Breast cancer is common and deadly if not treated in time. There hasn’t been public awareness of the highest order on breast cancer and yet women all over the world are found with this type of cancer every year.
According to research, at least 1.7 million women have been diagnosed with breast cancer. Some research states that incorporating artificial intelligence into disease diagnosis and treatment phases can help with clinical diagnostics and the results have been affirmative of this hypothesis.
How do scientists use AI for breast cancer detection?
Artificial intelligence is used in breast cancer detection by using data that is obtained from biopsy slides and radionics. This comes as a result of the support that has come from all corners of the world on this project. There was a need to manufacture an algorithm that is for learning and can understand mammograms by the reduction of false positive outcomes.
Artificial intelligence increases the odds of identifying metastatic breast cancer cells from biopsy slides of the lymph nodes. Needless to say, AI will operate differently on different populations.
One of the most popular ways of screening breast cancer is by the use of mammograms. This is an image of high resolution, which is taken and kept then used without inconveniencing any person and without limitations on body size and age.
The mammogram has full-field systems that have inputs and outputs. The inputs are raw images while the output consists of Post-processing. AI analyses the photos and detects breast density, mass, mass segmentation, and cancer risk assessment.
In breast cancer patients, breast masses are an obvious finding, which is why detection of them is one of the most important steps in what is called Computer-aided Diagnosis (CAD). Two calcifications can appear as small spots in a mammogram, and those are micro-calcification and macro-calcification.
At present time, the only calcifications capable of being spotted by Computer-aided Diagnosis are micro-calcifications.
Assessment of breast density is carried out using two-dimensional mammograms while in breast segmentation what is identified as true segments affects the diagnosis directly. So to separate the breast segmentations from breast masses, fuzzy contours are used, and they automatically do this in a mammogram.
However, breast segmentation is hard to spot because of varying irregularities from one person to another and it is the usage of artificial intelligence that greatly improves the prognosis.
Breast Cancer risk assessment is done by viewing certain factors. These are reproductive factors such as parity, menarche, and age during a first pregnancy, and menopause. There are also other factors like estrogen, individuals’ lifestyles, and family history.
Computer-aided diagnosis became part of the screening process two decades ago. There were numerous studies carried out to show the efficiency of CAD double reading against the single reading of radiologists.
It did not have much difference and there was an advantage of one over another, but the use of the combination of both proved more successful than using each individually. Other studies have shown that by using AI-based CAD, there is a greater scope to reach high sensitivity. As well it can be used to reduce time in reading digital breast tomosynthesis and it can also be utilized as a pre-screening tool when it comes to the exclusion of low-risk cancer mammograms.
In most cases, CAD has been functioning as decision support or alternate opinion in patient care, but there is a need for it to be subjected to proper scrutiny and it must demonstrate efficiency before it is integrated.
Artificial intelligence can also be used in immunotherapy. Immunotherapy is a way in which doctors make use of the patient’s immune system responses toward the treatment given to them. So AI can be used to predict responses to immunotherapy.
There have been studies that show that linking AI and immunotherapy responses bring light uniform trends across anatomical locations and all types of cancers.
Challenges of Artificial intelligence in breast cancer treatment
AI models use imaging when managing cancer. The challenge that arises with this is the underutilization of the patient’s history which is saved in different hospitals. User-friendly software and easy-to-access databases are incorporated into hospital systems, and it is difficult to combine engineering communities and medical efforts.
The use of AI brings the challenge of trust. Doctors are yet to trust AI so much to have it helps them in their decision-making. The doctors need adequate training to be able to utilize AI technology. The use of mobile applications has made things easier when it comes to getting data from patients.
Some applications are being used to monitor various things like heart rate and blood pressure and they have improved the quality of patient care and satisfaction. Using AI methods would mean a lot of unethical risks that would need to be considered in the areas of consent, data confidentiality, patient autonomy, and violation of privacy. But there are legal ways to control the malpractices that may arise and also prevent any kind of privacy and confidentiality violation that may occur.
Conclusion
Breast cancer has proven a burden to both patients and medical personnel but bringing in AI has helped in many steps of diagnosing cancer early and treatment, of course. It might have limitations, but when they are overcome, a lot in the breast cancer processes will be improved and so will the lives of patients of this burdensome disease.