We have all been reading about the FDA and their increase responsibilities for approval on medical devices where software is inherent in the functioning of a device and this story with the MelaFind device is a good example of that process. I first reported on the device back in August of 2009 with clinical trials starting. If this is approved it stands to be one heck of a device for dermatologists for detecting skin cancer.
According to this article, the FDA panel member was concerned that the device would replace human judgment and well we have that all over healthcare today with devices and it can be a tough call for sure.
Is the algorithm used to quantify tissue biomarkers stronger than digital pathology or are there cases when both are needed? Don’t forget that researchers and programmers write those algorithms too so there’s always room for error with software, like we have seen with radiation software, but it has not stopped that process though so perhaps we are on the same path here to a degree, but this is different with detecting skin cancer. To me it certainly looks like it can play a role and how it is used in a clinical surrounding is perhaps still to be adequately defined.
Well we are back to that “A” word again and more so substantiating how algorithms live with us through all decision making processes today and the luddites that can’t come to grip with this are a serious threat with illiteracy, as those algos are not going away and stand to save lives. Many drugs are designed algorithmically now until a lab is needed at a late stage, so it’s in the software and those algos. BD
MelaFind, a device designed for use by dermatologists for in-vivo detection of suspicious pigmented skin lesions, has been narrowly approved by an FDA advisory panel.