diff --git a/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/__init__.py b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/__init__.py
new file mode 100644
index 0000000000..e69de29bb2
diff --git a/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_solver.prototxt b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_solver.prototxt
new file mode 100644
index 0000000000..d76ba8c7f1
--- /dev/null
+++ b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_solver.prototxt
@@ -0,0 +1,24 @@
+# The train/test net protocol buffer definition
+train_net: "ipcai_train.prototxt"
+test_net: "ipcai_test.prototxt"
+# test_iter specifies how many forward passes the test should carry out.
+# In the case of MNIST, we have test batch size 100 and 100 test iterations,
+# covering the full 10,000 testing images.
+test_iter: 100
+# Carry out testing every 500 training iterations.
+test_interval: 500
+# The base learning rate, momentum and the weight decay of the network.
+base_lr: 0.01
+momentum: 0.0
+weight_decay: 0.0005
+# The learning rate policy
+lr_policy: "inv"
+gamma: 0.0001
+power: 0.75
+# Display every 100 iterations
+display: 1000
+# The maximum number of iterations
+max_iter: 10000
+# snapshot intermediate results
+snapshot: 5000
+snapshot_prefix: "snapshot"
diff --git a/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_test.prototxt b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_test.prototxt
new file mode 100644
index 0000000000..8bdd2313aa
--- /dev/null
+++ b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_test.prototxt
@@ -0,0 +1,77 @@
+layer {
+  name: "data"
+  type: "HDF5Data"
+  top: "data"
+  top: "label"
+  hdf5_data_param {
+    source: "ipcai_test_hdf5.h5_list.txt"
+    batch_size: 50
+  }
+}
+layer {
+  name: "fc1"
+  type: "InnerProduct"
+  bottom: "data"
+  top: "fc1"
+  inner_product_param {
+    num_output: 100
+    weight_filler {
+      type: "xavier"
+    }
+    bias_filler {
+      type: "constant"
+      value: 0.1
+    }
+  }
+}
+layer {
+  name: "relu1"
+  type: "ReLU"
+  bottom: "fc1"
+  top: "fc1"
+}
+layer {
+  name: "fc2"
+  type: "InnerProduct"
+  bottom: "fc1"
+  top: "fc2"
+  inner_product_param {
+    num_output: 100
+    weight_filler {
+      type: "xavier"
+    }
+    bias_filler {
+      type: "constant"
+      value: 0.1
+    }
+  }
+}
+layer {
+  name: "relu2"
+  type: "ReLU"
+  bottom: "fc2"
+  top: "fc2"
+}
+layer {
+  name: "score"
+  type: "InnerProduct"
+  bottom: "fc2"
+  top: "score"
+  inner_product_param {
+    num_output: 1
+    weight_filler {
+      type: "xavier"
+    }
+    bias_filler {
+      type: "constant"
+      value: 0.1
+    }
+  }
+}
+layer {
+  name: "loss"
+  type: "EuclideanLoss"
+  bottom: "score"
+  bottom: "label"
+  top: "loss"
+}
diff --git a/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_train.prototxt b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_train.prototxt
new file mode 100644
index 0000000000..7367414c4a
--- /dev/null
+++ b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/ipcai_train.prototxt
@@ -0,0 +1,77 @@
+layer {
+  name: "data"
+  type: "HDF5Data"
+  top: "data"
+  top: "label"
+  hdf5_data_param {
+    source: "ipcai_train_hdf5.h5_list.txt"
+    batch_size: 100
+  }
+}
+layer {
+  name: "fc1"
+  type: "InnerProduct"
+  bottom: "data"
+  top: "fc1"
+  inner_product_param {
+    num_output: 100
+    weight_filler {
+      type: "xavier"
+    }
+    bias_filler {
+      type: "constant"
+      value: 0.1
+    }
+  }
+}
+layer {
+  name: "relu1"
+  type: "ReLU"
+  bottom: "fc1"
+  top: "fc1"
+}
+layer {
+  name: "fc2"
+  type: "InnerProduct"
+  bottom: "fc1"
+  top: "fc2"
+  inner_product_param {
+    num_output: 100
+    weight_filler {
+      type: "xavier"
+    }
+    bias_filler {
+      type: "constant"
+      value: 0.1
+    }
+  }
+}
+layer {
+  name: "relu2"
+  type: "ReLU"
+  bottom: "fc2"
+  top: "fc2"
+}
+layer {
+  name: "score"
+  type: "InnerProduct"
+  bottom: "fc2"
+  top: "score"
+  inner_product_param {
+    num_output: 1
+    weight_filler {
+      type: "xavier"
+    }
+    bias_filler {
+      type: "constant"
+      value: 0.1
+    }
+  }
+}
+layer {
+  name: "loss"
+  type: "EuclideanLoss"
+  bottom: "score"
+  bottom: "label"
+  top: "loss"
+}
diff --git a/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_create_hdf5_database.py b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_create_hdf5_database.py
new file mode 100644
index 0000000000..cf8115835a
--- /dev/null
+++ b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_create_hdf5_database.py
@@ -0,0 +1,31 @@
+import h5py, os
+import caffe
+import numpy as np
+import pandas as pd
+
+from regression.preprocessing import preprocess
+
+
+def create_hdf5(path_to_simulation_results, hdf5_name):
+
+    df = pd.read_csv(path_to_simulation_results, header=[0, 1])
+
+    X, y = preprocess(df, snr=10.)
+    y = y.values
+
+    with h5py.File(hdf5_name,'w') as H:
+        H.create_dataset('data', data=X )  # note the name X given to the dataset!
+        H.create_dataset('label', data=y )  # note the name y given to the dataset!
+    with open(hdf5_name + '_list.txt','w') as L:
+        L.write(hdf5_name)  # list all h5 files you are going to use
+
+
+data_root = "/media/wirkert/data/Data/2016_02_02_IPCAI/results/intermediate"
+
+TRAIN_IMAGES = os.path.join(data_root,
+                            "ipcai_revision_colon_mean_scattering_train_all_spectrocam.txt")
+TEST_IMAGES = os.path.join(data_root,
+                           "ipcai_revision_colon_mean_scattering_test_all_spectrocam.txt")
+
+create_hdf5(TRAIN_IMAGES, "ipcai_train_hdf5.h5")
+create_hdf5(TEST_IMAGES, "ipcai_test_hdf5.h5")
diff --git a/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_create_lmdb_database.py b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_create_lmdb_database.py
new file mode 100644
index 0000000000..ad0160c8ce
--- /dev/null
+++ b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_create_lmdb_database.py
@@ -0,0 +1,52 @@
+
+import os
+
+import pandas as pd
+import lmdb
+import caffe
+
+from regression.preprocessing import preprocess
+
+
+def create_lmdb(path_to_simulation_results, lmdb_name):
+
+    df = pd.read_csv(path_to_simulation_results, header=[0, 1])
+
+    X, y = preprocess(df, snr=10.)
+    y = y.values * 1000
+
+    # We need to prepare the database for the size. We'll set it 10 times
+    # greater than what we theoretically need. There is little drawback to
+    # setting this too big. If you still run into problem after raising
+    # this, you might want to try saving fewer entries in a single
+    # transaction.
+    map_size = X.nbytes * 10
+
+    env = lmdb.open(lmdb_name, map_size=map_size)
+
+    with env.begin(write=True) as txn:
+        # txn is a Transaction object
+        for i in range(X.shape[0]):
+            datum = caffe.proto.caffe_pb2.Datum()
+            datum.channels = X.shape[1]
+            datum.height = 1
+            datum.width = 1
+            datum.data = X[i].tobytes()  # or .tostring() if numpy < 1.9
+            datum.label = int(y[i])
+            str_id = '{:08}'.format(i)
+
+            # The encode is only essential in Python 3
+            txn.put(str_id.encode('ascii'), datum.SerializeToString())
+
+
+data_root = "/media/wirkert/data/Data/2016_02_02_IPCAI/results/intermediate"
+
+TRAIN_IMAGES = os.path.join(data_root,
+                            "ipcai_revision_colon_mean_scattering_train_all_spectrocam.txt")
+TEST_IMAGES = os.path.join(data_root,
+                           "ipcai_revision_colon_mean_scattering_test_all_spectrocam.txt")
+
+create_lmdb(TRAIN_IMAGES, "ipcai_train_lmdb")
+create_lmdb(TEST_IMAGES, "ipcai_test_lmdb")
+
+
diff --git a/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_train_caffe.py b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_train_caffe.py
new file mode 100644
index 0000000000..56f3216002
--- /dev/null
+++ b/Modules/Biophotonics/python/iMC/scripts/ipcai_to_caffe/script_train_caffe.py
@@ -0,0 +1,83 @@
+from pylab import *
+
+import caffe
+from caffe import layers as L, params as P
+
+
+def ipcai(database, batch_size):
+    # our version of LeNet: a series of linear and simple nonlinear transformations
+    n = caffe.NetSpec()
+
+    n.data, n.label = L.HDF5Data(batch_size=batch_size, source=database, ntop=2)
+
+    n.fc1 = L.InnerProduct(n.data, num_output=100, weight_filler=dict(type='xavier'),
+                           bias_filler=dict(type='constant', value=0.1))
+    n.relu1 = L.ReLU(n.fc1, in_place=True)
+    n.fc2 = L.InnerProduct(n.relu1, num_output=100, weight_filler=dict(type='xavier'),
+                           bias_filler=dict(type='constant', value=0.1))
+    n.relu2 = L.ReLU(n.fc2, in_place=True)
+    n.score = L.InnerProduct(n.relu2, num_output=1, weight_filler=dict(type='xavier'),
+                             bias_filler=dict(type='constant', value=0.1))
+    n.loss = L.EuclideanLoss(n.score, n.label)
+
+    return n.to_proto()
+
+with open('ipcai_train.prototxt', 'w') as f:
+    f.write(str(ipcai('ipcai_train_hdf5.h5_list.txt', 100)))
+
+with open('ipcai_test.prototxt', 'w') as f:
+    f.write(str(ipcai('ipcai_test_hdf5.h5_list.txt', 50)))
+
+caffe.set_device(0)
+caffe.set_mode_gpu()
+
+### load the solver and create train and test nets
+solver = None  # ignore this workaround for lmdb data (can't instantiate two solvers on the same data)
+solver = caffe.SGDSolver('ipcai_solver.prototxt')
+
+# each output is (batch size, feature dim, spatial dim)
+print [(k, v.data.shape) for k, v in solver.net.blobs.items()]
+
+# just print the weight sizes (we'll omit the biases)
+print [(k, v[0].data.shape) for k, v in solver.net.params.items()]
+
+solver.net.forward()  # train net
+print solver.test_nets[0].forward()  # test net (there can be more than one)
+
+niter = 100000
+test_interval = 1000
+# losses will also be stored in the log
+train_loss = zeros(niter)
+test_acc = zeros(int(np.ceil(niter / test_interval)))
+output = zeros((niter, 8, 10))
+
+# the main solver loop
+for it in range(niter):
+    solver.step(1)  # SGD by Caffe
+
+    # store the train loss
+    train_loss[it] = solver.net.blobs['loss'].data
+
+    # store the output on the first test batch
+    # (start the forward pass at fc1 to avoid loading new data)
+    solver.test_nets[0].forward(start='fc1')
+    output[it] = solver.test_nets[0].blobs['score'].data[:8]
+
+    # run a full test every so often
+    # (Caffe can also do this for us and write to a log, but we show here
+    #  how to do it directly in Python, where more complicated things are easier.)
+    if it % test_interval == 0:
+        print 'Iteration', it, 'testing...'
+        mean = 0.
+        for i in range(100):
+            solver.test_nets[0].forward()
+            mean += np.sum(np.abs(np.squeeze(solver.test_nets[0].blobs['score'].data)
+                - solver.test_nets[0].blobs['label'].data))
+        mean = mean / 5000
+        test_acc[it // test_interval] = mean * 100. # %
+
+print "final testing accuracy: ", test_acc[-1]
+
+
+print solver.test_nets[0].blobs['score'].data
+print solver.test_nets[0].blobs['label'].data
\ No newline at end of file