TEDx Talks - Are we ready for an AI oncologist? | Pallavi Tiwari | TEDxOshkosh
The speaker, a biomedical engineer, discusses the transformative potential of AI in cancer treatment. Currently, cancer diagnosis often leads to uncertain futures and standard treatments like surgery, radiation, and chemotherapy, which may not be effective for all patients. The speaker's team at the University of Wisconsin-Madison is developing AI models that can assist doctors in making personalized treatment decisions. These models analyze MRI scans to distinguish between benign and malignant tumors with over 90% accuracy, surpassing traditional invasive methods. This technology aims to reduce unnecessary procedures and financial burdens on patients by predicting individual responses to treatments. The speaker emphasizes the need for more data to improve AI accuracy and encourages contributions from those affected by cancer to advance this research.
Key Points:
- AI can personalize cancer treatment, improving accuracy and reducing unnecessary procedures.
- Current cancer treatments are often one-size-fits-all and financially burdensome.
- AI models have shown over 90% accuracy in diagnosing brain tumors, surpassing traditional methods.
- The speaker calls for more data to enhance AI models' accuracy to 99%.
- Public contributions of medical data can significantly advance AI research in cancer care.
Details:
1. 🎗️ Rethinking Cancer Diagnosis: A Manageable Future
- The segment envisions a future where cancer diagnosis is not seen as a terrifying event but as a manageable condition, similar to diabetes or heart disease.
- There is an implication that advancements in treatment could make cancer more treatable and less fear-inducing.
2. 🔍 Current Cancer Treatments and Their Challenges
- Cancer diagnosis currently implies an uncertain future for most patients, highlighting the unpredictability in treatment outcomes.
- The majority of cancer patients receive a standardized treatment approach, commonly known as 'one size fits all,' which includes surgery, radiation, and chemotherapy.
- Surgery is often the first line of treatment, aimed at removing tumors but can be invasive and requires significant recovery time.
- Radiation therapy targets cancer cells but can also damage surrounding healthy tissue, leading to side effects.
- Chemotherapy uses drugs to kill cancer cells but often results in severe side effects due to its impact on both cancerous and healthy cells.
- Each component of the standard treatment regimen is essential, yet they are costly and may not guarantee effectiveness for all patients.
- There is a need for personalized treatment plans to improve patient outcomes and minimize unnecessary side effects.
- Healthcare providers face challenges in balancing treatment efficacy, cost, and quality of life for patients.
3. 🤖 AI's Role and Trust in Personalized Cancer Care
3.1. AI's Potential in Cancer Treatment
3.2. Trust Issues in AI-Driven Cancer Care
4. 📚 Case Study: Mr. Kumar's Journey
- Mr. Kumar, a dedicated high school teacher and soccer coach, lived a fulfilling life characterized by his passion for education and sports. However, his life took a dramatic turn when he was diagnosed with glioblastoma, a highly aggressive brain tumor.
- The prognosis for glioblastoma is severe, with an average life expectancy of 15 to 20 months post-diagnosis. This grim outlook presented Mr. Kumar with significant personal and professional challenges.
- Despite undergoing rigorous standard treatment, including surgery, radiation, and chemotherapy, Mr. Kumar faced uncertainty. An MRI scan revealed a lesion in his brain, which could either be a cancerous recurrence or a benign effect of the treatment, with a 40% chance of being non-cancerous.
- The uncertainty of his condition affected not only his health but also his role as a teacher and coach, impacting his interactions with students and the community he cherished.