DDSM 'leader board' for CAD

[Philip Teare]

Hi

I'm wondering if there is a list of best attempts at training machine learning CAD systems against the DDSM sets and what the accuracies may be. Simple classification rather than segmentation etc, but any data is of interest.

I realise context may come into it, but anything along these lines would be useful (and relevant papers of course). If you have done similar, what were the sensitivity and specificity rates you attained? I'll share mine if I'm able as soon as I can decode the thing!


Best
Phil

[GopalKaremore]

Dear Philip,

Have a look at my PhD thesis (2nd Chapter) where I summarized and compared the mammographic parenchymal texture techniques (through Machine Learning) used in various applications such as, microcalcification detection, mass characterization and tissue characterization. Emphasis is given on techniques that have been used for risk assessment application on screening mammograms. Performance is evaluated and compared by Receiver Operative Characteristic curve analysis on one of the most commonly used database available in public domain, such as mini-MIAS and DDSM.

(from Chapter 4-7) I also compared DDSM dataset with other case control study such as Nijmegen, HRT, and ER specific risk mammgram data.

Link: http://www.forskningsdatabasen.dk/en/catalog/2193075110

PDF Link: http://www.diku.dk/forskning/phd-studiet/phd/davidg_thesis.pdf/Thesis_Gopal_Latest.pdf

I hope it helps you.

Good Luck!

Sincerely,

Gopal Karemore