Current Medical Science 41(6):1105-1115,2021

Current
DOI Medical Science 41(6):2021
https://doi.org/10.1007/s11596-021-2474-3 1105

Application of Artificial Intelligence in Medicine: An Overview

Peng-ran LIU†, Lin LU†, Jia-yao ZHANG, Tong-tong HUO, Song-xiang LIU, Zhe-wei YE#
Department of Orthopedics, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan
430022, China

The Author(s) 2021

[Abstract] Artificial intelligence (AI) is a new technical discipline that uses computer technology
to research and develop the theory, method, technique, and application system for the simulation,
extension, and expansion of human intelligence. With the assistance of new AI technology, the
traditional medical environment has changed a lot. For example, a patient’s diagnosis based on
radiological, pathological, endoscopic, ultrasonographic, and biochemical examinations has
been effectively promoted with a higher accuracy and a lower human workload. The medical
treatments during the perioperative period, including the preoperative preparation, surgical period,
and postoperative recovery period, have been significantly enhanced with better surgical effects.
In addition, AI technology has also played a crucial role in medical drug production, medical
management, and medical education, taking them into a new direction. The purpose of this review
is to introduce the application of AI in medicine and to provide an outlook of future trends.
Key words: artificial intelligence; medicine; application; overview

Artificial intelligence (AI) is a new technical significant role in technical improvements, thus leading
discipline that uses computer technology to research to a new conception: Artificial Intelligence plus (AI
and develop the theory, method, technique, and plus). AI plus takes advantage of the achievements and
application system for the simulation, extension, and technology of AI and combines them with traditional
expansion of human intelligence. The conception of AI industries to create new productivity, innovation,
first appeared in 1950 from the scientist Alan Turing, and development. AI research has demonstrated that
who is named as the “father of artificial intelligence”; the output-input ratio in medicine is more promising
he developed the “Turing test” and described AI as than that of other fields[7]. The combination of AI and
similar to but more complex than the human brain[1, 2]. medicine (AI plus medicine) changes the traditional
With the development of AI in recent years, medical model and makes a revolutionary promotion.
especially the appearance of deep learning (a branch of AI plus medicine has also attracted much attention due
computer learning algorithms and the core composition to its potential prospects and future. Therefore, this
of a new generation of AI technology, which can paper aims to review the latest applications of AI plus
automatically learn from big-data-analysis and then medicine in recent years.
artificially and independently make decisions upon the
knowledge, including various neural networks such as 1 AI IN MEDICAL DIAGNOSIS
the deep belief network, convolutional neural network,
long- and short-term memory network, etc.), there has When a clinician diagnoses a patient with a certain
been a surge of interest regarding this new technical illness/condition with the assistance of AI, the time
discipline and the promotion of a series of artificial required for a diagnosis can be greatly reduced and the
intelligent systems in practical applications, like the diagnostic efficiency can be significantly improved.
Internist-1 system[3], MYCIN system[4], CASNET By analyzing the clinical data from radiology (like
system[5], and also some databases and record systems[6]. X-ray, CT, and MRI), pathological, endoscopic,
Now AI is widely applied to multiple fields and plays a ultrasonographic, and biochemical examinations for
related human body indicators, AI can output results
Peng-ran LIU, E-mail: lprlprlprwd@163.com; Lin LU, quickly and change the ineffective traditional medical
E-mail: lledu2014@163.com model, which is unable to give timely and accurate
#
Corresponding author, E-mail: yezhewei@hust.edu.cn conclusions, especially for complicated diagnoses. In
†
The authors contributed equally to this work. addition, as AI can solve problems in such a short time,
*
This project was supported by the National Natural Science doctors can make a more deliberate and reasonable
Foundation of China (No. 81974355). treatment plan according to the patient’s condition.
1106 Current Medical Science 41(6):2021

1.1 AI in Radiology pathological scanning techniques and the upgrade of

As a scientific and intuitive basis for medical related software, whole-slide imaging technology has
diagnosis, radiology is presently involved in the become a routine diagnostic method in pathological
diagnostic process of almost all diseases. The demand work. However, there are still problems about how
for a radiological diagnosis is increasing at a high rate to automatically and rapidly analyze and obtain an
per year; however, the cultivation of medical talent accurate diagnosis from the convenient pathological
cannot be achieved overnight and the number of images, thus demanding a prompt solution. Some
medical doctors with radiation medicine experience is scholars have pointed out that “AI is the next step and
increasing slowly. As the gap between the supply and future of precision pathology” and have proposed a
demand of medical doctors in this field is constantly new blueprint[17–21]. As a means to predict a diagnosis,
increasing, high occupation pressure and misdiagnosis the application of AI in pathology has shown
rates are observed. Therefore, it is of great practical bright prospects. During the analytical work, with
significance to find other ways such as AI to temper the augmentation of the AI algorithm, the processes of
urgent situation. In radiological diagnosis, numerous pathological image segmentation, tumor identification,
applications of AI have appeared in recent years. For and metastasis determination have been promoted, and
example, during AI deep learning research, Francesco the work is finished in a higher quality and a shorter
found a new algorithm with a high sensitivity for period[22]. In addition, research has shown that even in
the early screening, rapid diagnosis, and grading of some cases, the AI algorithm outperforms professional
retinal diseases[8], which also has been confirmed in clinical pathologists for diagnosis based on
other research[9, 10]. In addition, Gong[11] invented an pathological images[23]. For example, Hart has utilized
AI-aided diagnostic (CADx) system based on clinical the convolutional neural network to distinguish Spitz
CT data to classify malignant and benign lung nodules and conventional melanocytic lesions, two different
by analyzing 243 patients with confirmed pulmonary pathological types, and the results show an extremely
nodules; he demonstrated the feasibility of the CADx high accuracy[24]. Kosaraju also has proposed an
system for accurately distinguishing the nature of lung innovative multi-task model using AI deep learning
nodules and the practicability of the system for early that can simultaneously take multi-scale patch images
and subtle lung cancer diagnosis. Moreover, his team for pathological image analysis. After assessing the
combined the quantitative imaging (QI) features (AI pathological images of well-, moderately, and poorly
technology) and serum biomarkers of 173 patients differentiated gastric cancer, the efficient and accurate
with pulmonary nodules with the system to improve performance of the new model outperformed other
its performance of pulmonary nodule classification and contemporary AI methods[25]. Moreover, Coudray
finally got an optimistic result, which demonstrated a confirmed that the deep learning model could predict
better CADx system performance by using QI features six genetic mutations associated with cancer and assist
than only using the serum biomarkers[12]. There is also doctors to detect the subtype and gene mutation in cancer
a study for the effectiveness between clinical doctors diagnosis, with a high accuracy of 97%; the model
and AI by Rodriguez-Ruiz[13], who compared the code is published online[26]. Furthermore, AI has been
stand-alone performance of an AI system to that of used in the diagnosis of epithelial tumors, lung cancer,
radiologists in breast cancer diagnosis by X-ray; based basal cell carcinoma, and glomerulosclerosis[27–31].
on 2652 exams overall, the performance of the AI These developments emphasize the practicability of AI
system was statistically not inferior to that of clinical technology applied in pathology.
radiologists, meaning that the AI system in this study 1.3 AI in Endoscopy
reached an accuracy equal to that of clinical radiologists As for AI in endoscopy, with AI technology
with expertise in breast cancer diagnosis. Similarly, augmentation, there have been great advances in
breast cancer screening augmented with intelligent endoscopy detection that have changed the traditional
technology defeated clinical radiologists, showing a model and improved the efficiency. According to
lower misdiagnosis rate and a workload decrease of some experts, AI technology can effectively promote
88%[14]; additionally, the automatic preselect function the detection of lesions, colorectal polyps, as well as
significantly reduced the pressures on clinicians[15]. gastric and esophageal cancer using endoscopy[32].
Furthermore, in MRI data analysis, an AI algorithm Gulat and Emmanuel have pointed out that endoscopy
also has been shown to be superior to that of human is an attractive technology for AI augmentation with
observers, immensely improving the positive rate of immense potential; after deep learning, the AI system
subtle inflammatory detection in early rheumatoid can significantly enhance the diagnosis of stomach
arthritis[16]. and intestinal diseases, including Barrett’s esophagus,
1.2 AI in Pathology squamous carcinoma, and gastric cancer, by shortening
Pathology is the footstone of the diagnosis of the detection time and improving the diagnostic
tumors and other lesions. With the development of accuracy[33]. In addition, some scholars have collected
Current Medical Science 41(6):2021 1107
7556 clinical images by endoscopy and then analyzed stage, AI plus diagnosis has been the general trend
them by AI technology to provide a practical neural and will continue to flourish.
network algorithm to automatically detect bowel
lesions; the results showed that endoscopy combined 2 AI IN MEDICAL TREATMENT
with the new AI algorithm had a higher sensitivity and
more accurate localization of the bowel lesions than with 2.1 AI in Surgery
the traditional model[34]. With more and more research In the surgical field, the most outstanding
confirming the feasibility of AI plus endoscopy in the achievement and application of AI could not be without
diagnosis and classification of various diseases[35–37], the surgical AI system. About 20 years ago, PUMA-
there is obviously a future for this new technology. 560, Probot, AESOP, Robodoc, and Acrobot[50–52]
1.4 AI in Ultrasonography and Biochemical Exa- played an favorable auxiliary role in surgery. But
minations in the early phase, all of the surgical systems could
Similar to above, the application of AI technology not operate without human control, meaning that the
also raises the level of diagnosis using ultrasonography systems were just another kind of unintelligent scalpel
and biochemical tests. Although image-based computer- that was more flexible.
aided diagnosis (CAD) systems have already been With the development of AI technology, the
applied by doctors to diagnosis through ultrasound, the conception of an AI plus surgical system had been raised.
performance is largely dependent on the detection and In the contemporary era, the most groundbreaking
classification methods. Combined with AI technology, production of this conception is the Da Vinci surgical
methods have changed a lot. For example, Nguyen AI system. As a great invention unprecedented in
has proposed a new ultrasonographic image analysis human history, the emergence of the Da Vinci surgical
method based on AI that successfully enhanced the system makes surgical treatment more minimally
consequence of thyroid nodule classification[38]. Other invasive, with the advantages of a clearer image, more
scholars have also confirmed that the use of AI can accurate and convenient operation, and even remote
promote the traditional ultrasonographic detection of operation. This creative invention allows complex
tumors in the thyroid, breast, bronchia, puborectalis surgical operations to be performed through minimally
muscle, and urogenital hiatus as well as other obstetric invasive methods that have been difficult in the past.
and gynecological lesions, with a high efficiency and There are three components in the Da Vinci surgical
accuracy[39–44]. AI system: surgeon console, manipulator operating
What’s more, after a comprehensive learning system, and imaging system. In 2000, the U.S. Food
analysis of clinical examination data, based on and Drug Administration approved the application of
the model of big-data analysis, AI has also made the Da Vinci surgical system in clinical surgery. This AI
significant progress in clinical disease diagnosis. By system revolutionized the traditional surgical model.
utilizing deep sequencing to analyze genes that are For example, with the application of the Da Vinci
recurrently mutated in acute myeloid leukemia (AML) surgical AI system, thyroid surgery was improved
and large electronic health record database analysis, in terms of the postoperative cosmesis and voice
Abelson has developed an AML predictive system outcomes[53], maxillary surgery was improved in terms
contributing to earlier detection and monitoring[45]. of the accuracy and safety[54], gastric, nephritic, and
In addition, Sun successfully predicted the radiomic prostatic surgery was improved as shown by the high
signature and clinical outcomes in anti-PD-1/PD- surgical success rate but low complication rate[55–57],
L1 immunotherapy by analyzing the CT images and and lung cancer surgery was beneficial to the patients
RNA sequencing data from patients with malignant in terms of the postoperative recovery[58]. Except for
tumors using an AI algorithm[46]. By analyzing the the above points of surgical operation enhancement,
gene and related clinical features, AI also facilitates the compared to traditional surgical systems, the most
diagnosis of Noonan syndrome, a common autosomal outstanding characteristic of AI surgical systems
disorder, especially for atypical patients[47]. Moreover, particularly lies in “AI”, meaning that the surgical
by applying AI deep learning to set up a combined systems progressed from a nonintelligent form to an
model of multiple clinical tests like lung function test, intelligent form. Profited by the algorithms like deep
bronchial experiments, and some biochemical tests learning of AI technology, the histological diagnosis in
from 556 patients, the prediction and diagnosis of vivo and in situ during surgery stands on the stage of
initial asthma were remarkably promoted[48]. pathology, making efficient incisal edge pathological
However, there are also scholars who hold analysis and real-time tissue biopsy a reality[59]. By
the opposite opinion that AI has a place in clinical using deep learning, the AI algorithm can also self-
diagnosis, but the real influence of it may occur deduct based on the abundant experiments from
gradually over a long period and will not replace clinical surgeons and reconstruct clinical digitized data
clinicians in the short term[49]. Overall, at the current by uploading the surgical program to an AI surgical
1108 Current Medical Science 41(6):2021

system to assist the surgery intelligently, including operator’s confidence for the surgery[64–67]. Not only
surgical excision range formulation, postoperative in the preoperative preparation but also during the
organ residual volume insurance, and prediction of surgical period, 3DP still plays a crucial role in its next
lymph nodes with possibly positive metastasis[60]. stage–Surgical Guide. In a surgical operation, there are
The surgical planning and methods not only rely on always some demands of internal fixation and cutting
surgeons, but they also depend on the program using for orthotics or tumor excision, so problems may arise,
the intelligent algorithm[61]. Of course, in the current such as how to determine the appropriate angle and
stage, although AI surgical systems have achieved location of fixation for maximum effectiveness, and
partial intelligence, they also need human supervision where the cutting edge should be to retain more normal
to a certain extent. However, this point will be further tissue as far as possible. Through the preoperative
developed and be a hotspot with a bright future, and detection data, 3DP can produce an individualized
one day whole intelligence will be realized. surgical guide and a template for assisting the surgery.
2.2 AI in the Perioperative Period With 3DP template navigation in spine surgery, pedicle
The perioperative period is the period around screw insertion received a safety enhancement and
the whole operation, from the patient receiving was much easier than traditional methods; meanwhile,
surgical treatment to a basic recovery; it includes the risk of surrounding neurovascular damage was
three parts: preoperative preparation, surgical period, effectively reduced, and the radiation exposure was
and postoperative recovery period. During the whole also decreased[68, 69]. Moreover, applications of 3DP
process of the perioperative period, there are also lots of template navigation in osteotomy surgeries for tibial
achievements with the application of AI technologies. deformities and for total knee arthroplasty have been
2.2.1 Three-dimensional Printing (3DP) 3DP is a reported; compared with conventional methods, the
technology that partly employs AI technology during its operation time and efficiency were significantly
processes. It is one type of rapid prototyping technology improved with 3DP technology[70–72]. As for tumor
that uses powdered metal or other adhesive biomaterials excision, 3DP technology helps to precisely locate
to construct objects by layer-by-layer printing based on and confirm the cutting edge in bone tumor surgery,
digital model files (created from CT or MRI data with providing a satisfactory postoperative result[73] and also
AI technology). The clinical imaging data are imported reducing the risk of critical structure injury and saving
into intelligent software like MIMICS. After artificially more normal tissues[74]. What’s more, the newest
selecting the regions of interest by the operator, the stage of 3DP technology–Body Implant, magically
software can conveniently output a primary virtual accomplishes the reconstruction of human tissue
three-dimensional reconstruction by algorithm analysis with bioactive materials including scaffold materials,
for printing. Although further processing by humans functional cells, and active factors. After printing and
may be necessary at the current stage, we believe that sterilization, the implant can be applied during surgery
one day it will achieve whole intelligence. Once the to replace the injured and defective human tissue due
technology was applied in medicine, it vastly promoted to various reasons. For instance, the application of
the development of medicine, especially in surgery. polymers, bioceramics, and composites as bioinks that
For example, during the preoperative preparation, in print personalized bone scaffolds have been shown to
the face of some complex visceral injuries or bone enhance the surgical effect and patient satisfaction;
fractures, it is difficult for clinical surgeons to recognize among those receiving large mandibular defect
the key point by traditional detections in emergency reconstruction surgery, the defective tissue was fully
circumstances. However, combined with the initial repaired through 3DP technology[75]. In orthopedic
stage of 3DP technology–Model Printing, doctors can surgery, combined with the mirror-replication
hold the 1:1 real model of the injured part reestablished technique, 3DP technology also brings hope to the
from the actual CT scanning data, get more visual and intractable challenge of cranial defects and limb bone
intuitive information, make more detailed preoperative defects, which used to rely on bone transplantation
plans, and even practice a simulated surgery on the (a surgery with lots of complications) to reestablish
model in advance[62]. Scholars have indicated that bone using traditional surgical methods[76–78]. Similarly,
in cardiac and vascular surgery, 3DP technology 3DP substitution also has been widely applied in the
provides a characteristic patient-specific model that fields of neurodegenerative disease, skeletal muscle
recognizes complex anatomy and is helpful for injury reestablishment, arthroplasty, aortic valve replacement,
orientation, plan-making, and communication with and pelvic diseases in urology and gynecology[79–83].
patients[63]. In addition, much research has shown that Furthermore, in the near future, 3DP technology will
3DP plays a vital role in the preoperative preparation accomplish the goal of printing whole functional-living
for dental surgery, orthopedic surgery, spine surgery, organs, for the next stage–Organ Bioengineering[84].
urological surgery, and some tumor surgeries, from 2.2.2 Virtual Reality, Augmented Reality, and Mixed
strengthening preoperative planning to enhancing the Reality Virtual reality (VR), augmented reality (AR),
Current Medical Science 41(6):2021 1109
and mixed reality (MR) technologies are new types of complications during anesthesia. Combined with the
digital holographic image technologies that are similar application of AI technology, six aspects have been
to 3DP in that they partly employ AI technology mostly promoted and have received extensive attention:
to reconstruct clinical data during their processes. (1) anesthesia depth monitoring; (2) anesthesia control;
VR is a pure virtual digital image generated by an (3) adverse event prediction; (4) ultrasound assistance;
intelligent computer algorithm that can provide some (5) pain control; and (6) operating room management[102].
opportunities for surgeons to practice using a virtual AI technology increases the safety of monitoring,
system, without any serious consequences of operation delivery, and postoperative management, thus bringing
failure, thus improving their surgical abilities[85, 86]. promising developments for anesthesiology[103, 104].
However, due to the lack of real-world experience, 2.2.4 Rehabilitation Assistance In the field of
VR cannot be applied in real surgery[87]. AR is a postoperative rehabilitation, AI technology also
composition of intelligent augmented information and plays a crucial role in the process of recovery.
the real environment, and, in general, it is different from For instance, in the intensive care unit (ICU), the
VR according to its specialty in the real world. After application of AI wireless sensors can effectively
conversion of patient data and virtual reestablishment collect patient information, reduce false alarms, and
of the critical area, with the addition of the virtual relieve challenges in the ICU[105]. With the gradual
image to the real visual world, AR technology can diversification of AI technology, there have been many
validly assist surgery preoperatively or intraoperatively new tools (monitoring and remote management) in
by recognizing the complex anatomical structures and the field of nursing[106]. The AI-based medical devices
navigation during the operation[88, 89]. However, due can be helpful during patient recovery, meeting the
to unwieldy equipment for AR navigation systems, requirements of rehabilitation and expediting the
there are still limitations during surgery[90]. With the proceedings[107]. In addition, the application of AI robots
appearance of the newest digital holographic imaging also has accelerated limb rehabilitation in complex
technology MR, the combination of VR and MR breaks anthropopathic action guidance and helped patients to
the boundary between virtuality and reality, effectively obtain a better degree of recovery[108, 109]. What’s more,
solving the problem. There are three features in MR: AI technology also has been used to track progression
closed combination among virtuality and reality, real- and to monitor health, which may be beneficial for the
time interaction, and precise matching[91]. In the MR management of discharged patients[110, 111].
system, which is comprised of comparably portable
equipment (e.g., a wearable MR device, the Hololens, 3 AI IN DRUG PRODUCTION
and the latest Microsoft technological production[92]),
a real-time interactive location, and vivid visual In the traditional model, the production of drugs
experiences, the clinical surgeon can immerse requires a long period, including functional target
themselves in the mixed surgical world and formulate studies, drug ingredient design studies, performance
a better therapeutic schedule; in addition, the doctor- tests, clinical trials, testing, and promotion; thus,
patient communication is improved[93]. Due to these even after a long period of research, new drugs may
advantages, this new technology has of course been not necessarily work as well as expected. However,
applied in a variety of fields, such as spine, orthopedic, with the development of AI in recent years, the new
liver, kidney, and skull surgeries, for intraoperative technology has changed the traditional drug industry
guidance assistance, thus shortening the operation time in healthcare and has facilitated new drug discovery
and promoting the accuracy and safety of surgery[94–98]. and assembly[112, 113]. Moreover, following the gradual
Additionally, MR has even met the requirements of maturity of AI-generation drugs, both the novelty
telemedicine, which is essential for health care in rural and quality of drugs have reached new heights[114].
and remote areas, as information can be shared in real For instance, the combination of AI prediction
time using online chatting platforms[99]. Compared with models and vaccine design has efficiently accelerated
the 3DP technology mentioned above, except the more clinical trial processes and cut down the research and
accurate navigation, MR still has some advantages in development costs and time period[115]. Deep learning
timeliness due to the fact that 3DP production may technology-guided drug discovery can target proteins
take several hours for printing[90]. What’s more, MR as designed, which would once be impossible to
technology also can be used to assist with postoperative achieve[116]. Benefited by the strong logical deduction
rehabilitation and routine training[100, 101]. and automatic learning abilities of AI technology, the
2.2.3 Anesthesiology Assistance AI technology design and production of cancer drugs were profoundly
also has been widely applied in anesthesiology during optimized with a better therapeutic performance[117].
the perioperative period. Anesthesia is an important Additionally, the research of AI-assisted bioinformatics
part of the surgical procedure that helps provide a tools and methods also has provided a bright future
smooth operation; however, there are lots of risks and for small molecule drug therapy[118]. What’s more, the
1110 Current Medical Science 41(6):2021

3DP technology mentioned above also has brought results with better performance and confidence of the
enormous development in drug production. 3DP medical students[127]. Moreover, the AI simulation-
in drug production possesses patient-customized based surgical training system that combines AI and
characteristics as well as accomplishes the selection simulation together for studying surgical techniques has
of drug size, shape, and combination of different created a new educational tool with objective feedback,
pharmaceutical ingredients, which may be more which is beneficial for student learning[128]. Not only
convenient for clinical applications[119]. Through 3DP for assistance with learning but also for supervising, AI
technology, even the layers and percentage parameters technology had availed the monitoring of mental health
in tablet coating, therapeutic release rates and patterns and study performance of students, thus enabling
can also be predesigned, thus providing better curative universities to know the conditions of their students in
effects[120, 121]. a timelier fashion[129]. What’s more, the 3DP and MR
technology mentioned above can also provide medical
4 AI IN MEDICAL MANAGEMENT AND EDU- students with more vivid learning opportunities, which
CATION are unavailable in conventional textbook reading.
With the assistance of intelligent algorithms, three-
In the traditional model, the medical management dimensional reconstruction is different from two-
in hospital relies on the overall planning of the hospital dimensional books; therefore, students can utilize the
administrative department, and there always are some 3DP medical model for studying three-dimensional
management omissions and disadvantages like the anatomical structures and even practice operations
unreasonable distribution of medical resources. With on the model to improve their surgical skills[130, 131].
AI technology regulation, procedures have changed Furthermore, the MR technology can also help students
quite a bit. Some scholars have used the long-short- understand the human anatomy in a more intuitive
term memory neural network AI technology to build a manner with any size or layer that they control, thus
prediction model and to analyze the database of patient providing simulative surgery training with no risks[132].
hospital-stay time, successfully accomplishing the Currently, 3DP or MR-based assistance methods have
prediction of accurate waiting times in the emergency been widely applied in medical education.
department of their hospital, which effectively enhanced
the medical efficiency and patients’ subjective experience 5 AI IN CORONAVIRUS RESEARCH
and promoted medical resource redistribution[122].
Furthermore, research that analyzed the data of patient At the end of 2019, the new coronavirus disease
hospital-stay time, the route to hospital, as well as 2019 (COVID-19) outbreak brought threats to the world.
climatic and temporal elements with 10 AI algorithms Human health, safety, human civilization development,
was used to reduce the average hospitalization time by and the global economy were severely affected, and
7%, select the optimum number of hospital beds, and countless people died in the global disaster. Fortunately,
optimize hospital resources and required inputs[123]. huge achievements in the prophylaxis and treatment of
In addition, a real-time-prediction model based on COVID-19 were made by relying on several advanced
artificial neuron networks comparatively and accurately medical methods and highly developed technologies,
predicted the readmission rate, thus helping to make including AI technology. In the race to control the
preparations for patients and improving hospital spread of COVID-19, AI was used to work as human
management[124]. In short, AI technology has facilitated intelligence in order to address the following: early
patient counseling, hospital management, medical detection and diagnosis, treatment monitoring, contact
resource allocation, and ultimately individualized tracing, prediction of cases and mortality, development
clinical care[125]. of drugs and vaccines, medical workload reduction,
The education of medical students is the future and disease prevention[133]. During the fight against
and hope in medical developments; however, due COVID-19, due to the fact that the quantification and
to the massive and intricate professional knowledge localization of pulmonary lung CT data cannot be
required, the training period of medical students is long accurately and efficiently evaluated, Zhang developed
and difficult. The development of medical students a new system based on deep learning to analyze the CT
will be hindered if they only study medical books data of patients and concluded that the right lower lobe
and specimens. With the diversified application of AI of the lung is the high occurrence area of COVID-19
technology, the learning pattern of medical students pneumonia[134]. In addition, Aikaterini applied an
has been richer and more colorful. AI-based problem- AI machine learning algorithm in the analysis of
based learning has improved student’s learning and COVID-19 CT scans, and their findings indicated
understanding, thus enhancing their knowledge of that the algorithm could promote earlier detection and
clinical diseases[126]. The combination of studying medical care[135]. Moreover, Tivani proposed point-
surgery with an AI system also has provided promising of-care diagnostic services that blended radiology,
Current Medical Science 41(6):2021 1111
pathology, and artificial intelligence all together, further mmons.org/licenses/by/4.0/), which permits use, sharing,
assisting the diagnosis of COVID-19[136]. Additionally, adaptation, distribution and reproduction in any medium or
Sweta performed a quick intelligent screening for format, as long as you give appropriate credit to the original
potential drugs to treat COVID-19 with a drug- author(s) and the source, provide a link to the Creative
Commons licence, and indicate if changes were made.
repositioning method; this group was able to quickly
The images or other third party material in this article
detect drugs that may be useful by using both AI- and are included in the article’s Creative Commons licence,
pharmacology-based methods, thus demonstrating unless indicated otherwise in a credit line to the material. If
that this method could be helpful for COVID-19 material is not included in the article’s Creative Commons
drug design and research[137]. This method also has licence and your intended use is not permitted by statutory
been confirmed by other scholars, who established a regulation or exceeds the permitted use, you will need to
platform based on AI learning and prediction models to obtain permission directly from the copyright holder. To
identify the drugs on the market with a possibility for view a copy of this licence, visit http://creativecommons.
treating COVID-19; as a result, they found more than org/licenses/by/4.0/.
80 drugs with great potential[138]. Furthermore, there
Conflict of Interest Statement
has been much research on AI algorithm assistance,
The authors declare that they have no conflicts of
which pushed the quick development of COVID-19 interest.
vaccines[139–141].
In general, with suitable AI-based technology, the REFERENCES
process of early warning, diagnosis, drug research, and 1 Mintz Y, Brodie R. Introduction to artificial intelligence
medical control during the fight against COVID-19 in medicine. Minim Invasive Ther Allied Technol,
will be effectively guaranteed[142], and the pandemic 2019,28(2):73-81
will be overcome in the near future. 2 Kaul V, Enslin S, Gross SA. The history of artificial
intelligence in medicine. Gastrointest Endosc, 2020,
6 FUTURE OF AI 92(4):807-812
3 Miller RA, Pople HJ, Myers JD. Internist-1, an
AI technology is a high-tech production that experimental computer-based diagnostic consultant for
general internal medicine. N Engl J Med, 1982,307(8):
adapts to the development of the contemporary era;
468-476
therefore, it is the inevitable result of the advancement 4 Shortliffe EH, Davis R, Axline SG, et al. Computer-
of science and technology and follows the trend of based consultations in clinical therapeutics: explanation
time. There have been two industrial revolutions and rule acquisition capabilities of the MYCIN system.
in human society, steam revolution and electrical Comput Biomed Res, 1975,8(4):303-320
revolution, which both profoundly changed the way 5 Weiss S, Kulikowski CA, Safir A. Glaucoma consultation
of human life and promoted human civilization. Now by computer. Comput Biol Med, 1978,8(1):25-40
the scientific and technical revolution, including AI 6 Kulikowski CA. Beginnings of Artificial Intelligence
technology, has already shown an irresistible trend that in Medicine (AIM): Computational Artifice Assisting
Scientific Inquiry and Clinical Art - with Reflections
has grown vigorously like a raging fire. In the medical
on Present AIM Challenges. Yearb Med Inform, 2019,
field, with the assistance of new AI technologies, the 28(1):249-256
traditional medical environment has changed a lot, 7 Patel VL, Shortliffe EH, Stefanelli M, et al. The coming
and patient diagnosis using radiological, pathological, of age of artificial intelligence in medicine. Artif Intell
endoscopic, ultrasonographic, and biochemical Med, 2009,46(1):5-17
examinations were effectively promoted with a higher 8 Sorrentino FS, Jurman G, De NK, et al. Application of
accuracy and a lower human workload. The medical Artificial Intelligence in Targeting Retinal Diseases.
treatments during the perioperative period, including Curr Drug Targets, 2020,21(12):1208-1215
the preoperative preparation, surgical period, and 9 Heydon P, Egan C, Bolter L, et al. Prospective
postoperative recovery period, were significantly evaluation of an artificial intelligence-enabled algorithm
for automated diabetic retinopathy screening of 30 000
enhanced with better surgical effects. In addition, AI
patients. Br J Ophthalmol, 2021, 105(5):723-728
technology also has played a crucial role in medical 10 Xie Q, Liu Y, Huang H, et al. An innovative method
drug production, medical management, and medical for screening and evaluating the degree of diabetic
education, taking them into a new direction. The future retinopathy and drug treatment based on artificial
of AI has come, and we believe that the new revolution intelligence algorithms. Pharmacol Res, 2020,159:
will be swift like the wind and bring our medical field 104986
into an unprecedented new era. 11 Gong J, Liu JY, Sun XW, et al. Computer-aided
diagnosis of lung cancer: the effect of training data sets
Open Access on classification accuracy of lung nodules. Phys Med
This article is licensed under a Creative Commons Biol, 2018,63(3):35036
Attribution 4.0 International License https://creativeco- 12 Gong J, Liu JY, Jiang YJ, et al. Fusion of quantitative
imaging features and serum biomarkers to improve
1112 Current Medical Science 41(6):2021

performance of computer-aided diagnosis scheme histopathology images with a deep neural network. Br
for lung cancer: A preliminary study. Med Phys, J Dermatol, 2020,182(3):754-762
2018,45(12):5472-5481 31 Bueno G, Fernandez CM, Gonzalez L, et al.
13 Rodriguez RA, Lang K, Gubern MA, et al. Stand-Alone Glomerulosclerosis identification in whole slide
Artificial Intelligence for Breast Cancer Detection in images using semantic segmentation. Comput Methods
Mammography: Comparison With 101 Radiologists. J Programs Biomed, 2020,184:105273
Natl Cancer Inst, 2019,111(9):916-922 32 Namikawa K, Hirasawa T, Yoshio T, et al. Utilizing
14 McKinney SM, Sieniek M, Godbole V, et al. International artificial intelligence in endoscopy: a clinician's guide.
evaluation of an AI system for breast cancer screening. Expert Rev Gastroenterol Hepatol, 2020:1-18
Nature, 2020,577(7788):89-94 33 Gulati S, Emmanuel A, Patel M, et al. Artificial
15 Rodriguez RA, Lang K, Gubern MA, et al. Can we intelligence in luminal endoscopy. Ther Adv Gastrointest
reduce the workload of mammographic screening Endosc, 2020,13:2631774520935220
by automatic identification of normal exams with 34 Hwang Y, Lee HH, Park C, et al. An Improved
artificial intelligence? A feasibility study. Eur Radiol, Classification and Localization Approach to Small
2019,29(9):4825-4832 Bowel Capsule Endoscopy Using Convolutional Neural
16 Stoel BC. Artificial intelligence in detecting early RA. Network. Dig Endosc, 2020,33(4):598-607
Semin Arthritis Rheum, 2019,49(3S):S25-S28 35 He YS, Su JR, Li Z, et al. Application of artificial
17 Acs B, Rantalainen M, Hartman J. Artificial intelligence intelligence in gastrointestinal endoscopy. J Dig Dis,
as the next step towards precision pathology. J Intern 2019,20(12):623-630
Med, 2020,288(1):62-81 36 Chahal D, Byrne MF. A primer on artificial intelligence
18 Allen TC. Regulating Artificial Intelligence for a and its application to endoscopy. Gastrointest Endosc,
Successful Pathology Future. Arch Pathol Lab Med, 2020,92(4):813-820
2019,143(10):1175-1179 37 Sharma P, Pante A, Gross SA. Artificial intelligence in
19 Bera K, Schalper KA, Rimm DL, et al. Artificial endoscopy. Gastrointest Endosc, 2020,91(4):925-931
intelligence in digital pathology - new tools for 38 Nguyen DT, Pham TD, Batchuluun G, et al. Artificial
diagnosis and precision oncology. Nat Rev Clin Oncol, Intelligence-Based Thyroid Nodule Classification Using
2019,16(11):703-715 Information from Spatial and Frequency Domains. J
20 Serag A, Ion MA, Qureshi H, et al. Translational AI Clin Med, 2019,8(11):1976
and Deep Learning in Diagnostic Pathology. Front Med 39 Nguyen DT, Kang JK, Pham TD, et al. Ultrasound
(Lausanne), 2019,6:185 Image-Based Diagnosis of Malignant Thyroid
21 Zemouri R, Devalland C, Valmary DS, et al. Nodule Using Artificial Intelligence. Sensors (Basel),
Neural network: A future in pathology? Ann Pathol, 2020,20(7):1822
2019,39(2):119-129 40 Sun C, Zhang Y, Chang Q, et al. Evaluation of a deep
22 Wang S, Yang DM, Rong R, et al. Pathology Image learning-based computer-aided diagnosis system for
Analysis Using Segmentation Deep Learning distinguishing benign from malignant thyroid nodules
Algorithms. Am J Pathol, 2019,189(9):1686-1698 in ultrasound images. Med Phys, 2020,47(9):3952-3960
23 Komura D, Ishikawa S. Machine learning approaches for 41 Chen J, You H, Li K. A review of thyroid gland
pathologic diagnosis. Virchows Arch, 2019,475(2):131- segmentation and thyroid nodule segmentation methods
138 for medical ultrasound images. Comput Methods
24 Hart S N, Flotte W, Norgan AP, et al. Classification Programs Biomed, 2020,185:105329
of Melanocytic Lesions in Selected and Whole-Slide 42 Fujioka T, Mori M, Kubota K, et al. Breast Ultrasound
Images via Convolutional Neural Networks. J Pathol Image Synthesis using Deep Convolutional Generative
Inform, 2019,10:5 Adversarial Networks. Diagnostics (Basel), 2019,9(4):
25 Kosaraju SC, Hao J, Koh HM, et al. Deep-Hipo: Multi- 176
scale receptive field deep learning for histopathological 43 Chen CH, Lee YW, Huang YS, et al. Computer-aided
image analysis. Methods, 2020,179:3-13 diagnosis of endobronchial ultrasound images using
26 Coudray N, Ocampo PS, Sakellaropoulos T, et al. convolutional neural network. Comput Methods
Classification and mutation prediction from non-small Programs Biomed, 2019,177:175-182
cell lung cancer histopathology images using deep 44 Noort F, Vaart CH, Grob A, et al. Deep learning enables
learning. Nat Med, 2018,24(10):1559-1567 automatic quantitative assessment of puborectalis
27 Iizuka O, Kanavati F, Kato K, et al. Deep Learning muscle and urogenital hiatus in plane of minimal hiatal
Models for Histopathological Classification of Gastric dimensions. Ultrasound Obstet Gynecol, 2019,54(2):
and Colonic Epithelial Tumours. Sci Rep, 2020,10(1): 270-275
1504 45 Abelson S, Collord G, Ng S, et al. Prediction of acute
28 Kanavati F, Toyokawa G, Momosaki S, et al. Weakly- myeloid leukaemia risk in healthy individuals. Nature,
supervised learning for lung carcinoma classification 2018,559(7714):400-404
using deep learning. Sci Rep, 2020,10(1):9297 46 Sun R, Limkin EJ, Vakalopoulou M, et al. A radiomics
29 Wang S, Yang DM, Rong R, et al. Artificial Intelligence approach to assess tumour-infiltrating CD8 cells and
in Lung Cancer Pathology Image Analysis. Cancers response to anti-PD-1 or anti-PD-L1 immunotherapy:
(Basel), 2019,11(11): 1673 an imaging biomarker, retrospective multicohort study.
30 Jiang YQ, Xiong JH, Li HY, et al. Recognizing Lancet Oncol, 2018,19(9):1180-1191
basal cell carcinoma on smartphone-captured digital 47 Li X, Yao R, Tan X, et al. Molecular and phenotypic
Current Medical Science 41(6):2021 1113
spectrum of Noonan syndrome in Chinese patients. Clin in Minimally Invasive Spine Surgery. Curr Rev
Genet, 2019,96(4):290-299 Musculoskelet Med, 2019,12(4):425-435
48 Tomita K, Nagao R, Touge H, et al. Deep learning 67 Bangeas P, Tsioukas V, Papadopoulos V N, et al. Role
facilitates the diagnosis of adult asthma. Allergol Int, of innovative 3D printing models in the management
2019,68(4):456-461 of hepatobiliary malignancies. World J Hepatol,
49 Demircioglu A. Radiomics-AI-based image analysis. 2019,11(7):574-585
Pathologe, 2019,40(Suppl 3):271-276 68 Feng ZH, Li XB, Phan K, et al. Design of a 3D navigation
50 Jakopec M, Harris SJ, Rodriguez YB, et al. The first template to guide the screw trajectory in spine: a step-
clinical application of a “hands-on” robotic knee surgery by-step approach using Mimics and 3-Matic software. J
system. Comput Aided Surg, 2001,6(6):329-339 Spine Surg, 2018,4(3):645-653
51 Cowley G. Introducing “Robodoc”. A robot finds his 69 Kashyap A, Kadur S, Mishra A, et al. Cervical pedicle
calling--in the operating room. Newsweek, 1992,120(21): screw guiding jig, an innovative solution. J Clin Orthop
86 Trauma, 2018,9(3):226-229
52 Stefano GB. Robotic Surgery: Fast Forward to 70 Corona PS, Vicente M, Tetsworth K, et al. Preliminary
Telemedicine. Med Sci Monit, 2017,23:1856 results using patient-specific 3d printed models to
53 Tae K. Robotic thyroid surgery. Auris Nasus Larynx, improve preoperative planning for correction of post-
2020,48(3):331-338 traumatic tibial deformities with circular frames. Injury,
54 Stefanelli LV, Mandelaris GA, Franchina A, et al. 2018,49 Suppl 2:S51-S59
Accuracy Evaluation of 14 Maxillary Full Arch Implant 71 Sun ML, Zhang Y, Peng Y, et al. Accuracy of a Novel
Treatments Performed with Da Vinci Bridge: A Case 3D-Printed Patient-Specific Intramedullary Guide to
Series. Materials (Basel), 2020,13(12):2806 Control Femoral Component Rotation in Total Knee
55 Lenfant L, Wilson CA, Sawczyn G, et al. Single- Arthroplasty. Orthop Surg, 2020,12(2):429-441
Port Robot-Assisted Dismembered Pyeloplasty With 72 Zhou F, Xue F, Zhang S. The application of 3D printing
Mini-Pfannenstiel or Peri-Umbilical Access: Initial patient specific instrumentation model in total knee
Experience in a Single Center. Urology, 2020,143:147- arthroplasty. Saudi J Biol Sci, 2020,27(5):1217-1221
152 73 Park JW, Kang HG, Kim JH, et al. The application of
56 Winder A, Strauss DC, Jones RL, et al. Robotic surgery 3D-printing technology in pelvic bone tumor surgery. J
for gastric gastrointestinal stromal tumors: A single Orthop Sci, 2021,26(2):276-283
center case series. J Surg Oncol, 2020, doi: 10.1002/ 74 Gomez JM, Estades FJ, Meschian CS, et al. Internal
jso.26053. Online ahead of print Hemipelvectomy and Reconstruction Assisted by 3D
57 Jones R, Dobbs RW, Halgrimson WR, et al. Single Printing Technology Using Premade Intraoperative
port robotic radical prostatectomy with the da Vinci Cutting and Placement Guides in a Patient With
SP platform: a step by step approach. Can J Urol, Pelvic Sarcoma: A Case Report. JBJS Case Connect,
2020,27(3):10263-10269 2019,9(4):e60
58 Wang Y, Meng D, Sun X, et al. A Prospective Study 75 Salah M, Tayebi L, Moharamzadeh K, et al. Three-
of Da Vinci Surgical Robotic System with Chest Wall dimensional bio-printing and bone tissue engineering:
External Nursing Interventions. Chin J Lung Cancer technical innovations and potential applications in
(Chinese), 2020,23(6):487-491 maxillofacial reconstructive surgery. Maxillofac Plast
59 Zuo S, Yang GZ. Endomicroscopy for Computer and Reconstr Surg, 2020,42(1):18
Robot Assisted Intervention. IEEE Rev Biomed Eng, 76 Alkhaibary A, Alharbi A, Alnefaie N, et al. Cranioplasty:
2017,10:12-25 A Comprehensive Review of the History, Materials,
60 Navarrete AJ, Hashimoto DA. Current applications of Surgical Aspects, and Complications. World Neurosurg,
artificial intelligence for intraoperative decision support 2020,139:445-452
in surgery. Front Med, 2020,14(4):369-381 77 Vidal L, Kampleitner C, Brennan MA, et al.
61 Samareh A, Chang X, Lober WB, et al. Artificial Reconstruction of Large Skeletal Defects: Current
Intelligence Methods for Surgical Site Infection: Impacts Clinical Therapeutic Strategies and Future Directions
on Detection, Monitoring, and Decision Making. Surg Using 3D Printing. Front Bioeng Biotechnol, 2020,8:61
Infect (Larchmt), 2019,20(7):546-554 78 Xing F, Xiang Z, Rommens PM, et al. 3D Bioprinting
62 Tejo OA, Buj CI, Fenollosa AF. 3D Printing in Medicine for Vascularized Tissue-Engineered Bone Fabrication.
for Preoperative Surgical Planning: A Review. Ann Materials (Basel), 2020,13(10):2278
Biomed Eng, 2020,48(2):536-555 79 Rey F, Barzaghini B, Nardini A, et al. Advances in Tissue
63 Wang C, Zhang L, Qin T, et al. 3D printing in adult Engineering and Innovative Fabrication Techniques
cardiovascular surgery and interventions: a systematic for 3-D-Structures: Translational Applications in
review. J Thorac Dis, 2020,12(6):3227-3237 Neurodegenerative Diseases. Cells, 2020,9(7):1636
64 Nikoyan L, Patel R. Intraoral Scanner, Three- 80 Boso D, Maghin E, Carraro E, et al. Extracellular
Dimensional Imaging, and Three-Dimensional Matrix-Derived Hydrogels as Biomaterial for Different
Printing in the Dental Office. Dent Clin North Am, Skeletal Muscle Tissue Replacements. Materials
2020,64(2):365-378 (Basel), 2020,13(11):2483
65 Skelley NW, Smith MJ, Ma R, et al. Three-dimensional 81 Ettinger M, Windhagen H. Individual revision
Printing Technology in Orthopaedics. J Am Acad arthroplasty of the knee joint. Orthopade, 2020,49(5):
Orthop Surg, 2019,27(24):918-925 396-402
66 Yamaguchi JT, Hsu WK. Three-Dimensional Printing 82 Levin D, Mackensen GB, Reisman M, et al. 3D
1114 Current Medical Science 41(6):2021

Printing Applications for Transcatheter Aortic Valve Telementoring with Augmented Reality (STAR): A
Replacement. Curr Cardiol Rep, 2020,22(4):23 head-mounted display to improve surgical coaching and
83 Farmer ZL, Dominguez RJ, Mancinelli C, et al. confidence in remote areas. Surgery, 2020,167(4):724-
Urogynecological surgical mesh implants: New trends 731
in materials, manufacturing and therapeutic approaches. 100 Held J, Yu K, Pyles C, et al. Augmented Reality-Based
Int J Pharm, 2020,585:119512 Rehabilitation of Gait Impairments: Case Report. JMIR
84 Edgar L, Pu T, Porter B, et al. Regenerative medicine, Mhealth Uhealth, 2020,8(5):e17804
organ bioengineering and transplantation. Br J Surg, 101 Chen PJ, Penn IW, Wei SH, et al. Augmented reality-
2020,107(7):793-800 assisted training with selected Tai-Chi movements
85 Mirchi N, Bissonnette V, Ledwos N, et al. Artificial improves balance control and increases lower limb
Neural Networks to Assess Virtual Reality Anterior muscle strength in older adults: A prospective
Cervical Discectomy Performance. Oper Neurosurg randomized trial. J Exerc Sci Fit, 2020,18(3):142-147
(Hagerstown), 2020,19(1):65-75 102 Hashimoto DA, Witkowski E, Gao L, et al. Artificial
86 Sadeghi AH, Taverne Y, Bogers A, et al. Immersive Intelligence in Anesthesiology: Current Techniques,
virtual reality surgical planning of minimally invasive Clinical Applications, and Limitations. Anesthesiology,
coronary artery bypass for Kawasaki disease. Eur Heart 2020,132(2):379-394
J, 2020,41(34):3279 103 Seger C, Cannesson M. Recent advances in the
87 Fertleman C, Aubugeau WP, Sher C, et al. A Discussion technology of anesthesia. F1000Res, 2020,9:F1000
of Virtual Reality As a New Tool for Training Healthcare Faculty Rev-375
Professionals. Front Public Health, 2018,6:44 104 Kamdar N, Jalilian L. Telemedicine: A Digital Interface
88 Creighton FX, Unberath M, Song T, et al. Early for Perioperative Anesthetic Care. Anesth Analg, 2020,
Feasibility Studies of Augmented Reality Navigation 130(2):272-275
for Lateral Skull Base Surgery. Otol Neurotol, 105 Poncette AS, Mosch L, Spies C, et al. Improvements in
2020,41(7):883-888 Patient Monitoring in the Intensive Care Unit: Survey
89 Gibby J, Cvetko S, Javan R, et al. Use of augmented Study. J Med Internet Res, 2020,22(6):e19091
reality for image-guided spine procedures. Eur Spine J, 106 Angehrn Z, Haldna L, Zandvliet AS, et al. Artificial
2020,29(8):1823-1832 Intelligence and Machine Learning Applied at the Point
90 Hu HZ, Feng XB, Shao ZW, et al. Application and of Care. Front Pharmacol, 2020,11:759
Prospect of Mixed Reality Technology in Medical Field. 107 Dai B, Yu Y, Huang L, et al. Application of neural
Curr Med Sci, 2019,39(1):1-6 network model in assisting device fitting for low vision
91 Goo HW, Park SJ, Yoo SJ. Advanced Medical Use patients. Ann Transl Med, 2020,8(11):702
of Three-Dimensional Imaging in Congenital Heart 108 Averta G, Della SC, Valenza G, et al. Exploiting upper-
Disease: Augmented Reality, Mixed Reality, Virtual limb functional principal components for human-
Reality, and Three-Dimensional Printing. Korean J like motion generation of anthropomorphic robots. J
Radiol, 2020,21(2):133-145 Neuroeng Rehabil, 2020,17(1):63
92 Salmas M, Chronopoulos E, Chytas D. Comment on: “A 109 Zhao Y, Liang C, Gu Z, et al. A New Design Scheme for
Novel Evaluation Model for a Mixed-Reality Surgical Intelligent Upper Limb Rehabilitation Training Robot.
Navigation System: Where Microsoft HoloLens Meets Int J Environ Res Public Health, 2020,17(8):2948
the Operating Room”. Surg Innov, 2020:1611082121 110 De CH, Corradi F, Smeets C, et al. Wearable Monitoring
93 Wu X, Liu R, Yu J, et al. Mixed Reality Technology and Interpretable Machine Learning Can Objectively
Launches in Orthopedic Surgery for Comprehensive Track Progression in Patients during Cardiac
Preoperative Management of Complicated Cervical Rehabilitation. Sensors (Basel), 2020,20(12):3601
Fractures. Surg Innov, 2018,25(4):421-422 111 Ramezani R, Zhang W, Xie Z, et al. A Combination
94 Gu Y, Yao Q, Xu Y, et al. A Clinical Application Study of Indoor Localization and Wearable Sensor-Based
of Mixed Reality Technology Assisted Lumbar Pedicle Physical Activity Recognition to Assess Older Patients
Screws Implantation. Med Sci Monit, 2020,26:e924982 Undergoing Subacute Rehabilitation: Baseline Study
95 Chytas D, Chronopoulos E, Salmas M, et al. Comment Results. JMIR Mhealth Uhealth, 2019,7(7):e14090.
on: “Intraoperative 3D Hologram Support With Mixed 112 Bajorath J, Kearnes S, Walters WP, et al. Artificial
Reality Techniques in Liver Surgery”. Ann Surg, Intelligence in Drug Discovery: Into the Great Wide
2021,274(6):e761-e762 Open. J Med Chem, 2020,63(16):8651-8652
96 Zeiger J, Costa A, Bederson J, et al. Use of Mixed Reality 113 Brown N, Ertl P, Lewis R, et al. Artificial intelligence in
Visualization in Endoscopic Endonasal Skull Base chemistry and drug design. J Comput Aided Mol Des,
Surgery. Oper Neurosurg (Hagerstown), 2020,19(1):43- 2020,34(7):709-715
52 114 Zhavoronkov A. Medicinal Chemists versus Machines
97 Wu X, Liu R, Yu J, et al. Mixed Reality Technology- Challenge: What Will It Take to Adopt and Advance
Assisted Orthopedics Surgery Navigation. Surg Innov, Artificial Intelligence for Drug Discovery? J Chem Inf
2018,25(3):304-305 Model, 2020,60(6):2657-2659
98 Yoshida S, Sugimoto M, Fukuda S, et al. Mixed reality 115 Russo G, Reche P, Pennisi M, et al. The combination of
computed tomography-based surgical planning for artificial intelligence and systems biology for intelligent
partial nephrectomy using a head-mounted holographic vaccine design. Expert Opin Drug Discov, 2020:1-15
computer. Int J Urol, 2020, 26(6):681-682 116 Fernandez A. Artificial Intelligence Teaches Drugs
99 Rojas ME, Cabrera ME, Lin C, et al. The System for to Target Proteins by Tackling the Induced Folding
Current Medical Science 41(6):2021 1115
Problem. Mol Pharm, 2020,17(8):2761-2767 130 Bertin H, Huon JF, Praud M, et al. Bilateral sagittal
117 Liang G, Fan W, Luo H, et al. The emerging roles of split osteotomy training on mandibular 3-dimensional
artificial intelligence in cancer drug development and printed models for maxillofacial surgical residents. Br J
precision therapy. Biomed Pharmacother, 2020,128: Oral Maxillofac Surg, 2020,58(8):953-958
110255 131 Bohl MA, McBryan S, Pais D, et al. The Living Spine
118 Takakusagi Y, Takakusagi K, Sakaguchi K, et al. Phage Model: A Biomimetic Surgical Training and Education
display technology for target determination of small- Tool. Oper Neurosurg (Hagerstown), 2020,19(1):98-
molecule therapeutics: an update. Expert Opin Drug 106
Discov, 2020:1-13 132 Sappenfield JW, Smith WB, Cooper LA, et al.
119 Awad A, Fina F, Goyanes A, et al. 3D printing: Visualization Improves Supraclavicular Access to the
Principles and pharmaceutical applications of selective Subclavian Vein in a Mixed Reality Simulator. Anesth
laser sintering. Int J Pharm, 2020,586:119594 Analg, 2018,127(1):83-89
120 Pandey M, Choudhury H, Fern J, et al. 3D printing 133 Vaishya R, Javaid M, Khan IH, et al. Artificial
for oral drug delivery: a new tool to customize drug Intelligence (AI) applications for COVID-19 pandemic.
delivery. Drug Deliv Transl Res, 2020,10(4):986-1001 Diabetes Metab Syndr, 2020,14(4):337-339
121 Tsintavi E, Rekkas DM, Bettini R. Partial tablet coating 134 Zhang HT, Zhang JS, Zhang HH, et al. Automated
by 3D printing. Int J Pharm, 2020,581:119298 detection and quantification of COVID-19 pneumonia:
122 Cheng N, Kuo A. Using Long Short-Term Memory CT imaging analysis by a deep learning-based software.
(LSTM) Neural Networks to Predict Emergency Eur J Nucl Med Mol Imaging, 2020,47(11):2525-2532
Department Wait Time. Stud Health Technol Inform, 135 Sakagianni A, Feretzakis G, Kalles D, et al. Setting
2020,272:199-202 up an Easy-to-Use Machine Learning Pipeline for
123 Nas S, Koyuncu M. Emergency Department Capacity Medical Decision Support: A Case Study for COVID-19
Planning: A Recurrent Neural Network and Simulation Diagnosis Based on Deep Learning with CT Scans. Stud
Approach. Comput Math Methods Med, 2019,2019: Health Technol Inform, 2020,272:13-16
4359719 136 Mashamba TP, Drain PK. Point-of-Care Diagnostic
124 Saab A, Saikali M, Lamy JB. Comparison of Machine Services as an Integral Part of Health Services during
Learning Algorithms for Classifying Adverse-Event the Novel Coronavirus 2019 Era. Diagnostics (Basel),
Related 30-Day Hospital Readmissions: Potential 2020,10(7):449
Implications for Patient Safety. Stud Health Technol 137 Mohanty S, Harun AR, Mridul M, et al. Application of
Inform, 2020,272:51-54 Artificial Intelligence in COVID-19 drug repurposing.
125 Lin YW, Zhou Y, Faghri F, et al. Analysis and prediction Diabetes Metab Syndr, 2020,14(5):1027-1031
of unplanned intensive care unit readmission using 138 Ke YY, Peng TT, Yeh TK, et al. Artificial intelligence
recurrent neural networks with long short-term memory. approach fighting COVID-19 with repurposing drugs.
PLoS One, 2019,14(7):e218942 Biomed J, 2020,43(4):355-362
126 Wu D, Xiang Y, Wu X, et al. Artificial intelligence- 139 Kim J, Zhang J, Cha Y, et al. Advanced bioinformatics
tutoring problem-based learning in ophthalmology rapidly identifies existing therapeutics for patients with
clerkship. Ann Transl Med, 2020,8(11):700 coronavirus disease-2019 (COVID-19). J Transl Med,
127 Yang YY, Shulruf B. Expert-led and artificial intelligence 2020,18(1):257
(AI) system-assisted tutoring course increase confidence 140 Arash KA, Julia W, Milad S, et al. Artificial Intelligence
of Chinese medical interns on suturing and ligature for COVID-19 Drug Discovery and Vaccine Deve-
skills: prospective pilot study. J Educ Eval Health Prof, lopment. Front Artif Intell, 2020,3:65
2019,16:7 141 Elaziz MA, Hosny KM, Salah A, et al. New machine
128 Mirchi N, Bissonnette V, Yilmaz R, et al. The learning method for image-based diagnosis of
Virtual Operative Assistant: An explainable artificial COVID-19. PLoS One, 2020,15(6):e235187
intelligence tool for simulation-based training in surgery 142 Mali SN, Pratap AP. Targeting infectious Coronavirus
and medicine. PLoS One, 2020,15(2):e229596 Disease 2019 (COVID-19) with Artificial Intelligence
129 Dekker I, De Jong EM, Schippers MC, et al. Optimizing (AI) applications: Evidence based opinion. Infect
Students' Mental Health and Academic Performance: AI- Disord Drug Targets, 2021,21(4):475-477
Enhanced Life Crafting. Front Psychol, 2020,11:1063 (Received Sep. 4, 2020; accepted Dec. 1, 2020)