Breast Cancer Imaging AI Could Improve Accuracy Rate of Disease Detection by 37% [STUDY]

Breast cancer imaging has been proven to be useful for physicians in terms of breast cancer detection. Now, an artificial tool was made to improve the diagnosis capability of the method.

A new study revealed that an AI program can provide more accurate breast cancer imaging results to radiologists.

AI Sees Accurate Patterns on Breast Ultrasound Images

Some amazing findings were concluded during the trial using AI. In a study involving more than 44,700 completed ultrasound exams, the computer program was found out to be an effective tool in identifying breast cancer, Health Day reported on Friday, Sept. 24.

Additionally, the test showed a significant decrease in the needed tissue samples through artificial intelligence. It has accurately diagnosed the said condition by 37%. Moreover, it posed 27% accuracy in detecting tumors among patients.

According to a senior member of the study, Dr. Krzysztof Geras, the study has helped the experts in reading the breast ultrasound exams.

As previously stated, the chance of conducting a biopsy will be avoided or lessened just in case the breast cancer is benign.

In the process of using breast ultrasound exams, real-time images of the breast and even its related tissues are shown. This is made possible through high-frequency sound waves.

It's also interesting to learn that breast cancer imaging could be an alternative to the usual mammography test, according to Geras from Perlmutter Cancer Center.

Breast Ultrasound Could Unreliable

According to a Sept.24 report by Medical Xpress, radiation specialists frequently use ultrasound because it is cheap and easy to do. It is also believed to be much better than mammography when it comes to in-depth tissue study.

While many experts use it, breast ultrasound could be unreliable sometimes due to incorrect diagnosis. Somehow, this mistake could have an implication to the woman patient.

There are studies about breast ultrasound exams that later revealed the results to be benign following the biopsy report.

Another radiologist, Linda Moy, MD said that the addition of the machine-learning tool has helped their studies about breast cancer imaging. She noted that this AI could pave the way for the improvement of women's breast health in the future.

While the AI tool showed interesting results during the examination, the researchers would have to weigh a more balanced take on this.

Through the real-world conditions of the patients, more clinical trials should be done to prove its effectiveness.

Using the computer program, the breast exams yielded an average of 92% accuracy in diagnosing breast cancer.

To view the study entitled "Artificial intelligence system reduces false-positive findings in the interpretation of breast ultrasound exams," visit Nature.com.

Machine-Learning Model in Breast Cancer Detection

In 2017, another group of researchers found out a new way to lessen false-positive results in breast cancer screening. The experts said that the AI tool lowered the chance for the patients to undergo unnecessary surgeries.

To read more about health tech articles, you can view the top 10 AI healthcare startups for 2021.

This article is owned by Tech Times

Written by Joseph Henry