Medical data and operations are the most sensitive. While we trust AI in all sectors, the human health related area has a higher risk level. The risks are related to possible impact of error to persons life (misconducted diagnosis, robots malfunction during medical operations, etc.). As well the medical data falls into the most sensitive data category, linking it not only with human privacy questions, but biometry as well. While the potential of AI application is huge. Therefore the challenge invites to dive deeper in the analysis of AI application possibilities from legal regulation, ethical and technical aspects.
The direction for the challenge ideas, which could be converted into multiple final thesis or individual project topics is the following:
- A study of the possibilities and limitations of data mining technologies for the application of personalized medicine.
- Image augmentation strategies and its suitability for different type of medical images.
- Medical data anonymization strategies and its effect on the AI model accuracy.
- Early skin cancer detection using CNNs
- Neural networks for breast cancer identification and prognosis
Regarding more details on the challenge and topics contact Jelena Mamčenko (topic 1), Simona Ramanauskaitė (2-3 topics) and Jūratė Janutėnaitė Bogdanienė (topics 4-5), VILNIUS TECH.