This sample Java plugin runs a deep learning algorithm for mammography.
Note that contrarily to the "Basic" and "Dcm4Che" samples that can run
using Java 1.8, this Java plugin requires Java 11.

(1) Make sure to build the C++ plugin:

# cd ../../Plugin/
# mkdir Build
# cd Build
# cmake .. -DCMAKE_BUILD_TYPE=Release -DUSE_SYSTEM_ORTHANC_SDK=OFF
# make -j4


(2) Make sure to build the Java wrapper:

# cd ../../JavaSDK/
# mkdir Build
# cd Build
# cmake .. -DUSE_SYSTEM_ORTHANC_SDK=OFF -DCMAKE_INSTALL_PREFIX=..
# make -j4 install


(3) Compile the Java plugin using Maven:

# cd ../../Samples/MammographyDeepLearning
# mvn compile package

IMPORTANT: By default, this sample plugin targets GNU/Linux
distributions. To run this plugin on Microsoft Windows or Apple macOS,
make sure to uncomment the suitable classifier section in "pom.xml"
(i.e., "win-x86_64", "osx-x86_64", or "osx-aarch64").


(4) Start Orthanc:

On Ubuntu 20.04 or 22.04 (replace "mainline" with the version of the
Java plugin for Orthanc, cf. "../global.properties"):

# LD_LIBRARY_PATH=/usr/lib/jvm/java-11-openjdk-amd64/lib/server/ \
  ORTHANC_JAVA_VERSION=mainline \
  Orthanc ./configuration.json

IMPORTANT: The DICOMweb plugin must be installed in the current
directory to run the Stone Web viewer.


(5) Open the Orthanc Explorer interface at: http://localhost:8042/