Advancements in AI: Transforming Cancer Research and Treatment 

Introduction: 

This comprehensive collection aims to spotlight the latest breakthroughs in artificial intelligence (AI) within the field of cancer research, covering basic, translational, and clinical aspects. The primary focus is on precision oncology, exploring how AI tools can seamlessly transition from laboratories to clinics. The overarching goal is to elevate patient care and enhance clinical outcomes. Articles prioritized for inclusion will showcase innovative methodologies, address real-world challenges, and provide robust evidence using multicentric datasets. While deep learning remains central, the collection remains open to related fields. 

Image Processing:

Oncology heavily relies on image data, making the acquisition, interpretation, and analysis of such data critical at all stages of cancer detection and treatment. The shift to digitalization in medical imaging has paved the way for powerful AI methods, enabling high-throughput analysis and unlocking new possibilities in oncology. The focus on deep learning, especially convolutional neural networks, has revolutionized image-based cancer detection across various modalities. AI not only automates tasks like cancer detection but also generates image-based biomarkers, offering valuable insights into survival and treatment response. 

Language Processing: 

Textual data plays a crucial role in oncology, from electronic health records to patient-reported outcomes. The advent of transformer neural networks, particularly large language models (LLMs), has significantly enhanced natural language processing (NLP) capabilities. These models, trained on vast amounts of data, demonstrate human-level performance for various language understanding tasks. In this collection, we encourage articles employing LLM-based methods in oncology, showcasing their benefits in research and patient care. 

Enhancing Genomic Information with AI:

Next-generation sequencing has transformed cancer research, and AI applications offer opportunities to leverage high-throughput genomics data. AI can identify novel genomic and transcriptomic features, analyze pathogenicity of variants, and facilitate large-scale genotype–phenotype associations. This synergy between AI and genomics has the potential to unlock personalized care. We invite articles detailing meaningful discoveries in cancer genomics using AI and innovative methods to address challenges. 

The Future is Multimodal:

Transformer neural networks have facilitated the development of multimodal machine learning models, allowing simultaneous analysis of different data types. In oncology, decision-making inherently involves multiple modalities—images, text, genomic data, and more. Multimodal models offer a comprehensive approach, combining different data types to provide holistic insights. We welcome articles exploring novel technical approaches for multimodality in oncological applications. 

Evidence-Based Medicine and Patient-Centered Care:

While celebrating technological progress, we emphasize the importance of evidence-based medicine and patient-centered care. AI applications in oncology must generate solid scientific evidence to improve research and patient outcomes. Principles of empathetic human interactions, shared decision-making, and rigorous scientific consideration, along with ethical and privacy considerations, should guide AI development and implementation. 

Methods: 

In adherence to COPE guidelines, we disclose the use of GPT-4 (OpenAI) for spelling and grammar checks during the writing of this article.