Revolutionizing Cancer Care: AI and Cloud Tech at City of Hope

Transforming Cancer Care with AI and Cloud Technology​

In an era where technology and healthcare increasingly intertwine, City of Hope, one of the United States' largest and leading cancer research and treatment organizations, is setting an example in innovation. The hospital’s recent initiative, leveraging Microsoft Azure OpenAI Service and advanced optical character recognition (OCR) technology, points to a future where administrative burdens are significantly reduced and doctors are empowered to focus more on patient care.

A Century-Old Legacy Meets Modern Innovation​

Founded in 1913, City of Hope has decades of history in fighting cancer and improving patient outcomes. What sets this organization apart is not only its commitment to compassionate care but also a keen interest in integrating cutting-edge technologies in its operations. Recognizing that timely and efficient processing of massive amounts of medical data is a critical challenge, the organization embarked on a project that harnesses artificial intelligence (AI) and cloud computing to streamline patient onboarding and information management.

• City of Hope’s longstanding commitment to research and treatment paved the way for embracing digital transformation.
• Innovations spanning over 100 years have created an environment of continuous improvement in healthcare.
• The integration of modern AI technologies signifies a strategic step forward in addressing pervasive administrative challenges.

The Administrative Challenge in Modern Healthcare​

For any major healthcare provider, the influx of patient records is both staggering and complex. Doctors at City of Hope routinely encounter new patients carrying extensive medical histories—often containing 10 to 20 years of data that includes test results, progress notes, and treatment reports. Reviewing this trove of information is onerous, frequently forcing physicians to spend long hours combing through hundreds of document pages. This time-consuming process detracts from their core responsibility: engaging directly with patients.
The key pain points include:

• Volume and Complexity of Data: The sheer number of documents can overwhelm even the most seasoned medical professionals.
• Time-Intensive Processes: Manual review of lengthy histories draws attention away from patient care.
• Impact on Patient Interaction: Excessive time spent on administrative tasks inevitably reduces the doctor-patient interaction, affecting the quality of care.

City of Hope recognized that tackling this administrative burden required a robust solution—one that not only optimized data processing but also returned valuable time to doctors and their patients.

Leveraging Microsoft Azure OpenAI Service​

The collaboration between City of Hope and Microsoft illustrates a mutually beneficial relationship, where both technology and healthcare expertise merge to attain groundbreaking results. Central to the solution is Microsoft Azure OpenAI Service. This service provides the backbone for deploying large language models (LLMs) capable of digesting massive volumes of clinical documentation and summarizing key insights for the medical team.
Key features of this technological integration include:

• Automated Documentation Review: With the aid of LLMs, the AI system reads through comprehensive patient histories and extracts crucial data.
• Optical Character Recognition (OCR): This technique enhances the system’s ability to process both printed and handwritten text, bridging the gap between traditional paper records and digital text.
• Real-Time Insights: By processing data rapidly, the system provides near-instant summaries that allow doctors to quickly grasp a patient’s medical history.
• Reduced Manual Labor: The AI-driven approach minimizes the necessity for exhaustive manual review, enabling physicians to devote more time to diagnostics and patient discussions.

As Simon Nazarian, Chief Digital and Technology Officer at City of Hope, aptly noted, the collaboration with Microsoft is based on a shared vision for advancing healthcare using the best available technologies. He emphasized that this “mutually agreeable collaboration” has significantly enhanced the startup of innovative solutions in the hospital setting.

Transforming Patient Onboarding​

One of the most impressive outcomes of this digital transformation is the rapid onboarding of patients. In 2024 alone, City of Hope onboarded over 150,000 new patients. This milestone is not just a number—it is indicative of the efficiencies that AI and cloud technology can introduce into healthcare systems burdened by administrative overhead.

How Does AI Accelerate Onboarding?​

• Enhanced Data Processing: The system swiftly combs through decades-old records and synthesizes them into concise summaries.
• High Throughput with Accuracy: By using Azure OpenAI and OCR, the process scales to handle thousands of patients simultaneously while maintaining high data accuracy.
• Streamlined Referral Systems: Automated processing leads to quicker referrals, ensuring that patients receive timely evaluations and treatment.
• Reduced Non-Clinical Overtime: Medical experts are less encumbered by extended reviewing hours, fostering a better work-life balance and reducing burnout.

By integrating these advanced technologies, City of Hope not only improves operational efficiency but also creates a model for other healthcare institutions facing similar administrative challenges.

AI’s Impact on Clinical Decision-Making​

Beyond patient onboarding, AI’s transformative role extends to enhancing clinical decision-making. The early hints of AI-driven support systems in healthcare open new avenues to accelerate diagnoses and treatment plans. Consider the following benefits:

• Timely Summaries: With significant parts of patient histories distilled into quick-read summaries, clinicians are equipped with crucial insights at the point-of-care.
• Identification of Critical Data: AI algorithms can quickly flag abnormalities or treatment milestones that might otherwise be overlooked in lengthy medical records.
• Support for Multidisciplinary Teams: With summarized information, various specialists—from oncologists to radiologists—can communicate more efficiently, fostering a collaborative process in patient care.

AI enables medical personnel to transcend traditional limitations. The sentiment shared by Nasim Eftekhari, Executive Director of Applied AI and Data Science at City of Hope, encapsulates this shift; the aspiration of creating a system that could "digest all this information and provide an accurate summary" resonates deeply with the current trends in clinical automation and digital transformation.

The Role of Cloud Technology in Healthcare​

City of Hope’s adoption of Microsoft Azure underscores the growing reliance on cloud technology by healthcare organizations. Using Azure for AI implementations presents several distinct advantages:

• Scalability: Cloud services allow healthcare providers to scale resources dynamically in response to fluctuating demands without substantial overhead.
• Security and Compliance: Advanced security measures and compliance certifications help ensure that sensitive patient data is protected, which is paramount in healthcare.
• Flexible Integration: The Azure platform easily integrates with other hospital systems, fostering seamless communication across multiple platforms and devices.
• Innovation Foundation: With access to leading-edge machine learning models and computing power, hospitals can quickly deploy advanced applications without the need for extensive hardware investments.

For IT administrators on Windows platforms, these benefits are well understood. The same infrastructure that supports enterprise-level operations in business can now drive change in the patient care sphere. This evolution signals broader implications for healthcare providers looking to modernize legacy systems and harness the potential of AI.

Challenges and Future Perspectives​

No transformative journey is without its challenges. While the benefits of AI are substantial, integrating sophisticated systems into healthcare settings brings hurdles that must be navigated carefully.

Ensuring Ethical and Responsible AI Use​

Implementing AI in healthcare settings raises ethical considerations that must be addressed:

• Data Privacy: Strict enforcement of HIPAA and similar regulations ensures that patient information remains confidential.
• Bias and Accuracy: Responsible deployment of LLMs requires continual monitoring for biases or inaccuracies, which might affect patient outcomes.
• Human Oversight: Although AI speeds up data processing, final decision-making remains with medical professionals, ensuring that empathetic, human judgment complements automated insights.

The Road Ahead in AI-Driven Healthcare​

The evolution of AI in healthcare is just beginning. Future perspectives include:

• Enhanced Predictive Analytics: AI systems will not only summarize historical data but predict trends, helping in early diagnosis and even prevention.
• Personalized Treatment Plans: With comprehensive data analysis, treatment regimens could be tailored more precisely to each patient’s unique medical history.
• Extended Applications: Beyond administrative tasks, AI might assist in areas such as imaging analysis, real-time patient monitoring, and even robotic assistance.
• Interoperability Improvements: As more institutions adopt similar technologies, the seamless exchange of anonymized data might lead to a more connected healthcare ecosystem.

Such forward-looking applications demonstrate that City of Hope’s current initiative is just the tip of the iceberg in a broader shift that will redefine how healthcare is delivered.

Real-World Impact and Broader Implications​

The implications of City of Hope’s innovative use of AI extend beyond its immediate operational improvements:

• Enhancing Patient Outcomes: By reducing administrative delays and streamlining data review, patients experience quicker diagnoses and more timely interventions.
• Empowering Medical Staff: Physicians can realign their priorities, spending less time on paperwork and more on patient care, which can significantly improve the overall quality of treatment.
• Demonstrating Scalability: The ability to onboard a high volume of patients effectively serves as a real-world blueprint for other medical institutions seeking similar improvements.
• Catalyzing Broader Digital Transformation: The success at City of Hope highlights the relevance of cloud services and AI-driven systems in sectors with traditionally manual workflows. This example may inspire other hospitals to invest in similar technologies, ultimately fostering a more dynamic and responsive healthcare system.

Lessons for IT Professionals and Healthcare Administrators​

For IT professionals and healthcare administrators minting from the intersection of technology and service delivery, City of Hope’s story offers practical insights:

• Embrace Collaboration: The fruitful partnership between City of Hope and Microsoft illustrates that inter-industry collaboration can drive monumental changes.
• Invest in Scalable Technologies: By choosing cloud-based solutions like Microsoft Azure, organizations can future-proof their IT infrastructure and handle growing data loads efficiently.
• Prioritize User-Centric Design: Whether designing software or implementing new systems, the focus remains on enhancing user (and in this case, patient) experience.
• Stay Ahead of Compliance and Security: With sensitive data at stake, it’s critical to integrate robust security measures that align with contemporary regulatory standards.

Integrating Advanced Windows Technologies​

For our Windows enthusiasts and IT experts on WindowsForum, the technological underpinnings of City of Hope’s transformation may resonate deeply with ongoing trends in Windows 11 updates and Microsoft security patches. Windows systems, widely recognized for their robust enterprise capabilities, provide a reliable platform for hosting and interacting with such advanced cloud-based AI solutions. As healthcare increasingly depends on secure, efficient, and scalable technology, the adherence to modern Windows operating environments ensures that critical applications run seamlessly.

• Windows 11 introduces several enhancements in security and performance that support intensive applications like AI and large-scale data processing.
• Integration between Windows platforms and Microsoft Azure serves as a cornerstone for innovative digital ecosystems.
• For organizations prioritizing data security and operational agility, upgrading to the latest Windows environments is a strategic move that aligns with broader digital transformation trends.

Concluding Thoughts​

City of Hope’s innovative journey utilizing Microsoft Azure and advanced AI technologies is a testament to what happens when legacy meets cutting-edge tech. By automating the labor-intensive process of patient history review, the institution not only accelerates patient onboarding but also drastically improves the quality of patient care. Introducing responsible large language models and OCR to sift through years of complex medical data represents a paradigm shift in how healthcare systems manage information and deliver timely treatments.
The success story from City of Hope should serve as an inspiration across industries:

• It underscores the significant potential of cloud-based AI in reducing non-clinical burdens and empowering professionals.
• It challenges traditional workflows, encouraging organizations to adopt innovations that were once considered futuristic.
• It reminds us that technology’s true value lies in its ability to free up human resources—whether in healthcare or other sectors—allowing experts to focus on what truly matters: providing exceptional service.

As the healthcare landscape continues to evolve, it is clear that investments in technology—especially those that align with robust platforms such as Microsoft Azure and secure, up-to-date Windows systems—will drive the next wave of innovation. From enhanced patient care to streamlined administrative processes, City of Hope’s example serves as a blueprint for digital transformation, encouraging hospitals and IT professionals alike to embrace change with optimism and a keen eye on responsible, ethical innovation.
In summary, the convergence of AI and cloud technology at City of Hope represents more than just an operational improvement—it marks the dawn of a new era in how healthcare institutions can leverage digital tools to truly transform patient care, setting a powerful precedent for the healthcare community worldwide.

---

Source: Microsoft City of Hope uses Microsoft Azure and AI to rapidly onboard thousands of patients per year | Microsoft Customer Stories

 

[ChatGPT]

Cancer care, often characterized by complex, multidisciplinary treatments, has witnessed a technology-driven transformation in recent years, and nowhere is this more apparent than in the high-stakes environment of tumor boards. Tumor boards bring together specialists such as radiologists, pathologists, oncologists, and surgeons to collaboratively determine a patient’s optimal treatment plan. While the therapeutic advances in oncology are significant, the intricate preparation required for these meetings can contribute to a logistical bottleneck that affects both providers and patients. In this landscape, artificial intelligence is not simply another buzzword—it is rapidly proving to be a vital co-pilot, thanks to inventive solutions like Microsoft’s healthcare agent orchestrator.

The Shift: Tumor Board Preparation and the Administrative Burden​

In many cancer centers, preparing just one patient’s case for tumor board review can take healthcare professionals several painstaking hours. As Timothy Keyes, a data scientist at Stanford Health Care as well as a combined MD and PhD candidate at Stanford University School of Medicine, emphasizes, “They take a ton of time to prep.” The challenge is not only in the breadth of data—spanning electronic health records (EHRs), imaging scans, lab results, and current medical literature—but also in its fragmentation and the manual effort required to extract, validate, and synthesize pertinent information. Cases must be summarized efficiently for the tumor board, with each piece of information correctly attributed and easily traceable.
In an era when clinicians’ time is one of the most valuable commodities, these hurdles are not trivial. According to a 2024 research article in JAMA Oncology, clinicians spend up to half of their day on documentation and administrative tasks, often at the expense of direct patient contact. This inefficiency undermines both provider satisfaction and patient care, reinforcing the demand for novel technological solutions that streamline, rather than complicate, clinical workflows.

Microsoft’s Healthcare Agent Orchestrator: A New Paradigm​

Enter Microsoft’s healthcare agent orchestrator, now accessible to the broader health tech community via the Agent Catalog in Azure AI Foundry. Built to enable the rapid deployment and orchestration of autonomous AI agents, the platform acts as a digital liaison between disparate data sources and the human clinicians who must interpret them. The orchestrator handles the heavy lifting—pulling, organizing, and synthesizing information with impressive speed and accuracy.
At Stanford Health Care, Keyes and his team leveraged Microsoft’s new tools to create AI agents dedicated to the distinct components of tumor board preparation. The process began with astonishing efficiency: within just ten minutes, agents from the Azure AI Foundry Agent Catalog were deployed into Microsoft Teams for pilot testing. This seamless onboarding is a significant feature—by integrating with familiar apps like Teams or Word via Microsoft 365 Copilot, clinicians can simply express what they need in natural language, bypassing the friction typical of new technology adoption.

Inside the Agent Ecosystem: What Each Agent Does​

The orchestration of specialized agents enables a task-focused approach that traditional chat models have struggled to achieve, particularly in the complex, high-variance environment of healthcare data.

1. Data Organizer Agent​

This agent’s primary function is to transform a deluge of unstructured clinical notes, lab reports, medication records, and genomics data—each in varying formats—into a coherent, succinct abstract. Importantly, the system includes rigorous citation mechanisms to ensure that every piece of summarized information is traceable, enabling clinicians to verify data rapidly and reliably. This is critical in medicine, where data provenance and transparency can be a matter of life and death.

2. Radiology and Pathology Agents​

For image-heavy specialties, AI must navigate the idiosyncrasies of radiological scans (like CT or MRI) and pathology slides, which differ fundamentally in data type and clinical significance. By deploying leading specialized AI models on Azure, the radiology agent interprets imaging studies, while the pathology agent digests complex whole-slide images to provide actionable findings. Early pilot results at Stanford showed that these agents could surface relevant results with high accuracy, with Keyes personally validating citations against source notes and finding them correct.

3. Clinical Trials Agent​

Staying abreast of current and upcoming clinical trials is a persistent challenge for oncologists; eligibility criteria are complex and the volume of ongoing research is immense. The dedicated clinical trials agent in Microsoft’s orchestrator sifts through current registries, identifying which trials a patient might qualify for, linking results directly for further exploration. This feature is poised to not only speed up board preparation, but could also offer patients cutting-edge therapeutic opportunities that might otherwise be missed.

4. Medical Research Agent​

Staying current with the scientific literature—another responsibility that falls to busy clinicians—is supported by a reasoning-based agent that combs through scientific papers, extracting and presenting relevant findings complete with links for deeper investigation. It offloads the manual trawling of databases like PubMed, a process that historically adds further hours to case prep.

5. Report Creation Agent​

At the conclusion of the process, a report creation agent aggregates the work of its peer agents, delivering a professionally formatted case summary ready for presentation—typically as a Word document or PowerPoint slide deck. For many clinicians, this alone can spell the difference between an all-nighter and a manageable workload.

Time Savings and Efficiency: Quantifying the Impact​

According to Stanford’s early trials, “preparing a single patient’s case for a tumor board could take several hours; in testing, AI agents might make the work ten times faster.” Given that Stanford Health Care serves about 4,000 patients annually across more than a dozen tumor boards, this efficiency gain is projected to aggregate rapidly. Even a conservative estimate would suggest thousands of clinician-hours saved each year, translating into more time for direct patient care, research, and professional development—a trifecta that benefits every stakeholder in the healthcare ecosystem.
These time savings also accrue during multidisciplinary meetings themselves, where a dozen or more clinicians may convene for each case. As Keyes points out, “The agents will enable the work to be done easier, faster and more efficiently, which really matters when you’re talking about meetings with 10 clinicians in them, where time is really precious.”

Seamless Integration: No Steep Learning Curves​

One of the orchestrator’s most lauded strengths is its integration with Microsoft 365 Copilot. For clinicians already accustomed to tools within Microsoft’s productivity suite, the agents are ready to use—no additional applications, no cryptic interfaces to master. Simply put: if you can type a query in Teams or Word, you can leverage this technology. This “frictionless” user experience is a direct response to endemic burnout in healthcare, where each new software system typically brings more hurdles than help.
A 2023 survey by the American Medical Association found over 60% of clinicians felt existing EHR systems detracted from their efficiency due to poor usability and excessive administrative burden. Microsoft’s approach—embedding advanced AI functionality into established platforms—represents a potentially seismic shift in how digital health innovations are adopted at scale.

Maintaining Clinical Oversight: AI as an Assistant, Not an Autopilot​

The prospect of AI “autonomously” making healthcare decisions is as terrifying as it is tantalizing. Keyes is blunt about where he and his colleagues stand: “That is not really what we’re envisioning. We do want the clinicians in charge of a patient’s care. We always want them to be able to check.” The orchestrator is designed not to replace, but rather to augment, freeing doctors from the Sisyphean task of collation and documentation while preserving human oversight at every critical juncture.
Indeed, AI-generated content is always accompanied by citations and references, making verification straightforward and maintaining accountability—a profound reassurance in fields subject to regulatory scrutiny and ethical imperatives. In this paradigm, AI acts as the ultimate medical scribe: swift, tireless, and always transparent.

Strengths: The Human–Machine Symbiosis in Action​

- Massive Efficiency Gains​

Cloud-based orchestration means that complex multi-agent workflows, once measured in hours, can now be completed in minutes. For overstretched clinical teams, this isn’t just a “nice-to-have”—it’s a means of improving the consistency, reliability, and timeliness of patient care.

- Data Integrity and Transparency​

By structuring output with precise source citations, the orchestrator undercuts one of the most persistent challenges with conventional AI chatbots: unverifiable answers. Hot links to original data empower clinicians to trust—yet still verify—automated outputs.

- Lowering the Adoption Barrier​

With no need for clinicians to learn new apps or workflows, the orchestrator is positioned to win quick buy-in from even the most technology-weary users, suggesting a viable path to large-scale adoption in health systems.

- Expandability​

As the Agent Catalog expands and as more domain-specific models are validated, the same orchestration infrastructure could be extended to settings beyond oncology: cardiac boards, transplant committees, or even general inpatient team handoffs.

Potential Risks and Cautions​

Despite the clear promise, several risks and open questions remain.

- Clinical Validation and Generalizability​

As of this writing, Stanford Health Care is still testing its application of the healthcare agent orchestrator in a research setting. No AI-generated data has yet been used in real-time patient care. The key challenge remains external validation: Will these tools perform as reliably and accurately outside the controlled pilot environment? Historically, AI systems trained in one setting can falter in another—a well-known challenge in “AI chasm” research.

- Data Security and Privacy​

With agents aggregating data from EHRs, imaging, and genomics, robust safeguards must be in place to ensure HIPAA (Health Insurance Portability and Accountability Act) compliance and prevent unauthorized access. Microsoft states that Azure’s architecture meets stringent healthcare compliance standards, but the complexity of multi-agent, cross-system workflows could introduce new attack surfaces.

- Overreliance and Automation Bias​

There is a real risk that, with AI generating richly detailed, well-cited reports, some clinicians may be tempted to downshift their own critical scrutiny—especially as trust in the system grows. Safeguards and ongoing education will be essential to keep doctors engaged as active, skeptical participants.

- Algorithmic Bias​

Any AI operating on medical data can amplify pre-existing biases in that data. For example, underrepresented populations in clinical trials may be less likely to have trial matches suggested, or diagnostic interpretations could miss rare disease variants. Continuous auditing and systematic bias detection should underpin all deployments.

The Future of Agentic AI in Medicine​

As Keyes notes, “I would be excited at the idea of AI helping my doctors to be the best version of themselves and to liberate them from some of the time-consuming components of documentation so they can spend more time with me the patient.” This vision, increasingly echoed by clinicians worldwide, rests on a foundation of partnership—AI as an ally, not a replacement.
The next era in healthcare technology will be defined by the capacity to delegate not judgment, but labor. As platforms like Microsoft’s healthcare agent orchestrator mature, their true value may be measured less by raw speed and more by how much time, energy, and compassion they restore to those who matter most: patients and their caregivers.

Conclusion: A Promising Blueprint for AI in Healthcare​

Microsoft’s healthcare agent orchestrator, as piloted at Stanford Health Care, demonstrates a compelling synthesis of cutting-edge AI and the pragmatic needs of modern medicine. By focusing on collaborative agent frameworks, seamless integration, and rigorous clinical oversight, the orchestrator shows tremendous potential to transform not only tumor boards but a wide array of medical workflows. Its strengths—efficiency, transparency, user-friendliness, and expandability—position it at the forefront of digital health innovation.
Nevertheless, risks around validation, privacy, bias, and system reliability require persistent vigilance. As the orchestrator progresses from research pilot to real-world implementation, the lessons learned at institutions like Stanford will be critical to broader adoption and safe, effective use. In sum, AI agents in healthcare are moving from promise to practice. Their evolution will shape not only how medicine is delivered, but ultimately, who it serves—and how well.

---

Source: Microsoft 4 developers leading the way with AI agents