Deep learning-based prediction of response to HER2-targeted neoadjuvant chemotherapy from pre-treatment dynamic breast MRI: A multi-institutional validation study
Authors:
Nathaniel Braman,
Mohammed El Adoui,
Manasa Vulchi,
Paulette Turk,
Maryam Etesami,
Pingfu Fu,
Kaustav Bera,
Stylianos Drisis,
Vinay Varadan,
Donna Plecha,
Mohammed Benjelloun,
Jame Abraham,
Anant Madabhushi
Abstract:
Predicting response to neoadjuvant therapy is a vexing challenge in breast cancer. In this study, we evaluate the ability of deep learning to predict response to HER2-targeted neo-adjuvant chemotherapy (NAC) from pre-treatment dynamic contrast-enhanced (DCE) MRI acquired prior to treatment. In a retrospective study encompassing DCE-MRI data from a total of 157 HER2+ breast cancer patients from 5 i…
▽ More
Predicting response to neoadjuvant therapy is a vexing challenge in breast cancer. In this study, we evaluate the ability of deep learning to predict response to HER2-targeted neo-adjuvant chemotherapy (NAC) from pre-treatment dynamic contrast-enhanced (DCE) MRI acquired prior to treatment. In a retrospective study encompassing DCE-MRI data from a total of 157 HER2+ breast cancer patients from 5 institutions, we developed and validated a deep learning approach for predicting pathological complete response (pCR) to HER2-targeted NAC prior to treatment. 100 patients who received HER2-targeted neoadjuvant chemotherapy at a single institution were used to train (n=85) and tune (n=15) a convolutional neural network (CNN) to predict pCR. A multi-input CNN leveraging both pre-contrast and late post-contrast DCE-MRI acquisitions was identified to achieve optimal response prediction within the validation set (AUC=0.93). This model was then tested on two independent testing cohorts with pre-treatment DCE-MRI data. It achieved strong performance in a 28 patient testing set from a second institution (AUC=0.85, 95% CI 0.67-1.0, p=.0008) and a 29 patient multicenter trial including data from 3 additional institutions (AUC=0.77, 95% CI 0.58-0.97, p=0.006). Deep learning-based response prediction model was found to exceed a multivariable model incorporating predictive clinical variables (AUC < .65 in testing cohorts) and a model of semi-quantitative DCE-MRI pharmacokinetic measurements (AUC < .60 in testing cohorts). The results presented in this work across multiple sites suggest that with further validation deep learning could provide an effective and reliable tool to guide targeted therapy in breast cancer, thus reducing overtreatment among HER2+ patients.
△ Less
Submitted 22 January, 2020;
originally announced January 2020.
3D shape retrieval basing on representatives of classes
Authors:
M. Benjelloun,
E. W. Dadi,
E. M. Daoudi
Abstract:
In this paper, we present an improvement of our proposed technique for 3D shape retrieval in classified databases [2] which is based on representatives of classes. Instead of systematically matching the object-query with all 3D models of the database, our idea presented in [2] consist, for a classified database, to represent each class by one representative that is used to orient the retrieval pro…
▽ More
In this paper, we present an improvement of our proposed technique for 3D shape retrieval in classified databases [2] which is based on representatives of classes. Instead of systematically matching the object-query with all 3D models of the database, our idea presented in [2] consist, for a classified database, to represent each class by one representative that is used to orient the retrieval process to the right class (the class excepted to contain 3D models similar to the query). In order to increase the chance to fall in the right class, our idea in this work is to represent each class by more than one representative. In this case, instead of using only one representative to decide which is the right class we use a set of representatives this will contribute certainly to improving the relevance of retrieval results. The obtained experimental results show that the relevance is significantly improved.
△ Less
Submitted 27 December, 2018; v1 submitted 21 October, 2018;
originally announced October 2018.
