ISO/TR 24291:2021
(Main)Health informatics — Applications of machine learning technologies in imaging and other medical applications
Health informatics — Applications of machine learning technologies in imaging and other medical applications
This document lists examples of and defines categories of use cases for machine learning in medicine for clinical practice. The developments and applications of machine learning technologies for artificial intelligence consist of 1) data collection and curation, 2) pre-processing, 3) model training and validation, and 4) medicine depending on various kinds of specialty including radiology, pathology, emergency medicine, dermatology, ophthalmology, anaesthesia, surgery, etc., and clinical settings including repeated detection and/or diagnosis, real-time monitoring, and treatment prediction. This document covers categories applications of medicine in (4). It also defines the clinical usages and necessities of the artificial intelligence in medicine. (1) to (3) are not the scope of this document This document also excludes — basic research and other scientific areas, — use cases related to artificial intelligence methods other than machine learning (for example, symbolic artificial intelligence, expert systems), and — non-human results such as veterinary medicine.
Informatique de santé — Applications de technologies d'apprentissage automatique en imagerie et autres applications médicales
General Information
Buy Standard
Standards Content (Sample)
TECHNICAL ISO/TR
REPORT 24291
First edition
2021-03
Health informatics — Applications
of machine learning technologies
in imaging and other medical
applications
Informatique de santé — Applications de technologies
d'apprentissage automatique en imagerie et autres applications
médicales
Reference number
ISO/TR 24291:2021(E)
©
ISO 2021
---------------------- Page: 1 ----------------------
ISO/TR 24291:2021(E)
COPYRIGHT PROTECTED DOCUMENT
© ISO 2021
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting
on the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address
below or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 22 749 01 11
Email: copyright@iso.org
Website: www.iso.org
Published in Switzerland
ii © ISO 2021 – All rights reserved
---------------------- Page: 2 ----------------------
ISO/TR 24291:2021(E)
Contents Page
Foreword .iv
Introduction .v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Abbreviated terms . 4
5 Categories for defining use cases of machine learning in medicine .4
5.1 Categories based on technology . 4
5.1.1 General. 4
5.1.2 Robotics . 5
5.1.3 Continuous monitoring . 5
5.1.4 Machine learning . 5
5.1.5 Deep learning . 5
5.1.6 Image processing . 5
5.1.7 Natural language processing . 5
5.1.8 Audio recognition . 6
5.1.9 Bigdata analysis . 6
5.1.10 Prediction modeling . . 6
5.2 Categories based on medical specialty . 6
5.2.1 General. 6
5.2.2 Radiology and Pathology . 6
5.2.3 Dermatology . 6
5.2.4 Ophthalmology . 6
5.2.5 Internal Medicine . 7
5.2.6 Cardiology . 7
5.2.7 Neurology, Urology, Surgery . 7
5.2.8 Anaesthesiologist, Intensive Care Unit . 7
5.2.9 Emergency . 7
5.3 Categories based on medical usage . 7
5.3.1 General. 7
5.3.2 Clinical trials . 8
5.3.3 Clinical assistance . 8
5.3.4 Data-based precision medicine . 8
5.3.5 Medical Imaging and Diagnostic . 8
5.3.6 Hospital Management . 9
5.3.7 Robot surgery . 9
5.3.8 Drug development . 9
6 Use cases of artificial intelligence in medicine .10
6.1 General .10
6.2 AI Platform for Lung Cancer Screening and Reporting .10
6.3 AI based text to speech services with personal voices for speech impaired people .11
6.4 AI Platform for Chest CT-Scan Analysis .11
6.5 Support system for optimization and personification of drug therapy .11
6.6 WebioMed Clinical Decision Support System .12
Bibliography .14
© ISO 2021 – All rights reserved iii
---------------------- Page: 3 ----------------------
ISO/TR 24291:2021(E)
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out
through ISO technical committees. Each member body interested in a subject for which a technical
committee has been established has the right to be represented on that committee. International
organizations, governmental and non-governmental, in liaison with ISO, also take part in the work.
ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of
electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are
described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the
different types of ISO documents should be noted. This document was drafted in accordance with the
editorial rules of the ISO/IEC Directives, Part 2 (see www .iso .org/ directives).
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights. Details of
any patent rights identified during the development of the document will be in the Introduction and/or
on the ISO list of patent declarations received (see www .iso .org/ patents).
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and
expressions related to conformity assessment, as well as information about ISO's adherence to the
World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www .iso .org/
iso/ foreword .html.
This document was prepared by Technical Committee ISO/TC 215, Health informatics.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www .iso .org/ members .html.
iv © ISO 2021 – All rights reserved
---------------------- Page: 4 ----------------------
ISO/TR 24291:2021(E)
Introduction
Due to the rapid advancement of artificial intelligence, especially machine learning and deep learning,
defining categories of use cases in the clinical setting have started to be adopted to enhance healthcare
system and patients’ outputs. Therefore, it is crucial to define the categories of use cases for artificial
intelligence in the clinical setting to focus on application of artificial intelligence in medicine.
This document proposes categories of use cases of machine learning technologies for artificial
intelligence in medicine considering the property of artificial intelligence technology including
machine learning and deep learning and clinical settings especially requiring repeated detection and/
or diagnosis, real-time monitoring, and treatment prediction with images and continuous signals, etc.
This document will assist the health IT companies by reviewing the current status of machine learning
technologies for artificial intelligence in medicine and then by proposing a gap for a new application.
This document can be used to further develop the applications or the necessary standards of machine
learning technologies for artificial intelligence in medicine.
© ISO 2021 – All rights reserved v
---------------------- Page: 5 ----------------------
TECHNICAL REPORT ISO/TR 24291:2021(E)
Health informatics — Applications of machine learning
technologies in imaging and other medical applications
1 Scope
This document lists examples of and defines categories of use cases for machine learning in medicine
for clinical practice.
The developments and applications of machine learning technologies for artificial intelligence
consist of 1) data collection and curation, 2) pre-processing, 3) model training and validation, and 4)
medicine depending on various kinds of specialty including radiology, pathology, emergency medicine,
dermatology, ophthalmology, anaesthesia, surgery, etc., and clinical settings including repeated
detection and/or diagnosis, real-time monitoring, and treatment prediction.
This document covers categories applications of medicine in (4). It also defines the clinical usages and
necessities of the artificial intelligence in medicine.
(1) to (3) are not the scope of this document
This document also excludes
— basic research and other scientific areas,
— use cases related to artificial intelligence methods other than machine learning (for example,
symbolic artificial intelligence, expert systems), and
— non-human results such as veterinary medicine.
2 Normative references
There are no normative references in this document.
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
ISO and IEC maintain terminological databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https:// www .iso .org/ obp
— IEC Electropedia: available at http:// www .electropedia .org/
3.1
artificial intelligence
AI
branch of computer science devoted to developing data processing systems that perform functions
normally associated with human intelligence, such as reasoning, learning, and self-improvement
[SOURCE: ISO/IEC/IEEE 24765:2017, 3.234]
© ISO 2021 – All rights reserved 1
---------------------- Page: 6 ----------------------
ISO/TR 24291:2021(E)
3.2
big data
extensive datasets — primarily in the data characteristics of volume, variety, velocity, and/or
variability — that require a scalable technology for efficient storage, manipulation, management, and
analysis
Note 1 to entry: Big data is commonly used in many different ways, for example as the name of the scalable
technology used to handle big data extensive datasets.
[SOURCE: ISO/IEC 20546:2019, 3.1.2]
3.3
electronic medical records
EMR
electronic record derived from a computerized system used primarily for delivering patient care in a
clinical setting
3.4
clinical decision support
CDS
type of service that assists healthcare providers in making medical decisions, which typically requires
input of patient-specific clinical variables and provide patient-specific recommendations
[SOURCE: ISO/TS 22756:2020, 3.1]
3.5
clinical decision support system
CDSS
software designed to be a direct aid to clinical decision-making, in which the characteristics of an
individual patient are matched to a computerized clinical knowledge base, whereafter patient-specific
assessments or recommendations are presented to the clinician or the patient to aid in the process of
making evidence based clinical decisions
[SOURCE: ISO/TS 22756:2020, 3.2]
3.6
computer aided detection
CADe
health information technology system to provide physicians and other health professionals with
automated detection in medical records (i.e., images), that is, assistance with clinical diagnosis tasks
3.7
computer aided diagnosis
CADx
health information technology system to provide physicians and other health professionals with
automated diagnosis by using medical records including images, and EMR, that is, assistance with
clinical diagnosis tasks
3.8
computer aided differential diagnosis
CADD
health information technology system to provide physicians and other health professionals with
automated differential diagnosis by using medical records including images, and EMR
3.9
computed tomography
CT
radiographic scanning technique that uses a number of CT projections of an object at different angles in
order to allow calculation of a CT image
[SOURCE: ISO 15708-1:2017, 3.7]
2 © ISO 2021 – All rights reserved
---------------------- Page: 7 ----------------------
ISO/TR 24291:2021(E)
3.10
deep learning
approach to creating rich hierarchical representations through the training of neural networks (3.16)
with many hidden layers
Note 1 to entry: Deep learning is also known as deep neural network learning
[SOURCE: ISO/IEC TR 29119-11:2020, 3.1.26, modified — Note 1 has been modified.]
3.11
image processing
process of applying any operation to a pictorial representation of objects or data
for a given purpose
Note 1 to entry: Examples of operations include scene analysis, image compression, image restoration, image
enhancement, preprocessing, quantizing, spatial filtering, and construction of two- and three-dimensional
models of objects.
[SOURCE: ISO/IEC 2382:2015, 2125939, modified — Admitted term and Note 3 to entry deleted.]
3.12
machine learning
process using computational techniques to enable systems to learn from data or experience
[SOURCE: ISO/IEC TR 29119-11:2020, 3.1.43]
3.13
magnetic resonance imaging
MRI
imaging technique that uses static and time varying magnetic fields to provide images of tissue by the
magnetic resonance of nuclei
[SOURCE: ISO 14630:2012, 3.5]
3.14
natural language processing
NLP
technology used to determine and identify key words and phrases within processing audio data (e.g.
call centres) and free-form text (e.g., the body of an email)
Note 1 to entry: This technology is able to reduce words to their base constructs and perform other actions, such
as stemming, along with locating similar words or phrases without user intervention. This technology also varies
greatly from standard IDR technology due to the ability to automatically update rules as determined by the users
without the need for technical intervention. This technology is best suited for unstructured documents.
[SOURCE: ISO/TR 22957:2018, 3.7]
3.15
artificial neural network
neural network
neural net
ANN
network of primitive processing elements connected by weighted links with adjustable weights, in
which each element produces a value by applying a nonlinear function to its input values, and transmits
it to other elements or presents it as an output value
Note 1 to entry: Whereas some neural networks are intended to simulate the functioning of neurons in the nervous
system, most neural networks are used in artificial intelligence as realizations of the connectionist model.
Note 2 to entry: Examples of nonlinear functions are a threshold function, a sigmoid function, and a polynomial
function.
[SOURCE: ISO/IEC 2382:2015, 2120625, modified — Notes to entry 3 to 5 deleted.]
© ISO 2021 – All rights reserved 3
---------------------- Page: 8 ----------------------
ISO/TR 24291:2021(E)
3.16
prediction
output of an algorithm after it has been trained on a historical dataset and applied to new data when
forecasting the likelihood of a particular outcome
3.17
robotics
techniques involved in designing, building, and using robots
[SOURCE: ISO/IEC/IEEE 24765:2017, 3.3554]
3.18
speech recognition
automatic speech recognition
conversion, by a functional unit, of a speech signal to a representation of the content of the speech
Note 1 to entry: The content to be recognized can be expressed as a proper sequence of words or phonemes.
[SOURCE: ISO/IEC 19794-13:2018, 3.22]
4 Abbreviated terms
CBIR Content-based case retrieval
CNN Convolutional Neural Net
DB Database
EMR Electronic Medical Records
ICU Intensive Care Unit
IDR Intelligent Document Recognition
IoT Internet of Things
OR Operation Room
5 Categories for defining use cases of machine learning in medicine
5.1 Categories based on technology
5.1.1 General
AI techniques used in medicine can be summarized as in Table 1.
4 © ISO 2021 – All rights reserved
---------------------- Page: 9 ----------------------
ISO/TR 24291:2021(E)
Table 1 — Technology based categories of artificial intelligence and their purposes
Technology Purpose
Providing high quality treatments by increasing the precision and accuracy
Robotics
of the surgical process.
Within golden time, proper treatment could be performed by continuously
Continuous monitoring
monitoring of patient condition and alerting nurses.
Machine learning Predict response by analyzing data affecting treatment outcomes.
Self-learning ability to process large amounts of medical imaging records,
Deep learning
reducing uncertainty in medical treatment decisions.
Process large-scale medical images and apply them to detect diseases, diag-
Image processing
nosis, etc.
Translate long descriptive character sets such as electric medical records to
Natural language processing
be interpreted.
By recognizing voice and language of patient, dictate important information
Audio recognition
in electric medical records.
Process vast patient health records held by healthcare organizations and
Bigdata analysis
provide tailored recommendations to patients and providers.
Prediction modeling Apply AI models to predict outcomes such as predicting risk disorders.
5.1.2 Robotics
In robotics, AI can provide high quality treatments by increasing the precision and accuracy of the
surgical process. For example, it can control the trajectory, depth, and speed of the robot movements
with high precision and can go where traditional tools cannot. It can also reduce the burdens of the
surgeons during surgery by providing the same, repetitive movements without fatigue.
5.1.3 Continuous monitoring
Proper treatment within golden time could be performed by continuously monitoring of patient
condition and alerting nurses by AI. AI model with continuous monitoring data also can alert the
clinicians before onset.
5.1.4 Machine learning
By using traditional machine learning methods, AI can be used to predict response by analysing data
affecting treatment outcomes.
5.1.5 Deep learning
By using deep learning, self-learning ability to process large amounts of imaging and audios records
in medicine, reducing uncertainty in medical treatment decisions including computer aided detection,
computer aided diagnosis, computer aided differential diagnosis, and clinical decision support system.
Deep learning can handle multiple different types of clinical data such as images, texts, and signals at
the same time.
5.1.6 Image processing
In image processing, AI can be used to process large-scale medical images and apply them to detect
diseases, diagnosis, etc. AI for clinical image handling has demonstrated its performance in clinical
settings.
5.1.7 Natural language processing
In NLP, AI can be used to translate long descriptive character sets such as electric medical records to be
interpreted, i.e. extracting the information from unstructured electronic medical records.
© ISO 2021 – All rights reserved 5
---------------------- Page: 10 ----------------------
ISO/TR 24291:2021(E)
5.1.8 Audio recognition
In audio recognition, by recognizing voice and language of patient, AI can automatically dictate
important information in electronic medical records.
5.1.9 Bigdata analysis
In bigdata analysis, AI can be used to process vast patient health records held by healthcare
organizations and provide tailored recommendations to patients and providers.
5.1.10 Prediction modeling
In prediction modeling, AI models can be applied to predict outcomes such as predicting risk disorders.
5.2 Categories based on medical specialty
5.2.1 General
The applications of AI in medicine can be categorized by medical specialty as in Table 2.
Table 2 — Typical use cases of artificial intelligence in medicine in each medical specialty
Specialty Use Cases
Quantification, computer aided detection, diagnosis, and differential diagnosis on radio-
Radiology
logic images; Automated dictation system.
Quantification, computer aided detection, diagnosis, and differential diagnosis on patho-
Pathology
logic images; Automated dictation system.
Dermatology Skin cancer detection and classification.
Ophthalmology Eye disease detection and classification in retina image and OCT.
Internal Medicine Complication/outcome prediction; clinical decision support system.
Quantification, computer aided detection, diagnosis, and differential diagnosis on cardi-
Cardiology
ologic images; Complication/outcome prediction; clinical decision support system.
Neurology, Urology,
Complication/outcome prediction; clinical decision support system.
Surgery
Anaesthesiologist,
Intensive Care Unit Continuous monitoring system; Complication/outcome prediction.
(ICU)
Emergency Triage system; Continuous monitoring system; Complication/outcome prediction.
5.2.2 Radiology and Pathology
In Radiology and Pathology, AI can be used for fully automated quantification, computer aided
detection, diagnosis, and differential diagnosis on radiologic or pathologic images and automated
dictation system.
5.2.3 Dermatology
In Dermatology, skin cancer detection and classification could be one of typical use cases.
5.2.4 Ophthalmology
In Ophthalmology, AI can be used for eye disease detection and classification in retina image and OCT.
6 © ISO 2021 – All rights reserved
---------------------- Page: 11 ----------------------
ISO/TR 24291:2021(E)
5.2.5 Internal Medicine
In Internal Medicine, complication and outcome prediction, and clinical decision support system could
be one of important use cases.
5.2.6 Cardiology
In Cardiology, quantification, computer aided detection, diagnosis, and differential diagnosis on
cardiologic images, complication and outcome prediction, and clinical decision support system could be
one of use cases.
5.2.7 Neurology, Urology, Surgery
In Neurology, Urology, Surgery, AI can be used for complication and outcome prediction, and clinical
decision support system.
5.2.8 Anaesthesiologist, Intensive Care Unit
In Anaesthesiologist, Intensive Care Unit (ICU), continuous monitoring system and complication and
outcome prediction could be one of important applications.
5.2.9 Emergency
In Emergency, automated triages system, continuous monitoring system, complication and outcome
prediction could be one of important applications.
5.3 Categories based on medical usage
5.3.1 General
The examples of AI in medicine can be categorized by their medical usages including clinical trials,
clinical assistance, data-based precision medicine, diagnostic imaging, logistics in ORs and wards,
robot surgery, and drug development in medicine (see Table 3 and Figure 1).
Table 3 — Categories of use cases of medical usages
Usage Sub-usage Purpose
Case selection AI-based search techniques help find the right disease and patient,
Clinical trials
reduce the time to prepare for clinical trials, and improve objectivity.
Assistant service Convergence of IoT technology, voice recognition technology and
and wellness with artificial intelligence technology, efficient reservation, diagnosis and
IoT medical treatment process, business information update and custom-
Clinical Assis-
ized curation.
tance
Automated dictation Speech recognition and document generation technology that can
automate diagnosis and recording of readings, and read and structure
medical terminology.
Genome and genet- Prediction, diagnosis and treatment by analysing and modeling asso-
ics ciations of big data based on genome, multimodal medical imaging and
clinical pathology for personalized medicine.
Prediction of
Data based pre- Assist physicians in making final decisions by notifying them of case-
complication and
cision medicine based risks or complications in treatment and drug use.
mixed drug
Recommendation of Recommend additional diagnostic testing processes to increase ac-
exam curacy and lower risk, utilizing artificial intelligence and big data for
precise diagnosis and treatment.
© ISO 2021 – All rights reserved 7
---------------------- Page: 12 ----------------------
ISO/TR 24291:2021(E)
Table 3 (continued)
Usage Sub-usage Purpose
Normal selection Early diagnose medical images in advance to reduce clinical burden by
differentiating the true normal cases.
Content based case Helps diagnose by searching and visualizing similar cases from numer-
retrieval ous cases in DB.
Medical Imaging
Generating reading Artificial intelligence-based medical image analysis and natural lan-
and Diagnostic
reports guage processing technology are integrated to generate readings that
can assist the reading of image specialists.
Computer aided Application of AI techniques to diagnostic image big data, pathogenesis
diagnosis analysis, prognostic prediction.
- Efficient operation of logistics, opera
...
TECHNICAL ISO/TR
REPORT 24291
First edition
Health informatics - Applications
of machine learning technologies
in imaging and other medical
applications
Informatique de santé — Applications de technologies
d'apprentissage automatique en imagerie et autres applications
médicales
PROOF/ÉPREUVE
Reference number
ISO/TR 24291:2021(E)
©
ISO 2021
---------------------- Page: 1 ----------------------
ISO/TR 24291:2021(E)
COPYRIGHT PROTECTED DOCUMENT
© ISO 2021
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting
on the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address
below or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 22 749 01 11
Email: copyright@iso.org
Website: www.iso.org
Published in Switzerland
ii PROOF/ÉPREUVE © ISO 2021 – All rights reserved
---------------------- Page: 2 ----------------------
ISO/TR 24291:2021(E)
Contents Page
Foreword .iv
Introduction .v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Abbreviated terms . 4
5 Categories for defining use cases of machine learning in medicine .4
5.1 Categories based on technology . 4
5.1.1 General. 4
5.1.2 Robotics . 5
5.1.3 Continuous monitoring . 5
5.1.4 Machine learning . 5
5.1.5 Deep learning . 5
5.1.6 Image processing . 5
5.1.7 Natural language processing . 5
5.1.8 Audio recognition . 5
5.1.9 Bigdata analysis . 6
5.1.10 Prediction modeling . . 6
5.2 Categories based on medical specialty . 6
5.2.1 General. 6
5.2.2 Radiology and Pathology . 6
5.2.3 Dermatology . 6
5.2.4 Ophthalmology . 6
5.2.5 Internal Medicine . 6
5.2.6 Cardiology . 7
5.2.7 Neurology, Urology, Surgery . 7
5.2.8 Anaesthesiologist, Intensive Care Unit . 7
5.2.9 Emergency . 7
5.3 Categories based on medical usage . 7
5.3.1 General. 7
5.3.2 Clinical trials . 8
5.3.3 Clinical assistance . 8
5.3.4 Data-based precision medicine . 8
5.3.5 Medical Imaging and Diagnostic . 8
5.3.6 Hospital Management . 9
5.3.7 Robot surgery . 9
5.3.8 Drug development . 9
6 Use cases of artificial intelligence in medicine .10
6.1 General .10
6.2 AI Platform for Lung Cancer Screening and Reporting .10
6.3 AI based text to speech services with personal voices for speech impaired people .11
6.4 AI Platform for Chest CT-Scan Analysis .11
6.5 Support system for optimization and personification of drug therapy .11
6.6 WebioMed Clinical Decision Support System .12
Bibliography .14
© ISO 2021 – All rights reserved PROOF/ÉPREUVE iii
---------------------- Page: 3 ----------------------
ISO/TR 24291:2021(E)
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out
through ISO technical committees. Each member body interested in a subject for which a technical
committee has been established has the right to be represented on that committee. International
organizations, governmental and non-governmental, in liaison with ISO, also take part in the work.
ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of
electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are
described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the
different types of ISO documents should be noted. This document was drafted in accordance with the
editorial rules of the ISO/IEC Directives, Part 2 (see www .iso .org/ directives).
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights. Details of
any patent rights identified during the development of the document will be in the Introduction and/or
on the ISO list of patent declarations received (see www .iso .org/ patents).
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and
expressions related to conformity assessment, as well as information about ISO's adherence to the
World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www .iso .org/
iso/ foreword .html.
This document was prepared by Technical Committee ISO/TC 215, Health informatics.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www .iso .org/ members .html.
iv PROOF/ÉPREUVE © ISO 2021 – All rights reserved
---------------------- Page: 4 ----------------------
ISO/TR 24291:2021(E)
Introduction
Due to the rapid advancement of artificial intelligence, especially machine learning and deep learning,
defining categories of use cases in the clinical setting have started to be adopted to enhance healthcare
system and patients’ outputs. Therefore, it is crucial to define the categories of use cases for artificial
intelligence in the clinical setting to focus on application of artificial intelligence in medicine.
This document proposes categories of use cases of machine learning technologies for artificial
intelligence in medicine considering the property of artificial intelligence technology including
machine learning and deep learning and clinical settings especially requiring repeated detection and/
or diagnosis, real-time monitoring, and treatment prediction with images and continuous signals, etc.
This document will assist the health IT companies by reviewing the current status of machine learning
technologies for artificial intelligence in medicine and then by proposing a gap for a new application.
This document can be used to further develop the applications or the necessary standards of machine
learning technologies for artificial intelligence in medicine.
© ISO 2021 – All rights reserved PROOF/ÉPREUVE v
---------------------- Page: 5 ----------------------
TECHNICAL REPORT ISO/TR 24291:2021(E)
Health informatics - Applications of machine learning
technologies in imaging and other medical applications
1 Scope
This document lists examples of and defines categories of use cases for machine learning in medicine
for clinical practice.
The developments and applications of machine learning technologies for artificial intelligence
consist of 1) data collection and curation, 2) pre-processing, 3) model training and validation, and 4)
medicine depending on various kinds of specialty including radiology, pathology, emergency medicine,
dermatology, ophthalmology, anaesthesia, surgery, etc., and clinical settings including repeated
detection and/or diagnosis, real-time monitoring, and treatment prediction.
This document covers categories applications of medicine in (4). It also defines the clinical usages and
necessities of the artificial intelligence in medicine.
(1) to (3) are not the scope of this document
This document also excludes
— basic research and other scientific areas,
— use cases related to artificial intelligence methods other than machine learning (for example,
symbolic artificial intelligence, expert systems), and
— non-human results such as veterinary medicine.
2 Normative references
There are no normative references in this document.
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
ISO and IEC maintain terminological databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https:// www .iso .org/ obp
— IEC Electropedia: available at http:// www .electropedia .org/
3.1
artificial intelligence
AI
branch of computer science devoted to developing data processing systems that perform functions
normally associated with human intelligence, such as reasoning, learning, and self-improvement
[SOURCE: ISO/IEC/IEEE 24765:2017, 3.234]
© ISO 2021 – All rights reserved PROOF/ÉPREUVE 1
---------------------- Page: 6 ----------------------
ISO/TR 24291:2021(E)
3.2
big data
extensive datasets — primarily in the data characteristics of volume, variety, velocity, and/or
variability — that require a scalable technology for efficient storage, manipulation, management, and
analysis
Note 1 to entry: Big data is commonly used in many different ways, for example as the name of the scalable
technology used to handle big data extensive datasets.
[SOURCE: ISO/IEC 20546:2019, 3.1.2]
3.3
electronic medical records
EMR
electronic record derived from a computerized system used primarily for delivering patient care in a
clinical setting
3.4
clinical decision support
CDS
type of service that assists healthcare providers in making medical decisions, which typically requires
input of patient-specific clinical variables and provide patient-specific recommendations
[SOURCE: ISO/TS 22756:2020, 3.1]
3.5
clinical decision support system
CDSS
software designed to be a direct aid to clinical decision-making, in which the characteristics of an
individual patient are matched to a computerized clinical knowledge base, whereafter patient-specific
assessments or recommendations are presented to the clinician or the patient to aid in the process of
making evidence based clinical decisions
[SOURCE: ISO/TS 22756:2020, 3.2]
3.6
computer aided detection
CADe
health information technology system to provide physicians and other health professionals with
automated detection in medical records (i.e., images), that is, assistance with clinical diagnosis tasks
3.7
computer aided diagnosis
CADx
health information technology system to provide physicians and other health professionals with
automated diagnosis by using medical records including images, and EMR, that is, assistance with
clinical diagnosis tasks
3.8
computer aided differential diagnosis
CADD
health information technology system to provide physicians and other health professionals with
automated differential diagnosis by using medical records including images, and EMR
3.9
computed tomography
CT
radiographic scanning technique that uses a number of CT projections of an object at different angles in
order to allow calculation of a CT image
[SOURCE: ISO 15708-1:2017, 3.7]
2 PROOF/ÉPREUVE © ISO 2021 – All rights reserved
---------------------- Page: 7 ----------------------
ISO/TR 24291:2021(E)
3.10
deep learning
approach to creating rich hierarchical representations through the training of neural networks (3.16)
with many hidden layers
Note 1 to entry: Deep learning is also known as deep neural network learning
[SOURCE: ISO/IEC TR 29119-11:2020, 3.1.26, modified — Note 1 has been modified.]
3.12
image processing
process of applying any operation to a pictorial representation of objects or data
for a given purpose
Note 1 to entry: Examples of operations include scene analysis, image compression, image restoration, image
enhancement, preprocessing, quantizing, spatial filtering, and construction of two- and three-dimensional
models of objects.
[SOURCE: ISO/IEC 2382:2015, 2125939, modified — Admitted term and Note 3 to entry deleted.]
3.13
machine learning
process using computational techniques to enable systems to learn from data or experience
[SOURCE: ISO/IEC TR 29119-11:2020, 3.1.43]
3.14
magnetic resonance imaging
MRI
imaging technique that uses static and time varying magnetic fields to provide images of tissue by the
magnetic resonance of nuclei
[SOURCE: ISO 14630:2012, 3.5]
3.15
natural language processing
NLP
technology used to determine and identify key words and phrases within processing audio data (e.g.
call centres) and free-form text (e.g. the body of an email)
Note 1 to entry: This technology is able to reduce words to their base constructs and perform other actions, such
as stemming, along with locating similar words or phrases without user intervention. This technology also varies
greatly from standard IDR technology due to the ability to automatically update rules as determined by the users
without the need for technical intervention. This technology is best suited for unstructured documents.
[SOURCE: ISO/TR 22957:2018, 3.7]
3.16
artificial neural network
neural network
neural net
ANN
network of primitive processing elements connected by weighted links with adjustable weights, in
which each element produces a value by applying a nonlinear function to its input values, and transmits
it to other elements or presents it as an output value
Note 1 to entry: Whereas some neural networks are intended to simulate the functioning of neurons in the nervous
system, most neural networks are used in artificial intelligence as realizations of the connectionist model.
Note 2 to entry: Examples of nonlinear functions are a threshold function, a sigmoid function, and a polynomial
function.
[SOURCE: ISO/IEC 2382:2015, 2120625, modified — Notes to entry 3 to 5 deleted.]
© ISO 2021 – All rights reserved PROOF/ÉPREUVE 3
---------------------- Page: 8 ----------------------
ISO/TR 24291:2021(E)
3.17
prediction
output of an algorithm after it has been trained on a historical dataset and applied to new data when
forecasting the likelihood of a particular outcome
3.18
robotics
techniques involved in designing, building, and using robots
[SOURCE: ISO/IEC/IEEE 24765:2017, 3.3554]
3.19
speech recognition
automatic speech recognition
conversion, by a functional unit, of a speech signal to a representation of the content of the speech
Note 1 to entry: The content to be recognized can be expressed as a proper sequence of words or phonemes.
[SOURCE: ISO/IEC 19794-13:2018, 3.22]
4 Abbreviated terms
CBIR Content-based case retrieval
CNN Convolutional Neural Net
DB Database
EMR Electronic Medical Records
ICU Intensive Care Unit
IDR Intelligent Document Recognition
IoT Internet of Things
OR Operation Room
5 Categories for defining use cases of machine learning in medicine
5.1 Categories based on technology
5.1.1 General
AI techniques used in medicine can be summarized as in Table 1.
Table 1 — Technology based categories of artificial intelligence and their purposes
Technology Purpose
Providing high quality treatments by increasing the precision and accuracy
Robotics
of the surgical process.
Within golden time, proper treatment could be performed by continuously
Continuous monitoring
monitoring of patient condition and alerting nurses.
Machine learning Predict response by analyzing data affecting treatment outcomes.
Self-learning ability to process large amounts of medical imaging records,
Deep learning
reducing uncertainty in medical treatment decisions.
4 PROOF/ÉPREUVE © ISO 2021 – All rights reserved
---------------------- Page: 9 ----------------------
ISO/TR 24291:2021(E)
Table 1 (continued)
Technology Purpose
Process large-scale medical images and apply them to detect diseases, diag-
Image processing
nosis, etc.
Translate long descriptive character sets such as electric medical records to
Natural language processing
be interpreted.
By recognizing voice and language of patient, dictate important information
Audio recognition
in electric medical records.
Process vast patient health records held by healthcare organizations and
Bigdata analysis
provide tailored recommendations to patients and providers.
Prediction modeling Apply AI models to predict outcomes such as predicting risk disorders.
5.1.2 Robotics
In robotics, AI can provide high quality treatments by increasing the precision and accuracy of the
surgical process. For example, it can control the trajectory, depth, and speed of the robot movements
with high precision and can go where traditional tools cannot. It can also reduce the burdens of the
surgeons during surgery by providing the same, repetitive movements without fatigue.
5.1.3 Continuous monitoring
Proper treatment within golden time could be performed by continuously monitoring of patient
condition and alerting nurses by AI. AI model with continuous monitoring data also can alert the
clinicians before onset.
5.1.4 Machine learning
By using traditional machine learning methods, AI can be used to predict response by analysing data
affecting treatment outcomes.
5.1.5 Deep learning
By using deep learning, self-learning ability to process large amounts of imaging and audios records
in medicine, reducing uncertainty in medical treatment decisions including computer aided detection,
computer aided diagnosis, computer aided differential diagnosis, and clinical decision support system.
Deep learning can handle multiple different types of clinical data such as images, texts, and signals at
the same time.
5.1.6 Image processing
In image processing, AI can be used to process large-scale medical images and apply them to detect
diseases, diagnosis, etc. AI for clinical image handling has demonstrated its performance in clinical
settings.
5.1.7 Natural language processing
In NLP, AI can be used to translate long descriptive character sets such as electric medical records to be
interpreted, i.e. extracting the information from unstructured electronic medical records.
5.1.8 Audio recognition
In audio recognition, by recognizing voice and language of patient, AI can automatically dictate
important information in electric medical records.
© ISO 2021 – All rights reserved PROOF/ÉPREUVE 5
---------------------- Page: 10 ----------------------
ISO/TR 24291:2021(E)
5.1.9 Bigdata analysis
In bigdata analysis, AI can be used to process vast patient health records held by healthcare
organizations and provide tailored recommendations to patients and providers.
5.1.10 Prediction modeling
In prediction modeling, AI models can be applied to predict outcomes such as predicting risk disorders.
5.2 Categories based on medical specialty
5.2.1 General
The applications of AI in medicine can be categorized by medical specialty as in Table 2.
Table 2 — Typical use cases of artificial intelligence in medicine in each medical specialty
Specialty Use Cases
Quantification, computer aided detection, diagnosis, and differential diagnosis on radio-
Radiology
logic images; Automated dictation system.
Quantification, computer aided detection, diagnosis, and differential diagnosis on patho-
Pathology
logic images; Automated dictation system.
Dermatology Skin cancer detection and classification.
Ophthalmology Eye disease detection and classification in retina image and OCT.
Internal Medicine Complication/outcome prediction; clinical decision support system.
Quantification, computer aided detection, diagnosis, and differential diagnosis on cardi-
Cardiology
ologic images; Complication/outcome prediction; clinical decision support system.
Neurology, Urology,
Complication/outcome prediction; clinical decision support system.
Surgery
Anaesthesiologist,
Intensive Care Unit Continuous monitoring system; Complication/outcome prediction.
(ICU)
Emergency Triage system; Continuous monitoring system; Complication/outcome prediction.
5.2.2 Radiology and Pathology
In Radiology and Pathology, AI can be used for fully automated quantification, computer aided
detection, diagnosis, and differential diagnosis on radiologic or pathologic images and automated
dictation system.
5.2.3 Dermatology
In Dermatology, skin cancer detection and classification could be typical one of use cases.
5.2.4 Ophthalmology
In Ophthalmology, AI can be used for eye disease detection and classification in retina image and OCT.
5.2.5 Internal Medicine
In Internal Medicine, complication and outcome prediction, and clinical decision support system could
be one of important use cases.
6 PROOF/ÉPREUVE © ISO 2021 – All rights reserved
---------------------- Page: 11 ----------------------
ISO/TR 24291:2021(E)
5.2.6 Cardiology
In Cardiology, quantification, computer aided detection, diagnosis, and differential diagnosis on
cardiologic images, complication and outcome prediction, and clinical decision support system could be
one of use cases.
5.2.7 Neurology, Urology, Surgery
In Neurology, Urology, Surgery, AI can be used for complication and outcome prediction, and clinical
decision support system.
5.2.8 Anaesthesiologist, Intensive Care Unit
In Anaesthesiologist, Intensive Care Unit (ICU), continuous monitoring system and complication and
outcome prediction could be one of important applications.
5.2.9 Emergency
In Emergency, automated triages system, continuous monitoring system, complication and outcome
prediction could be one of important applications.
5.3 Categories based on medical usage
5.3.1 General
The examples of AI in medicine can be categorized by their medical usages including clinical trials,
clinical assistance, data-based precision medicine, diagnostic imaging, logistics in ORs and wards,
robot surgery, and drug development in medicine (see Table 3 and Figure 1).
Table 3 — Categories of use cases of medical usages
Usage Sub-usage Purpose
Case selection AI-based search techniques help find the right disease and patient,
Clinical trials
reduce the time to prepare for clinical trials, and improve objectivity.
Assistant service Convergence of IoT technology, voice recognition technology and
and wellness with artificial intelligence technology, efficient reservation, diagnosis and
IoT medical treatment process, business information update and custom-
Clinical Assis-
ized curation.
tance
Automated dictation Speech recognition and document generation technology that can
automate diagnosis and recording of readings, and read and structure
medical terminology.
Genome and genet- Prediction, diagnosis and treatment by analysing and modeling asso-
ics ciations of big data based on genome, multimodal medical imaging and
clinical pathology for personalized medicine.
Prediction of
Data based pre- Assist physicians in making final decisions by notifying them of case-
complication and
cision medicine based risks or complications in treatment and drug use.
mixed drug
Recommendation of Recommend additional diagnostic testing processes to increase ac-
exam curacy and lower risk, utilizing artificial intelligence and big data for
precise diagnosis and treatment.
© ISO 2021 – All rights reserved PROOF/ÉPREUVE 7
---------------------- Page: 12 ----------------------
ISO/TR 24291:2021(E)
Table 3 (continued)
Usage Sub-usage Purpose
Normal selection Early diagnose medical images in advance to reduce clinical burden by
differentiating the true normal cases.
Content based case Helps diagnose by searching and visualizing similar cases from numer-
retrieval ous cases in DB.
Medical Imaging
Generating reading Artificial intelligence-based medical image analysis and natural lan-
and Diagnostic
reports guage processing technology are integrated to generate readings that
can assist the reading of image s
...
Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.