Federated Learning for Multi-Center Imaging Diagnostics: A Study in Cardiovascular Disease

Abstract Deep learning models can enable accurate and efficient disease diagnosis, but have thus far been hampered by the data scarcity present in the medical world. Automated diagnosis studies have been constrained by underpowered single-center datasets, and although some results have shown promise, their generalizability to other institutions remains questionable as the data heterogeneity between institutions is not taken into account. By allowing models to be trained in a distributed manner that preserves patients’ privacy, federated learning promises to alleviate these issues, by enabling diligent multi-center studies. We present the first federated learning study on the modality of cardiovascular magnetic resonance (CMR) and use four centers derived from subsets of the M&M and ACDC datasets, focusing on the diagnosis of hypertrophic cardiomyopathy (HCM). We adapt a 3D-CNN network pretrained on action recognition and explore two different ways of incorporating shape prior information to the model, and four different data augmentation setups , systematically analyzing their impact on the different collaborative learning choices. We show that despite the small size of data (180 subjects derived from four centers), the privacy preserving federated learning achieves promising results that are competitive with traditional centralized learning. We further find that federatively trained models exhibit increased robustness and are more sensitive to domain shift effects.

Download Full-text

3D Deep Learning on Medical Images: A Review

Sensors ◽

10.3390/s20185097 ◽

2020 ◽

Vol 20 (18) ◽

pp. 5097 ◽

Cited By ~ 10

Author(s):

Satya P. Singh ◽

Lipo Wang ◽

Sukrit Gupta ◽

Haveesh Goli ◽

Parasuraman Padmanabhan ◽

...

Keyword(s):

Machine Learning ◽

Deep Learning ◽

Medical Imaging ◽

Medical Images ◽

Medical Image Analysis ◽

Disease Diagnosis ◽

Imaging Data ◽

Learning Models ◽

Processing Technologies ◽

3D Cnn

The rapid advancements in machine learning, graphics processing technologies and the availability of medical imaging data have led to a rapid increase in the use of deep learning models in the medical domain. This was exacerbated by the rapid advancements in convolutional neural network (CNN) based architectures, which were adopted by the medical imaging community to assist clinicians in disease diagnosis. Since the grand success of AlexNet in 2012, CNNs have been increasingly used in medical image analysis to improve the efficiency of human clinicians. In recent years, three-dimensional (3D) CNNs have been employed for the analysis of medical images. In this paper, we trace the history of how the 3D CNN was developed from its machine learning roots, we provide a brief mathematical description of 3D CNN and provide the preprocessing steps required for medical images before feeding them to 3D CNNs. We review the significant research in the field of 3D medical imaging analysis using 3D CNNs (and its variants) in different medical areas such as classification, segmentation, detection and localization. We conclude by discussing the challenges associated with the use of 3D CNNs in the medical imaging domain (and the use of deep learning models in general) and possible future trends in the field.

Download Full-text

Benchmarking Deep Learning Models and Automated Model Design for COVID-19 Detection with Chest CT Scans

10.1101/2020.06.08.20125963 ◽

2020 ◽

Author(s):

Xin He ◽

Shihao Wang ◽

Shaohuai Shi ◽

Xiaowen Chu ◽

Jiangping Tang ◽

...

Keyword(s):

Deep Learning ◽

Health Care Professionals ◽

Data Augmentation ◽

Model Performance ◽

Chest Ct ◽

Ct Scans ◽

Data Sets ◽

Learning Models ◽

3D Cnn ◽

Novel Coronavirus

AbstractCOVID-19 pandemic has spread all over the world for months. As its transmissibility and high pathogenicity seriously threaten people’s lives, the accurate and fast detection of the COVID-19 infection is crucial. Although many recent studies have shown that deep learning based solutions can help detect COVID-19 based on chest CT scans, there lacks a consistent and systematic comparison and evaluation on these techniques. In this paper, we first build a clean and segmented CT dataset called Clean-CC-CCII by fixing the errors and removing some noises in a large CT scan dataset CC-CCII with three classes: novel coronavirus pneumonia (NCP), common pneumonia (CP), and normal controls (Normal). After cleaning, our dataset consists of a total of 340,190 slices of 3,993 scans from 2,698 patients. Then we benchmark and compare the performance of a series of state-of-the-art (SOTA) 3D and 2D convolutional neural networks (CNNs). The results show that 3D CNNs outperform 2D CNNs in general. With extensive effort of hyperparameter tuning, we find that the 3D CNN model DenseNet3D121 achieves the highest accuracy of 88.63% (F1-score is 88.14% and AUC is 0.940), and another 3D CNN model ResNet3D34 achieves the best AUC of 0.959 (accuracy is 87.83% and F1-score is 86.04%). We further demonstrate that the mixup data augmentation technique can largely improve the model performance. At last, we design an automated deep learning methodology to generate a lightweight deep learning model MNas3DNet41 that achieves an accuracy of 87.14%, F1-score of 87.25%, and AUC of 0.957, which are on par with the best models made by AI experts. The automated deep learning design is a promising methodology that can help health-care professionals develop effective deep learning models using their private data sets. Our Clean-CC-CCII dataset and source code are available at:https://github.com/arthursdays/HKBU HPML COVID-19.

Download Full-text

Automatic Optic Disc Segmentation Based on Modified Local Image Fitting Model with Shape Prior Information

Journal of Healthcare Engineering ◽

10.1155/2019/2745183 ◽

2019 ◽

Vol 2019 ◽

pp. 1-10 ◽

Cited By ~ 2

Author(s):

Yuan Gao ◽

Xiaosheng Yu ◽

Chengdong Wu ◽

Wei Zhou ◽

Xiaoliang Lei ◽

...

Keyword(s):

Optic Disc ◽

Prior Information ◽

Saliency Detection ◽

Computational Cost ◽

Retinal Disease ◽

Disease Diagnosis ◽

Contour Extraction ◽

Shape Prior ◽

Boundary Extraction ◽

Contour Evolution

Accurate optic disc (OD) detection is an essential yet vital step for retinal disease diagnosis. In the paper, an approach for segmenting OD boundary without manpower named full-automatic double boundary extraction is designed. There are two main advantages in it. (1) Since the performances and the computational cost produced by iterations of contour evolution of active contour models- (ACM-) based approaches greatly depend on the initialization, this paper proposes an effective and adaptive initial level set contour extraction approach using saliency detection and threshold techniques. (2) In order to handle unreliable information generated by intensity in abnormal retinal images caused by diseases, a modified LIF approach is presented by incorporating the shape prior information into LIF. We test the effectiveness of the proposed approach on a publicly available DIARETDB0 database. Experimental results demonstrate that our approach outperforms well-known approaches in terms of the average overlapping ratio and accuracy rate.

Download Full-text

Levenshtein Augmentation Improves Performance of SMILES Based Deep-Learning Synthesis Prediction

10.26434/chemrxiv.12562121 ◽

2020 ◽

Author(s):

Dean Sumner ◽

Jiazhen He ◽

Amol Thakkar ◽

Ola Engkvist ◽

Esben Jannik Bjerrum

Keyword(s):

Neural Networks ◽

Pattern Recognition ◽

Deep Learning ◽

Recurrent Neural Networks ◽

Data Augmentation ◽

State Of The Art ◽

Sequence Similarity ◽

Learning Models ◽

Underlying Network

<p>SMILES randomization, a form of data augmentation, has previously been shown to increase the performance of deep learning models compared to non-augmented baselines. Here, we propose a novel data augmentation method we call “Levenshtein augmentation” which considers local SMILES sub-sequence similarity between reactants and their respective products when creating training pairs. The performance of Levenshtein augmentation was tested using two state of the art models - transformer and sequence-to-sequence based recurrent neural networks with attention. Levenshtein augmentation demonstrated an increase performance over non-augmented, and conventionally SMILES randomization augmented data when used for training of baseline models. Furthermore, Levenshtein augmentation seemingly results in what we define as <i>attentional gain </i>– an enhancement in the pattern recognition capabilities of the underlying network to molecular motifs.</p>

Download Full-text

Deep Learning in Disease Diagnosis: Models and Datasets

Current Bioinformatics ◽

10.2174/1574893615999201002124021 ◽

2020 ◽

Vol 15 ◽

Author(s):

Deeksha Saxena ◽

Mohammed Haris Siddiqui ◽

Rajnish Kumar

Keyword(s):

Biological Sciences ◽

Machine Learning ◽

Deep Learning ◽

Disease Diagnosis ◽

Learning Models ◽

Data Types ◽

Related Data ◽

Abstract Level ◽

Experimental Validations ◽

Selection Of

Background: Deep learning (DL) is an Artificial neural network-driven framework with multiple levels of representation for which non-linear modules combined in such a way that the levels of representation can be enhanced from lower to a much abstract level. Though DL is used widely in almost every field, it has largely brought a breakthrough in biological sciences as it is used in disease diagnosis and clinical trials. DL can be clubbed with machine learning, but at times both are used individually as well. DL seems to be a better platform than machine learning as the former does not require an intermediate feature extraction and works well with larger datasets. DL is one of the most discussed fields among the scientists and researchers these days for diagnosing and solving various biological problems. However, deep learning models need some improvisation and experimental validations to be more productive. Objective: To review the available DL models and datasets that are used in disease diagnosis. Methods: Available DL models and their applications in disease diagnosis were reviewed discussed and tabulated. Types of datasets and some of the popular disease related data sources for DL were highlighted. Results: We have analyzed the frequently used DL methods, data types and discussed some of the recent deep learning models used for solving different biological problems. Conclusion: The review presents useful insights about DL methods, data types, selection of DL models for the disease diagnosis.

Download Full-text

Crop Disease Diagnosis using Deep Learning Models

2020 Global Conference on Wireless and Optical Technologies (GCWOT) ◽

10.1109/gcwot49901.2020.9391605 ◽

2020 ◽

Author(s):

Waleej Haider ◽

Aqeel Ur Rehman ◽

Ahmed Maqsood ◽

Syed Zurain Javed

Keyword(s):

Deep Learning ◽

Disease Diagnosis ◽

Learning Models ◽

Crop Disease

Download Full-text

Data Augmentation Through Monte Carlo Arithmetic Leads to More Generalizable Classification in Connectomics

10.1101/2020.12.16.423084 ◽

2020 ◽

Author(s):

Gregory Kiar ◽

Yohan Chatelain ◽

Ali Salari ◽

Alan C. Evans ◽

Tristan Glatard

Keyword(s):

Monte Carlo ◽

Data Augmentation ◽

Learning Models ◽

Reduction Techniques ◽

Numerical Noise ◽

Numerical Instabilities ◽

Dimensionality Reduction Techniques ◽

Improved Performance ◽

Low Dimensional ◽

Machine Learning Models

AbstractMachine learning models are commonly applied to human brain imaging datasets in an effort to associate function or structure with behaviour, health, or other individual phenotypes. Such models often rely on low-dimensional maps generated by complex processing pipelines. However, the numerical instabilities inherent to pipelines limit the fidelity of these maps and introduce computational bias. Monte Carlo Arithmetic, a technique for introducing controlled amounts of numerical noise, was used to perturb a structural connectome estimation pipeline, ultimately producing a range of plausible networks for each sample. The variability in the perturbed networks was captured in an augmented dataset, which was then used for an age classification task. We found that resampling brain networks across a series of such numerically perturbed outcomes led to improved performance in all tested classifiers, preprocessing strategies, and dimensionality reduction techniques. Importantly, we find that this benefit does not hinge on a large number of perturbations, suggesting that even minimally perturbing a dataset adds meaningful variance which can be captured in the subsequently designed models.

Download Full-text