UMIN-CTR Clinical Trial

Basic information
Official scientific title of the study	Development of imaging diagnosis system for emergency patients by artificial intelligence
Title of the study (Brief title)	Development of imaging diagnosis system
Region	Japan

Condition
Condition	Emergency diseases and injuries
Classification by specialty	Emergency medicine
Classification by malignancy	Others
Genomic information	NO

Objectives
Narrative objectives1	Development of imaging diagnosis systems for emergency patients by artificial intelligence
Basic objectives2	Others
Basic objectives -Others	Development of systems and verification of accuracy
Trial characteristics_1	Exploratory
Trial characteristics_2	Pragmatic
Developmental phase	Not applicable

Assessment
Primary outcomes	Sensitivity and specificity of imaging diagnosis
Key secondary outcomes

Base
Study type	Others,meta-analysis etc

Study design
Basic design
Randomization
Randomization unit
Blinding
Control
Stratification
Dynamic allocation
Institution consideration
Blocking
Concealment

Intervention
No. of arms
Purpose of intervention
Type of intervention
Interventions/Control_1
Interventions/Control_2
Interventions/Control_3
Interventions/Control_4
Interventions/Control_5
Interventions/Control_6
Interventions/Control_7
Interventions/Control_8
Interventions/Control_9
Interventions/Control_10

Eligibility
Age-lower limit	Not applicable
Age-upper limit	Not applicable
Gender	Male and Female
Key inclusion criteria	All the patients whose CT data are preserved in the server
Key exclusion criteria	None
Target sample size	80000

Research contact person
Name of lead principal investigator	Shigeki Kushimoto
Organization	Tohoku University Graduate School of Medicine
Division name	Emergency and Critical Care Medicine
Address	1-1 Seiryomachi, Aoba-ku, Sendai, 980-8574, Japan
TEL	022-717-7489
Email	kussie@emergency-medicine.med.tohoku.ac.jp

Public contact
Name of contact person	Daisuke Kudo
Organization	Tohoku University Graduate School of Medicine
Division name	Emergency and Critical Care Medicine
Address	1-1 Seiryomachi, Aoba-ku, Sendai, 980-8574, Japan
TEL	022-717-7489
Homepage URL
Email	kudodaisuke@med.tohoku.ac.jp

Sponsor
Institute	Tohoku University
Institute
Department

Funding Source
Organization	Self-funding by the profit organization which is included in the joint research team.
Organization
Division
Category of Funding Organization	Profit organization
Nationality of Funding Organization

Other related organizations
Co-sponsor	Hokkaido University Graduate School of Medicine Diverta Inc.
Name of secondary funｄer(s)

Secondary IDs
Secondary IDs	NO
Study ID_1
Org. issuing International ID_1
Study ID_2
Org. issuing International ID_2
IND to MHLW

Institutions
Institutions	東北大学病院（宮城）/Tohoku University Hospital 北海道大学病院（北海道）/Hokkaido University Hospital

Other administrative information
Date of disclosure of the study information	2017 Year 12 Month 18 Day

Progress
Recruitment status	Preinitiation
Date of protocol fixation	2017 Year 12 Month 08 Day
Anticipated trial start date	2018 Year 01 Month 01 Day
Last follow-up date
Date of closure to data entry
Date trial data considered complete
Date analysis concluded

Related information
URL releasing protocol
Publication of results	Unpublished
URL releasing results
Results
Other related information	Patients and Methods Subjects are the patients admitted to the hospitals between October 2006 and September 2017 and whose CT data are preserved in the server. Imaging diagnosis includes all emergency diseases and injuries. We will collect the data including CT imaging, imaging diagnosis, and clinical information. We also collect the data of patients without abnormal CT findings as controls for deep learning. The area of CT image includes head, face, neck, chest, abdomen, and pelvis. 1st step We will input the data to the machine learning software. The machine learning software will analyze and classify the data, then it will create algorithms for imaging diagnosis. We estimate that data from 70,000 patients will be needed to create the algorithms. 2nd step We will use the data from different patients in this step. We will examine sensitivity and specificity of the algorithms that will be created in the 1st step by comparing with the imaging diagnosis previously reported by radiologists. 3rd step We will repeat the 1st and 2nd steps in order to improve sensitivity and specificity of the algorithms for imaging diagnosis.

Management information
Registered date	2017 Year 12 Month 16 Day
Last modified on	2017 Year 12 Month 18 Day

Link to view the page
URL(English)	https://upload.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000034700