Page 226 - ITU KALEIDOSCOPE, ATLANTA 2019

S5.2 A healthcare cost calculator for older patients over the first year after renal transplantation
Rui Fu, Nicholas Mitsakakis and Peter C. Coyte, University of Toronto, Canada
Forecasting tools that accurately predict post-transplantation healthcare use of older end-stage

renal disease (ESRD) patients are needed at the time of transplantation in order to ensure smooth
care delivery in the post-transplant period. We addressed this need by developing a machine-
learning-based calculator that predicts the cost of healthcare for older recipients of a deceased-
donor kidney over the first year following transplantation. Regression tree and regularized linear
regression methods, including ridge regression, lasso regression and elastic net regression were
explored on all cases of deceased-donor renal transplants performed for patients aged over 60 in
Ontario, Canada between March 31, 2002 and April 31, 2013 (N=1328), The optimal model
(lasso) identified age, membership of one of 14 regionalized Local Health Integration Networks,
blood type, sensitization, having diabetes as the primary case of ESRD, total healthcare costs in
the 12-month pre-workup period and the 6-month workup period to be inputs to the cost
calculator. This cost calculator, in conjunction with clinical outcome information, will aid health
system planning and performance to ensure better management of recipients of scarce kidneys.

S5.3 Automatic plan generating system for geriatric care based on mapping similarity and global
optimization

Fei Ma, Chengliang Wang and Zhuo Zeng, Chongqing University, China
The smart home is an effective means of providing geriatric care to increase the ability of the

elderly to live independently and ensure their health in daily life. However, the smart home is not
widely used because it is arduous to obtain a sensing devices selection plan. In this paper, the
accuracy of service selection and cost savings assumes enormous importance. Therefore, we
propose an automatically plan generating system for the elderly based on semantic similarity,
intuitionistic fuzzy theory, and global optimization algorithm, aiming at searching for an
optimized plan. Experiment results indicate that our approach can satisfy care demands and
provide an optimized plan of sensing devices selection.

Session 6: Data and artificial intelligence era

S6.1 Invited paper - Preparing for the AI era under the digital health framework
Shan Xu, Chunxia Hu and Dong Min, China Academy of Information and Communication

Technology (CAICT), China

Information and communication technology (ICT) for health has shown great potential to
improve healthcare efficiency, especially artificial intelligence (AI). To better understand the
influence of ICT technology on health, a framework of the digital health industry has been
proposed in this paper. Factors from the health industry and the ICT part are extracted to study
the interaction between two groups of component factors. Health factors include service and
management; and ICT factors include sensors, networks, data resources, platforms, applications
and solutions. The interaction between ICT and health can be traced through the development
history, from the stage of institutional informationization to regional informationization, and
finally to service intelligentization. Following such a developmental roadmap, AI was chosen as
one of the most powerful technologies to study the penetration effect and key development
trends from the perspectives of data, computing power and algorithms. The health industry will
be much improved or redefined in the coming AI era. To better understand the strengths,
weaknesses and limitations of AI for health, exogenous factors are discussed at the end of the
paper; preparations on collaboration mechanism; standardization and regulation have been
proposed for the sustainable development of digital health in the AI era.

– 206 –