By
Matthew Hearfield
October 22, 2024
Updated
November 20, 2024
In this episode of AI to Z, Anna Frazzetto is joined by Swathi Young, Chief Technology Officer at Allwyn Corporation, to explore how AI is driving transformative change in healthcare. Swathi draws on her extensive experience with machine learning in cancer research to illustrate AI’s potential to improve patient outcomes.
Throughout the conversation, Swathi underscores the need to approach AI with a focus on real-world impact and accountability, making the episode a must-listen for leaders interested in harnessing AI for meaningful change in healthcare and beyond.
Listen to the full episode below:
AI's impact on cancer care
What they were traditionally doing was taking the CT scans and measuring that visceral fat manually by a radiologist. We were able to automate that using a vision machine learning algorithm.
Swathi Young explains how AI is driving breakthroughs in healthcare, particularly through machine learning in cancer care. She discusses projects where AI algorithms were used to analyse CT scans for lung cancer patients, helping create treatment plans by identifying factors like visceral fat in the L3 vertebrae that affect recovery.
Real-world AI applications
We used historical data, found patterns using machine learning, and found like top ten causes of other diseases and correlation.
This episode highlights examples of AI improving patient outcomes by automating complex tasks such as predicting readmissions for lung cancer patients and refining diagnosis in conditions like diabetic retinopathy. These real-world applications illustrate AI's capacity to enhance medical diagnosis and treatment with greater precision.
Another impactful application discussed was the use of AI in detecting diabetic retinopathy, an eye disease linked to diabetes. Machine learning models were able to detect early signs of this disease, potentially preventing severe vision loss. Additionally, AI was applied to improve breast cancer detection for women of colour, reducing the high false-positive rates that this demographic often faces.
Responsible AI and
ethical considerations
A critical theme is the need for "responsible AI" to mitigate bias in data and ensure fairness in AI-driven decisions. Swathi talks about the importance of incorporating diverse perspectives in AI development to prevent the amplification of human biases, especially in healthcare, where data may disproportionately affect certain demographic groups.Structured AI
implementation process
Swathi Young discusses her structured, five-step approach for successfully implementing AI solutions. The first step she emphasizes is aligning AI projects with business goals, ensuring that the AI initiative is focused on solving meaningful problems. She explains that businesses often rush into AI projects without a clear focus, which can lead to inefficiencies or misaligned outcomes.
Another critical step involves assessing the organization’s AI capabilities, including both personnel and infrastructure, and leveraging external expertise if needed. Collaboration with universities for pilot projects is also suggested as a way to jump-start AI initiatives on a limited budget.
The future of AI in healthcare
The episode touches on AI's future potential to revolutionize patient care by improving early diagnosis and creating more personalized treatments based on a patient’s genetic and medical history.
Key takeaways
-
AI's impact on cancer care: AI is revolutionising cancer care by automating tasks like CT scan analysis, helping optimize treatment plans based on patient-specific factors.
-
Real-world AI applications: Machine learning is improving healthcare outcomes by predicting lung cancer patient readmissions and aiding early detection in diseases like diabetic retinopathy.
-
Responsible AI and ethical considerations: Swathi stresses the need for responsible AI development to avoid amplifying biases, especially in healthcare, by ensuring inclusive data and diverse perspectives.
-
Challenges and ethical concerns: A five-step process for AI projects emphasizes starting with business problems and assessing organizational capabilities to ensure meaningful, scalable AI solutions.
Article and quotes have been edited for brevity and clarity
