor transformations that traditional radiological assessments might overlook, offering a critical edge in personalized treatment planning.

NeoPred’s clinical implications extend beyond mere prognostication. By delivering early and reliable predictions of pathological response one to two weeks before surgery, the model facilitates evidence-based adjustments in perioperative strategies, potentially sparing patients from unnecessary surgical morbidity or enabling timely alterations in systemic therapy. The integration of AI-generated quantitative metrics into multidisciplinary team discussions promises to streamline workflows, promote objective risk stratification, and enhance collaborative decision-making—a transformational step toward precision oncology in thoracic surgery.

This breakthrough AI initiative also forms an integral component of a broader ecosystem of computational tools masterminded by Prof. He’s team, designed to address the multifaceted challenges of lung cancer management. Technologies span early detection platforms combining cell-free DNA methylation assays and low-dose CT scans, advanced proteomics for plasma biomarker identification, natural language processing algorithms for electronic health record mining, and sophisticated deep learning models predicting gene mutations directly from histopathology slides. Each innovation complements NeoPred’s intent by fortifying the continuum from screening to therapeutic response monitoring.

Technological sophistication characterizes the NeoPred model’s architecture. It employs dual 3D convolutional neural networks independently trained on pre-treatment and pre-surgery CT imagery, facilitating dedicated feature extraction at each time point. A fusion mechanism consolidates outputs alongside clinical indicators, enhancing predictive cohesion. Notably, this workflow reflects a trend toward multimodal AI solutions that transcend the limits of isolated data sources, fostering holistic patient modeling and dynamic disease surveillance.

Importantly, the study not only quantified AI’s diagnostic superiority but also examined human-AI collaboration paradigms. Nine expert thoracic surgeons were asked to interpret CT scans initially without and subsequently with access to NeoPred’s predictive heatmaps. Post-exposure to AI insights, surgeons’ diagnostic accuracy surged to 82%, with an impressive AUC elevation to 0.829. This synergy exemplifies how AI tools can augment clinical expertise rather than supplant it, cultivating a collaborative interface that leverages machine precision and human judgment to optimize outcomes.

Beyond the realm of neoadjuvant therapy evaluation, NeoPred’s methodology portends future adaptability to other oncologic contexts where preoperative response assessment remains elusive. Its “blind spot complementarity,” as shown by identifying responders hidden within stable radiological presentations, positions the model as a prototype for AI-mediated phenotyping capable of unraveling tumor heterogeneity and microenvironmental complexities embedded within imaging phenotypes.

The integration of federated learning platforms within the ecosystem, such as the CAIMEN system fostered across over 40 institutions, underscores the team’s commitment to data security and collaborative model refinement without compromising patient privacy. This decentralized approach to AI training not only expands the diversity of representative data but also propels diagnostic accuracy across geographic and institutional boundaries, promising equitable precision medicine deployment.

As lung cancer remains the leading cause of cancer mortality worldwide, innovations like NeoPred usher in a new era where AI-driven insights guide therapeutic timelines, refine surgical candidacy, and personalize treatment trajectories with unprecedented granularity and timeliness. The convergence of multimodal imaging, deep learning, and clinical data modeling exemplified by NeoPred reflects the burgeoning potential of AI to revolutionize oncologic care, bringing hope for improved survival and quality of life to patients confronting NSCLC.

In sum, NeoPred exemplifies a paradigm shift in thoracic oncology, converging computational power and clinical acumen to anticipate tumor response before surgery, optimize treatment efficacy, and ultimately enhance patient-centered care. The pioneering work by Professor Jianxing He and colleagues marks a significant stride in operationalizing AI as a tangible, high-impact tool within modern cancer management, heralding broader adoption and continual innovation in this dynamic field.

Subject of Research: People

Article Title: NeoPred: dual-phase CT AI forecasts pathologic response to neoadjuvant chemo-immunotherapy in NSCLC

News Publication Date: 31-May-2025

Web References: http://dx.doi.org/10.1136/jitc-2025-011773

References:
[1] Zheng J, Yan Z, Wang R, et al. NeoPred: dual-phase CT AI forecasts pathologic response to neoadjuvant chemo-immunotherapy in NSCLC. J Immunother Cancer. 2025;13(5):e011773. doi:10.1136/jitc-2025-011773.

Keywords: Lung cancer, Immunotherapy, Adjuvants