Predictive model for determining the indications for automated 3D ultrasound for screening patients at low risk of developing breast tumors

Automatic ultrasound examination of the breast (3D ultrasound) has become an important tool in the diagnosis of breast cancer. It is believed that 3D ultrasound has high reproducibility, low dependence on the operator, less time spent on obtaining images, and automatic three-dimensional reconstructi...

Full description

Saved in:
Bibliographic Details
Main Authors: A. E. Garanina, A. V. Kholin
Format: Article
Language:Russian
Published: QUASAR, LLC 2024-06-01
Series:Исследования и практика в медицине
Subjects:
Online Access:https://www.rpmj.ru/rpmj/article/view/1001
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832568628643364864
author A. E. Garanina
A. V. Kholin
author_facet A. E. Garanina
A. V. Kholin
author_sort A. E. Garanina
collection DOAJ
description Automatic ultrasound examination of the breast (3D ultrasound) has become an important tool in the diagnosis of breast cancer. It is believed that 3D ultrasound has high reproducibility, low dependence on the operator, less time spent on obtaining images, and automatic three-dimensional reconstruction of the entire breast.Purpose of the study. To develop indications for 3D ultrasound based on predictive screening models for patients with a low risk of developing breast tumors based on the identification of the most significant risk factors.Patients and methods. A retro-prospective clinical study has been conducted from February 2019 to May 2023. A total of 2794 patients were included in the study. All patients underwent clinical examination, palpation, collected information on socio-demographic data and potential risk factors for breast cancer, and 2D ultrasound was also performed. The group under the age of 40 included 1,511 patients, of whom 628 underwent 3D ultrasound. The sample of 40 years and older included 1,283 patients, 655 of whom underwent 3D ultrasound. Mammography was performed in patients aged 40 and older. Quantitative and qualitative indicators of anamnesis and clinical examination, as well as MMH results in patients over 40 years old, were recorded. Based on these data, a logistic regression was compiled, followed by the selection of the most significant model by cutting off insignificant factors according to the p-level of significance and presenting the model as a ROC curve.Results. The most significant risk factors for the detection of breast cancer were identified. Based on their screening with 3D ultrasound in a group up to 40 years of age, it can be used in 95.96 % and is not indicated in 4.04 %. The presented model in the group up to 40 years worked correctly in 99.21 %. While screening with 3D ultrasound in a group of 40 years and older in 84.26 % is appropriate and not indicated in 15.74 %. The presented model worked correctly in 97.12 %.Conclusion. The study identified important pre-diagnostic factors for the choice of a diagnostic algorithm for breast examination in women of different age groups, and determined the indications for 3D ultrasound. The developed algorithms will help optimize screening and referral for additional examinations, which is of practical importance for improving diagnostics and optimizing healthcare resources.
format Article
id doaj-art-34ba9e841b8b4d1391217b3421404a98
institution Kabale University
issn 2410-1893
language Russian
publishDate 2024-06-01
publisher QUASAR, LLC
record_format Article
series Исследования и практика в медицине
spelling doaj-art-34ba9e841b8b4d1391217b3421404a982025-02-03T00:57:40ZrusQUASAR, LLCИсследования и практика в медицине2410-18932024-06-01112576810.17709/2410-1893-2024-11-2-5540Predictive model for determining the indications for automated 3D ultrasound for screening patients at low risk of developing breast tumorsA. E. Garanina0A. V. Kholin1I. I. Mechnikov North-Western State Medical University;<p> СМТ Clinic AO, Polyclinic Complex<p> St. Petersburg, Russian FederationI. I. Mechnikov North-Western State Medical University;<p> St. Petersburg, Russian FederationAutomatic ultrasound examination of the breast (3D ultrasound) has become an important tool in the diagnosis of breast cancer. It is believed that 3D ultrasound has high reproducibility, low dependence on the operator, less time spent on obtaining images, and automatic three-dimensional reconstruction of the entire breast.Purpose of the study. To develop indications for 3D ultrasound based on predictive screening models for patients with a low risk of developing breast tumors based on the identification of the most significant risk factors.Patients and methods. A retro-prospective clinical study has been conducted from February 2019 to May 2023. A total of 2794 patients were included in the study. All patients underwent clinical examination, palpation, collected information on socio-demographic data and potential risk factors for breast cancer, and 2D ultrasound was also performed. The group under the age of 40 included 1,511 patients, of whom 628 underwent 3D ultrasound. The sample of 40 years and older included 1,283 patients, 655 of whom underwent 3D ultrasound. Mammography was performed in patients aged 40 and older. Quantitative and qualitative indicators of anamnesis and clinical examination, as well as MMH results in patients over 40 years old, were recorded. Based on these data, a logistic regression was compiled, followed by the selection of the most significant model by cutting off insignificant factors according to the p-level of significance and presenting the model as a ROC curve.Results. The most significant risk factors for the detection of breast cancer were identified. Based on their screening with 3D ultrasound in a group up to 40 years of age, it can be used in 95.96 % and is not indicated in 4.04 %. The presented model in the group up to 40 years worked correctly in 99.21 %. While screening with 3D ultrasound in a group of 40 years and older in 84.26 % is appropriate and not indicated in 15.74 %. The presented model worked correctly in 97.12 %.Conclusion. The study identified important pre-diagnostic factors for the choice of a diagnostic algorithm for breast examination in women of different age groups, and determined the indications for 3D ultrasound. The developed algorithms will help optimize screening and referral for additional examinations, which is of practical importance for improving diagnostics and optimizing healthcare resources.https://www.rpmj.ru/rpmj/article/view/1001breast cancerultrasoundautomated volumetric breast scanning
spellingShingle A. E. Garanina
A. V. Kholin
Predictive model for determining the indications for automated 3D ultrasound for screening patients at low risk of developing breast tumors
Исследования и практика в медицине
breast cancer
ultrasound
automated volumetric breast scanning
title Predictive model for determining the indications for automated 3D ultrasound for screening patients at low risk of developing breast tumors
title_full Predictive model for determining the indications for automated 3D ultrasound for screening patients at low risk of developing breast tumors
title_fullStr Predictive model for determining the indications for automated 3D ultrasound for screening patients at low risk of developing breast tumors
title_full_unstemmed Predictive model for determining the indications for automated 3D ultrasound for screening patients at low risk of developing breast tumors
title_short Predictive model for determining the indications for automated 3D ultrasound for screening patients at low risk of developing breast tumors
title_sort predictive model for determining the indications for automated 3d ultrasound for screening patients at low risk of developing breast tumors
topic breast cancer
ultrasound
automated volumetric breast scanning
url https://www.rpmj.ru/rpmj/article/view/1001
work_keys_str_mv AT aegaranina predictivemodelfordeterminingtheindicationsforautomated3dultrasoundforscreeningpatientsatlowriskofdevelopingbreasttumors
AT avkholin predictivemodelfordeterminingtheindicationsforautomated3dultrasoundforscreeningpatientsatlowriskofdevelopingbreasttumors