Wednesday, November 2, 2016

Artificial Intelligence Expedites Breast Cancer Prediction

            A team of researchers led by Wong and Jenny C. Chang, M.D created an artificial intelligence software that scans patient’s medical charts and mammograms, interprets it, and accurately predicts if the patient needs a breast biopsy.
            Currently, women are tested thru mammograms for breast cancer with 50% yield of false positives. This leads to numerous erroneous diagnoses and unnecessary breast biopsies. Not to mention the emotional distress on patients. The new software has an accuracy of 99% and is 30 times faster than human doctors. 
            This new technology means faster and more accurate results. Doctors would benefit greatly from having the help from this AI software. All fields of study benefit greatly with the help of computers and machines to process information faster and more accurately than the human brain. This development is exciting for the medical field, especially in the field of diagnostics. Enabling physicians to enter the data from patients charts into a software that accurately diagnoses patients should be the future of medical practice.
            Personally, I want to go into Physician Assistant school and work with plastic surgeons. Lately, I have had the chance to shadow plastic surgeons and I met quite a few breast cancer patients. After witnessing first-hand how traumatic the experience can be, I believe our society would benefit greatly of more tools such as this software that increases accuracy and speed which diagnoses are given. This is an amazing medical development.


Post by SRH.

8 comments:

  1. This post is very interesting! I really liked how you gave a personal experience along with the information, it made it easier to relate and easier to understand. I think it is also very good to know about these things, as the risks of cancer becomes more and more prevalent every day, and it is wise to know how we can be notified of these things.

    M.E.

    ReplyDelete
  2. This article seems to be such a positive thing since there is a new software that is 99% accurate! I found it interesting how the future of this diagnostic is inserting the patients information into the software in order to retrieve results. I think this will be a great advance in that field. This seems to be very helpful, since many people are unaware of this new software. The more the software is able to be 'advertised' the greater chance the health organization can be to detect and fight breast cancer better.

    TL

    ReplyDelete
  3. I am actually just reading about this i had no idea there was another way to detect breast cancer, i haven't heard it in the news or clinics yet is this something thats still new and going on more trails before its released or has it already been released to the public?
    Ruth Guerra

    ReplyDelete
    Replies
    1. These results were published in a study recently (August 2016), which can be found here http://onlinelibrary.wiley.com/doi/10.1002/cncr.30245/abstract. The authors wrote a computer program that uses natural language processing algorithms--basically, it compares images of mammograms with other mammograms that are known to show cancer. And it does it very quickly, much faster than humans can read the scans, and perhaps more accurately.

      The findings look exciting and promising, but I think there will be more testing and perfecting of the software before it becomes a standardized method for diagnosing the cancer. I think a great way to first implement the software would be to prioritize scans for the doctors to read first, since I think it will still take some time before people trust computers to make cancer diagnoses. Still, the results of the study are fascinating.

      Delete
  4. Having this software is a great way to cut costs of medical bills in the future. this will also require the doctors to become more efficient in other fields were they are needed most. i was just beginning to wonder how and when algorithms are going to take place of certain medical procedures.

    MC

    ReplyDelete
  5. I had picked this article because of its application on current breast cancer treatment. As we work on ways to fight cancer, we can also work on how to improve the current diagnosis and treatments available. SRH

    ReplyDelete
  6. Does this article state what increased the accuracy of these tests to be that they are now 99% of the time correct? How long does it say until these testing methods are in effect and why have they not come around sooner being that there was a 50/50 shot of being diagnosed correctly? –JA

    ReplyDelete