“AI is a dealbreaker in many fields, and health care is no exception,” said Santosh. “Students, research scholars, professors and health care professionals can all benefit from this publication.”

The book, titled “Active Learning to Minimize the Possible Risk of Future Epidemics” explores the development of an AI-guided tool that could be used during an epidemic for mass screening of infected cases to mitigate the risk of spreading the disease to a large population.

In the possible scenario of future epidemics, a major challenge is the extensive time it takes to obtain thoroughly annotated data, requiring a time frame spanning months to years. Instead of waiting for big data, Santosh and Nakarmi employed active learning – also known as human or expert-in-the-loop learning, where a model is able to learn samples from day-one, like humans.

“The uniqueness resides in how we incorporated mentoring into deep learning models, similar to the guidance we receive from parents, teachers or coaches,” Santosh said of the research. “Human intervention, termed mentoring, occurs only when errors surface and there is limited data.”

Nakarmi, who graduated with his master’s degree in computer science in December 2023, said publishing a book was not only a confidence boost, but also a highly beneficial learning experience.

“I feel proud of the accomplishment of having a book published while pursuing my master’s degree, and all thanks to Dr. Santosh for the invaluable support,” said Nakarmi. “The experience has helped me build confidence in writing, extending to both short and long manuscripts, and grasp in-depth knowledge of active learning/machine learning. This skill set will undoubtedly help me when pursuing my Ph.D. degree.”

“It is rare to have a master’s student co-author a book – specifically scientific writing,” said Santosh. “A book with technicality, which we call implementation, provides students with a comprehensive understanding of project completion, writing, implementation and meeting deadlines. This experience also indicates their readiness to undertake complex projects in the future.”

In addition to writing and publishing “Active Learning to Minimize the Possible Risk of Future Epidemics,” Nakarmi said the experiences he had while in USD’s Department of Computer Science were unmatched.

“I got the opportunity to publish a conference paper, a journal paper and a book,” said Nakarmi. “Also, I had an opportunity to work as a machine learning intern at Synthetik Applied Technologies, in Austin, Texas, and participate as a member of the Technology Readiness Acceleration Center.

“In my opinion, the courses that the USD Department of Computer Science provides that are specific to data science and AI, the different clubs that students can participate in, the accessible faculty, the in-house High-Performance Computing (HPC) system and the visionary leadership of the department make it stand out,” he continued.

This is the second time Santosh has co-authored a book with a graduate student. In fall 2022, he and alumnus Casey Wall ’22 published “AI, Ethical Issues and Explain ability—Applied Biometrics.”

In 2020, Santosh published a similar publication on the intersection of health care and active learning, titled "AI-Driven Tools for Coronavirus Outbreak: Need of Active Learning and Cross-Population Train/Test Models on Multitudinal/Mutimodal Data."

"As of October 2023, this research article received enough citations to place it in the top 1% in the academic field of clinical medicine," said Santosh.

Of his 250 peer-reviewed research articles and 14 books, Santosh has published over 50 research articles and several books on medical imaging informatics and AI for health care, specifically focusing on COVID-19 and related diseases.

Press Contact
Hanna DeLange
Contact Email usdnews@usd.edu
Contact Website website