add Havaei reference to readme

README.md

@@ -59,7 +59,7 @@ Automatic tumor segmentation has the potential to decrease lag time between diag
 
 ## High Grade Gliomas
 
-<b>Glioblastoma cases each year (US)</b><sup>[4](#references)</sup>: 12,000  
+<b>Glioblastoma cases each year (US)</b><sup>[5](#references)</sup>: 12,000  
 <b>Median survival</b>: 14.6 months  
 <b>Five-year survival rate</b>: < 10%
 
@@ -70,7 +70,7 @@ High-grade malignant brain tumors are generally associated with a short life exp
 
 ## Convolutional Neural Networks
 
-Convolutional Neural Networks(CNNs) are a powerful tool in the field of image recognition. They were inspired in the late 1960s by the elucidation of how the [mammalian visual cortex works](https://en.wikipedia.org/wiki/Receptive_field): many networks neurons sensitive to a given 'receptive field' tiled over the entire visual field<sup>[1](#references)</sup>. This aspect of CNNs contributes to their high flexibility and spatial invariance, making them ideal candidates for semantic segmentatiaon of images with high disparity in locations of objects of interest. CNNs are a powerful tool in machine learning that are well suited for the challenging problem tackled in this project.
+Convolutional Neural Networks(CNNs) are a powerful tool in the field of image recognition. They were inspired in the late 1960s by the elucidation of how the [mammalian visual cortex works](https://en.wikipedia.org/wiki/Receptive_field): many networks neurons sensitive to a given 'receptive field' tiled over the entire visual field<sup>[2](#references)</sup>. This aspect of CNNs contributes to their high flexibility and spatial invariance, making them ideal candidates for semantic segmentatiaon of images with high disparity in locations of objects of interest. CNNs are a powerful tool in machine learning that are well suited for the challenging problem tackled in this project.
 
 ### Model Architecture ([code](https://github.com/naldeborgh7575/brain_segmentation/blob/master/Segmentation_Models.py))
 
@@ -131,8 +131,8 @@ When I began building the model I built an architecture based on one built by [H
 ## References
 
     1. Havaei, M. et. al, Brain Tumor Segmentation with Deep Neural Networks. arXiv preprint arXiv:1505.03540, 2015.
-    1. Hubel, D. and Wiesel, T. Receptive fields and functional architecture of monkey striate cortex. Journal of Physiology 1968.
-    2. Kistler et. al, The virtual skeleton database: an open access repository for biomedical research and collaboration. JMIR, 2013.
-    3. Menze et al., The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS), IEEE Trans. Med. Imaging, 2015.
-    4. Stupp et al., Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. The Lancet Onc., 2009.
-    5. Tustison, NJ. et. al, N4ITK: improved N3 bias correction. IEEE Trans Med Imaging, 2010.
+    2. Hubel, D. and Wiesel, T. Receptive fields and functional architecture of monkey striate cortex. Journal of Physiology 1968.
+    3. Kistler et. al, The virtual skeleton database: an open access repository for biomedical research and collaboration. JMIR, 2013.
+    4. Menze et al., The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS), IEEE Trans. Med. Imaging, 2015.
+    5. Stupp et al., Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. The Lancet Onc., 2009.
+    6. Tustison, NJ. et. al, N4ITK: improved N3 bias correction. IEEE Trans Med Imaging, 2010.

Segmentation_Models.py

@@ -202,7 +202,6 @@ def fit_model(self, X_train, y_train, X5_train = None, save=True):
         else:
             self.model_comp.fit(X_train, Y_train, batch_size=self.batch_size, nb_epoch=self.n_epoch, validation_split=0.1, show_accuracy=True, verbose=1, callbacks=[checkpointer])
 
-
     def save_model(self, model_name):
         '''
         INPUT string 'model_name': name to save model and weigths under, including filepath but not extension