Page 160 - ITU KALEIDOSCOPE, ATLANTA 2019

2019 ITU Kaleidoscope Academic Conference

The management process in the health industrial framework These preparation factors sit alongside the health direct
is divided into two parts: equipment and staff. The workflow in healthcare institutions; thus, they can be defined
corresponding applications of AI are also listed as below. as exogenous factors that play an external effect on the health
industry. Figure 6 shows the complete framework of digital
➢ Staff and institution management: Personnel in health health with the consideration of exogenous factors including
institutions usually includes physicians of various collaboration mechanisms, standardization and regulation,
specialties, nurses, technicians specializing in specific etc.
equipment, administrative financial clerks and other
support personnel. Intelligent institution management
application could either refer to specific problems like
scheduling the nurse personnel, performance appraisal,
workload distribution, task assignment, patient
feedback collection and analysis, etc. [12]. Besides,
auxiliary talent training is another direction for AI to
improve staff management. Royal Philips' annual
health survey shows that in Singapore, about 37% of
medical professionals can use artificial intelligence to
support administrative tasks, only 28 % of them have
the digital literacy to use it for diagnosis. Auxiliary
talent training could be a customized education and
collaboration platform. InferScholar Center released in
March 2019 [30] is equipped with advanced models and
visualization tools for the clinical research. Ali Health
is trying to break down various clinical case data into a
three-dimensional “virtual patient” in the physician
training system of Ali ET Medical Brain [31]. Figure 6 – Effects from exogenous factors

➢ Equipment and drug operation: Health institutions 5.1 Collaboration mechanism
frequently and continuously use large equipment to
measure patients’ health data. Intelligently detecting The collaboration mechanism refers to the way that related
the operation of the equipment with IoT sensors and AI stakeholders collaborate and contribute together. AI for
analysis could avoid emergencies such as equipment health is an interdisciplinary integration and innovation
failures. Moreover, drug development could also be between the ICT and the health industry. Expertise from the
supported by AI. New drug development requires an health and AI community are of great importance to promote
averaged investment of 2.87 billion US dollars [32]. this cross-domain task. The mechanism can be considered in
Only 5 out of the 5 000 can be used in animal three aspects listed below.
experiments on average, and finally only 1 of them can
enter the clinical trial stage [33]. AI could help with ➢ Top-level design: National or overview strategy of AI
including target screening, drug mining and drug for health industrial plan can act as a guidance to gather
optimization to improve development efficiency. more industrial power. This top-level design may
Computer simulation calculates the ability of small include goal definition, demand analysis, strategic
molecules to the drug target, increasing the screening direction, priorities, timeline and role division. Many
speed and success rate, and eventually reducing the AI strategies were deployed, but not AI for health. This
development cycles and overall costs. kind of top-level design will give an overarching view
of the industry development and help form an industry
5. CORRESPONDING PREPARATION consensus to integrate scattered opportunities and
create a common blueprint.
With the increasing penetration of AI into health processes,
and considerable resources allocated to exploring the use of ➢ Information exchange: AI for health provides end-to-
AI for health, the era of AI for health is coming. 9.5 billion end service, which give higher requirements on product
dollars of venture investment is reported in global digital development. Effective ways for information exchange
health in 2018, which is an increase of more than 30% over can help gather more news from different roles in the
the previous year; and the corporate finance provided to production chain. Competition provides early
digital health companies totaled $13 billion, which is a 58% incentives to prototype in the early stages; industry
increase from $8.2 billion in 2017 [34]. Yet, due to the alliance is another flexible way to extensively gather
complexity of AI models, it is difficult to fully understand industrial forces and seek cooperation opportunities in
their strengths, weaknesses and limitations. The mature stages.
corresponding preparation is necessary, requiring serious
consideration. ➢ Financing: To improve the health level of mankind is a
huge project that needs strong support including

– 140 –