Connect with us


How ChatGPT is Transforming Cancer Care




In recent years, the blend of artificial intelligence and healthcare has led to exciting advancements in cancer care. At the core of this change is generative AI, which can analyze vast amounts of patient data and generate insights that improve diagnosis and treatment. As generative AI continues to evolve, especially in its ability to work with various types of data, it’s opening new possibilities for better diagnoses, more effective treatments, and improved patient outcomes. This article explores how a generative AI system, ChatGPT, is transforming cancer care, bringing new hope and innovative solutions to the forefront.

Color Health's Vision: A ChatGPT for Cancer Care

Imagine having a version of ChatGPT that not only understands complex medical knowledge but also has detailed information about your patients. Picture this advanced ChatGPT aiding doctors in diagnosing cancer with remarkable precision, tailoring treatment plans based on a patient's genetic profile, and foreseeing potential complications before they occur. This futuristic vision is becoming a reality, through a collaboration between Color Health, a genetic testing startup, and OpenAI, the creators of ChatGPT.

This collaboration has led to the development of a groundbreaking “copilot” for doctors—a specialized version of ChatGPT specifically trained and optimized for oncology. This innovative tool harnesses the power of ChatGPT-4o to develop personalized screening and diagnostic plans for patients. By merging patient medical data with the latest clinical insights, this copilot enables healthcare professionals to make well-informed decisions about cancer screening and treatment.

Building ChatGPT for Caner Care

To build this groundbreaking tool, OpenAI employs a technique known as retrieval-augmented generation (RAG), which enables ChatGPT to extract information from external medical sources rather than relying on pre-existing knowledge. The RAG is empowered with comprehensive patient information and medical knowledge using a diverse array of data sources, including clinical notes, medical documents, patient histories, and the latest research studies. Using this RAG method, the ChatGPT meticulously extracts and normalizes valuable information, such as a patient's family history and individual risk factors, along with pertinent medical knowledge from these documents. The remarkable capability of ChatGPT-4o to comprehend multimodal information—ranging from clinical notes and medical drawings to PDF documents—enables it to gather insights from various data types. Once this knowledge is assimilated, ChatGPT is employed to answer critical questions like, “What screenings should the patient undergo?” in much the same way a standard ChatGPT responds to user prompts.

Additionally, the built-in ability of ChatGPT to generate and complete documents enables it to streamline the necessary paperwork for diagnostic workups. This includes creating medical necessity documents and obtaining insurance pre-authorizations. By integrating and automating these tasks, the ChatGPT not only enhances the efficiency of the diagnostic process but also frees up valuable time for healthcare providers, enabling them to focus more on patient care.

How Color Health Employs ChatGPT for Cancer Care

While there are numerous compelling applications of ChatGPT for cancer care, Color Health has identified two primary use cases for it: early cancer detection and effective patient management during treatment. In the first use case, Color Health faces the challenge of many individuals missing necessary screenings despite the availability of validated tools and guidelines. This gap often arises due to irregular doctor visits or insufficient adjustments in screenings. The ChatGPT serves as an expert oncologist's assistant, ensuring crucial screenings are not overlooked.

In the second use case, Color Health recognizes the urgency once someone is diagnosed with cancer. In this situation, time is critical, and everyday counts. Pre-treatment workups are essential but can be slow and frustrating for patients, leading to delays and incomplete information for doctors. The ChatGPT could step in by identifying necessary tests before the oncology appointment, streamlining the treatment process and reducing delays.

By building a specialized ChatGPT for doctors, Color Health aims to bridge these gaps in cancer care, ensuring that more patients receive the necessary screenings and timely treatments.

Ensuring Quality and Safety

While this ChatGPT offers significant opportunities for enhancing cancer care, ensuring quality and safety is paramount. To achieve this, OpenAI and Color Health have adopted two key approaches: the copilot and doctor-in-the-loop models. The copilot concept is inspired by programming copilots and emphasizes that the copilot is designed not to replace the doctor but to augment their capabilities and enable them to handle more complex tasks efficiently. Conversely, the doctor-in-the-loop approach ensures that the copilot's output is reviewed by clinicians before being delivered to patients. This collaborative model not only improves the copilot's accuracy and reliability but also maintains crucial human oversight in patient care. By combining the strengths of AI with human expertise, Color Health aims to enhance the overall quality and safety of cancer care.

Besides these approaches, it is crucial to thoroughly evaluate this technology in clinical settings before deploying it in the real world. To assess its impact, Color Health is collaborating with the University of California, San Francisco Helen Diller Family Comprehensive Cancer Center (UCSF HDFCCC). The initial implementation will involve a retrospective evaluation, followed by a targeted rollout. Depending on the evaluation results, there is potential to integrate ChatGPT into clinical workflows for all new cancer cases at UCSF. This rigorous evaluation process ensures that the system copilot meets the highest standards of effectiveness and safety before widespread implementation.

The Bottom Line

The integration of generative AI, exemplified by ChatGPT, into cancer care represents a transformative leap in healthcare. By harnessing advanced AI techniques, Color Health and OpenAI are developing tools that significantly enhance diagnostic accuracy and treatment efficiency. The copilot model, with its doctor-in-the-loop approach, ensures that AI augments human expertise rather than replacing it, maintaining critical oversight and improving patient outcomes. As this technology undergoes rigorous evaluation in clinical settings, its potential to transform cancer care becomes increasingly clear. With comprehensive patient data and cutting-edge clinical insights, ChatGPT is poised to bridge gaps in early detection and patient management, offering new hope and improved care for cancer patients worldwide.

Dr. Tehseen Zia is a Tenured Associate Professor at COMSATS University Islamabad, holding a PhD in AI from Vienna University of Technology, Austria. Specializing in Artificial Intelligence, Machine Learning, Data Science, and Computer Vision, he has made significant contributions with publications in reputable scientific journals. Dr. Tehseen has also led various industrial projects as the Principal Investigator and served as an AI Consultant.