Liver Mass
Conditions
Interventions
The examiners first review the pre-recorded B-mode US videos and answer the location and diagnosis of liver mass without AI. Subsequently, they refer to the information presented by AI within the same
s before and under the AI support.
Sponsors
The Japan Society of Ultrasonics in Medicine
Eligibility
Sex/Gender
All
Inclusion criteria
Inclusion criteria: For the construction of a database of B-mode US video images (1080 video images in total) for the exploratory test: Patients undergoing thorough examination and follow-up for hepatic lesions (liver mass) include the following: 1) Patients who are scheduled for abdominal ultrasound examination with the purpose of investigating newly detected liver mass. 2) Patients who have already been diagnosed with liver cysts, hepatic hemangiomas, hepatocellular carcinoma, or metastatic liver cancer based on imaging findings such as CT or MRI, or pathological evidence from procedures like biopsies, and are scheduled for regular follow-up using abdominal ultrasound." For the exploratory test of the AI for supporting US diagnosis of liver mass (20 examinors): Board-certified fellows (5 persons) and registered medical sonographers (5 persons) qualified by the Japan Society of Ultrasonics in Medicine (JSUM). Non-board certified fellow (5 persons) and Non-registered medical sonographer (5 persons) by JSUM.
Exclusion criteria
Exclusion criteria: Participants in the construction of database for B-mode US video images of liver mass include: 1) Cases where a definitive diagnosis of liver mass could not be established. 2) Cases where modification of imaging findings is anticipated due to treatment of liver tumors. 3) Cases where consent from the patient has been withdrawn. 4) Cases where obtaining consent from the individual is difficult. 5) Other cases deemed inappropriate by the attending physician.
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| Evaluation of improvement for accuracy, sensitivity, specificity, and Matthews correlation coefficient in the discrimination of malignant tumors in B-mode US examination under the support of AI | — |
Secondary
| Measure | Time frame |
|---|---|
| 1) Evaluation of improvement for the accuracy of liver tumor differentiation among four types of liver lesions (hepatocellular carcinoma, metastatic liver cancer, hepatic hemangioma, and hepatic cyst) under the support of AI. 2) Evaluation of improvement for disease-specific sensitivity and specificity in the differentiation of liver mass among four types of liver lesions under the support of AI. 3) Evaluation for precision, recall, and F-value in the detection of liver mass under the support of AI. 4) Stratified analysis of skilled (Board certified fellows and registered medical sonographers of the Japan Society of Ultrasonics in Medicines) vs. non-Skilled (non-certified fellows and non-registered medical sonographer) in primary and secondary outcomes 5) Construction of database of US video images for the clinical trial of AI-aided US diagnosis of liver tumor | — |
Countries
Japan
Contacts
Public ContactNaoshi Nishida
Kindai University Faculty of Medicine, Department of Gastroenterology and Hepatology
Outcome results
None listed