An artificial intelligence approach for breast cancer early risk assessment

Ying Han, Evrim Yuzgec, Mohammad T. Khasawneh

Research output: Contribution to conferencePaperpeer-review

1 Scopus citations

Abstract

Breast cancer, more than ever before, is growing among women to become one of the most common cancer types affecting their quality of life. In 2012, the American Cancer Society has estimated that there will be more than 200,000 new cases of invasive breast cancer with nearly 40,000 breast cancer related deaths. This high incident rate of breast cancer requires research focused on potential factors leading to disease development. Over the years, substantial studies have been conducted on tools for breast cancer detection and prediction models of patient survivability. There is also a need for a tool to educate currently healthy people about potential factors to help them maintain a healthy life and prevent breast cancer at early stages. Therefore, this study proposes an artificial intelligence-based risk assessment tool for estimating breast cancer risk, which has the potential to help healthy women be aware of possible risks associated with their current lifestyle and physical condition. Backpropagation (BP), Learning Vector Quantization (LVQ), Probabilistic Neural Networks (PNN), and a decision tree are applied to classify a population based on their risk factors. Results showed that BP network achieved the highest classification accuracy of 96%, sensitivity of 98.2%, and specificity of 94.6%.

Original languageEnglish
Pages1712-1719
Number of pages8
StatePublished - 2013
EventIIE Annual Conference and Expo 2013 - San Juan, Puerto Rico
Duration: 18 May 201322 May 2013

Conference

ConferenceIIE Annual Conference and Expo 2013
Country/TerritoryPuerto Rico
CitySan Juan
Period18/05/1322/05/13

Keywords

  • Artificial intelligence
  • Breast cancer
  • Decision tree
  • Neural networks
  • Risk assessment

Fingerprint

Dive into the research topics of 'An artificial intelligence approach for breast cancer early risk assessment'. Together they form a unique fingerprint.

Cite this