Machine learning has advanced substantially in the past few years, and there are many generic solutions freely available to classify text and images. The solutions are so straightforward to set up and run that having a software background is no longer necessary to perform machine learning experimentation. These systems are being adapted in many ways, and it seems only natural that those in the medical field may wish to see how machine learning might help with their research. This research examines if off-the-shelf machine learning systems are suitable for research by medical professionals who do not have software backgrounds. If all doctors who wish to experiment with machine learning could have an adequate system available, the impact on research could be substantial. This investigation applies a commonly available machine learning practice lab to medical images. As part of this investigation, we evaluated the TensorFlow for Poets (TFP) tutorial from Google Code Labs with openly available medical images provided by Kaggle Inc. While we would not recommend our test results as a basis for diagnosing medical conditions, the results were encouraging enough to suggest that using off-the-shelf systems can offer a promising opportunity to expand machine learning research into those with medical, but not software backgrounds.
|Title of host publication||Information Management and Big Data - 6th International Conference, SIMBig 2019, Proceedings|
|Editors||Juan Antonio Lossio-Ventura, Nelly Condori-Fernandez, Jorge Carlos Valverde-Rebaza|
|Number of pages||13|
|State||Published - 1 Jan 2020|
|Event||Communications in Computer and Information Science - |
Duration: 1 Jan 2020 → …
|Name||Communications in Computer and Information Science|
|Conference||Communications in Computer and Information Science|
|Period||1/01/20 → …|
Bibliographical notePublisher Copyright:
© Springer Nature Switzerland AG 2020.
- Image classification
- Machine learning