AI vs Cancer

In partnership with

Pioneering medicine starts with investors like you.

Cizzle Biotech is transforming lung cancer detection with a patented, 95% accurate blood test. By joining our raise, you’ll gain ownership in a company backed by partnerships with Bio-Techne and Moffitt Cancer Center. You can build your portfolio with a stake in a medical breakthrough.

Join the Raise

^{Read the Offering information carefully before investing. It contains details of the issuer’s business, risks, charges, expenses, and other information, which should be considered before investing. Obtain a Form C and Offering Memorandum at invest.cizzlebio.com}

Can AI Really Diagnose Cancer?

The short answer is: Not yet. While AI has made significant strides, it’s not ready to independently diagnose complex conditions like cancer.

A study involving ChatGPT-3.5, a general-purpose AI, showed that while it can provide some information about cancer types and treatments, it can also mix up correct and incorrect answers. This highlights the risks of using general-purpose models in critical areas like cancer diagnosis.

On the other hand, AI tools that combine specialized models trained on medical data are emerging. Products like Microsoft’s Azure Health Bot and IBM’s watsonx Assistant are examples. These models integrate credible medical sources like the US National Library of Medicine and CDC, making them useful for tasks like triage. However, they still lack the depth needed to fully understand a patient’s medical history or lifestyle—critical factors in diagnosis. For now, a doctor’s examination is essential.

Realistic AI Use Cases in Healthcare 

Instead of diagnosing cancer, AI chatbots can be more useful for symptom collection and connecting users with healthcare providers. These tools assess symptoms and suggest whether someone should monitor their condition, seek urgent care, or schedule an appointment. For example, we created a HIPAA-compliant chatbot using Azure Health Bot for one client. It helps users assess their symptoms and directs them to telehealth platforms like Teladoc or MDLive.

A key challenge for medical chatbots is interpreting human input, especially subjective symptoms like fatigue. Pre-set answer options or follow-up questions (e.g., "When did the fatigue start?") can improve accuracy and help the bot gather more context to make better recommendations.

AI in Cancer Imaging: A Growing Field

AI has proven effective in analyzing medical images like X-rays and mammograms. Over 700 radiology AI tools have been FDA-approved, and that number grows each year. In one project, we helped develop a machine learning (ML) algorithm to assist with early lung cancer detection. The system analyzed sputum samples, with a 92% sensitivity in detecting cancer in high-risk patients with small lung nodules.

When developing AI for cancer imaging, a high-quality dataset is crucial. The data should be diverse (covering different ages, genders, and ethnicities) and properly annotated. For example, oncologists need to label tumor boundaries and tissue characteristics accurately, as poor annotations can lead to misidentifications.

AI in Treatment Planning: A Promising Future

AI tools for treatment planning, such as the PERCEPTION project, show great promise in oncology. PERCEPTION analyzes individual cancer cells to understand how tumors evolve and resist treatment. In clinical trials, it successfully predicted which cancer patients would respond to certain treatments, offering a glimpse into how personalized cancer care could evolve.

However, AI models are not yet ready for widespread use in clinical settings. Even high-accuracy models (80–90%) are not sufficient in healthcare, where small errors can have significant consequences. Nonetheless, AI’s potential to revolutionize treatment planning remains promising.

The Future of AI in Cancer Care

AI is quickly becoming a valuable tool in cancer care, from diagnostics to treatment personalization. As both patients and clinicians gain more trust in AI, we can expect its role to grow in precision medicine, making cancer care more tailored and effective in the future.