IBM for Clinical Support Used Daily
Overall Satisfaction with IBM Cloud Object Storage
I am using a cloud storage solution to build object detection machine learning models to use in applications for slide analysis for pathology. I would like to highlight cells and get their count, mean cell volume, and any shape discrepancies and use augmented reality to project the findings onto the slide in real time to enhance the performance of the pathologist.
Pros
- Machine Learning and cloud based HIPPA compliant storage solutions for back end data support for mobile and web based applications
- IBM does offer fast machine learning interfaces for natural language processing which I will be using for a chronic cough project.
Cons
- Better help functions or API support, sometimes the documentation is a little hard to follow.
- Better GUI for machine learning construction, Python can be a bit heavy to run and use and command line interfaces are tough when you want feedback to how something is processing.
- Whole slide analysis from the developer program had some problems retrieving the correct version of sci-kit so I was unable to finish the example project to build a web server of tiled images from an SVS whole slide image for analysis. Did not see a help button but would love to have help to finish.
- Good so far
Yes most things are SQL so that is a big help along with Kubernetes.
Very effective for faster response time
Yes enterprise levels have been set independently but I have so many doctors wanting side project applications built for clinical support this is the best avenue for these sorts of things.
Speed and help with different data types
Comments
Please log in to join the conversation