UMIN-CTR Clinical Trial

Basic information
Public title	Creating an AI model for hoarseness classification using speech analysis in the perioperative period
Acronym	Creating an AI model for hoarseness classification using speech analysis in the perioperative period
Scientific Title	Creating an AI model for hoarseness classification using speech analysis in the perioperative period
Scientific Title:Acronym	Creating an AI model for hoarseness classification using speech analysis in the perioperative period
Region	Japan

Condition
Condition	Thyroid Surgery Esophageal Cancer Surgery Dissociative Aortic Aneurysm Surgery
Classification by specialty	Surgery in general Vascular surgery Oto-rhino-laryngology Anesthesiology Intensive care medicine
Classification by malignancy	Others
Genomic information	NO

Objectives
Narrative objectives1	With the recent development of artificial intelligence (AI) technology, speech analysis systems and machine learning have become an integral part of our lives. We hypothesized that by using speech analysis systems and machine learning, it would be possible to predict the diagnosis of antegrade nerve palsy in the perioperative period using the patient's voice. In this study, we aim to create a hoarseness classification AI model using a speech analysis system. If we can identify antegrade nerve palsy (hoarseness) by voice analysis, we can easily predict the diagnosis of antegrade nerve palsy without causing patient distress, and reduce complications in the perioperative period.
Basic objectives2	Safety,Efficacy
Basic objectives -Others
Trial characteristics_1
Trial characteristics_2
Developmental phase

Assessment
Primary outcomes	The purpose of this study is to create an AI model for hoarseness classification using a speech analysis system.
Key secondary outcomes

Base
Study type	Observational

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	20 years-old <=
Age-upper limit	Not applicable
Gender	Male and Female
Key inclusion criteria	Patients scheduled for esophageal cancer surgery, dissecting aortic aneurysm surgery, or thyroid surgery at Yamagata University Hospital.
Key exclusion criteria	Patients who were not able to cooperate in the study.
Target sample size	200

Research contact person
Name of lead principal investigator	1st name Tatsuya Middle name Last name Hayasaka
Organization	Yamagata University Hospital
Division name	Department of Anesthesia
Zip code	9909585
Address	2-2-2, Iida-Nishi, Yamagata City
TEL	023-628-5400
Email	hayasakatatsuya1101@gmail.com

Public contact
Name of contact person	1st name Tatsuya Middle name Last name Hayasaka
Organization	Yamagata University Hospital
Division name	Department of Anesthesia
Zip code	9909585
Address	2-2-2, Iida-Nishi, Yamagata City
TEL	023-628-5400
Homepage URL
Email	hayasakatatsuya1101@gmail.com

Sponsor
Institute	Department of Anesthesiology, Yamagata University School of Medicine
Institute
Department

Funding Source
Organization	Department of Anesthesiology, Yamagata University School of Medicine
Organization
Division
Category of Funding Organization	Self funding
Nationality of Funding Organization

Other related organizations
Co-sponsor
Name of secondary funder(s)

IRB Contact (For public release)
Organization	Yamagata University Medical Ministry Council
Address	2-2-2 Iida-Nishi, Yamagata City, Yamagata Prefecture
Tel	0236285015
Email	ikekenkyu@jm.kj.yamagata-u.ac.jp

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

Other administrative information
Date of disclosure of the study information	2021 Year 05 Month 24 Day

Related information
URL releasing protocol
Publication of results	Unpublished

Result
URL related to results and publications
Number of participants that the trial has enrolled
Results
Results date posted
Results Delayed
Results Delay Reason
Date of the first journal publication of results
Baseline Characteristics
Participant flow
Adverse events
Outcome measures
Plan to share IPD
IPD sharing Plan description

Progress
Recruitment status	Open public recruiting
Date of protocol fixation	2021 Year 05 Month 24 Day
Date of IRB	2021 Year 05 Month 01 Day
Anticipated trial start date	2021 Year 05 Month 24 Day
Last follow-up date	2023 Year 05 Month 30 Day
Date of closure to data entry
Date trial data considered complete
Date analysis concluded

Other

Other related information

Patients who will undergo thyroid surgery, esophageal cancer surgery, or aortic aneurysm resection at Yamagata University Hospital between June 2021 and June 2023 will be included in the study. Before the surgery (from admission to the day before the surgery), the voice of the target patients (according to previous studies, "A-I-U-E-O", "the word 'Jack and the Beanstalk'", and a section of ATR503 sentence (about 2-3 minutes)) will be collected. After completion of the surgery, vocal fold movements will be recorded by laryngeal fiber, which is performed in normal practice (normal vocal fold movement and presence of antegrade nerve palsy will be the correct labels). After the next day of surgery, collect the voice as before the surgery. The voices of patients with a difference in voice are classified as positive, and those with no difference in voice are classified as negative. Using 80% of the total data as train data, an AI model is created based on the positive/negative data and laryngeal fiber findings. We used 20% of the total data as test data to draw ROC curve and calculate AUC. As a secondary evaluation, we will use preoperative patient data (age, gender, height, weight, body temperature, heart rate, blood pressure, oxygenation capacity, etc.) and intraoperative findings such as surgical site to examine the correlation with recurrent nerve palsy by deep learning.

Management information
Registered date	2021 Year 05 Month 05 Day
Last modified on	2021 Year 12 Month 02 Day

Link to view the page
URL(English)	https://upload.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000050258