In the field of medical imaging, 3D deep learning models play a crucial role in building powerful predictive models of disease progression. However, the size of these models presents significant challenges, both in terms of computational resources and data requirements. Moreover, achieving high-quality pretraining of 3D models proves to be even more challenging. To address these issues, hybrid 2.5D approaches provide an effective solution for utilizing 3D volumetric data efficiently using 2D models. Combining 2D and 3D techniques offers a promising avenue for optimizing performance while minimizing memory requirements. In this paper, we explore 2.5D architectures based on a combination of convolutional neural networks (CNNs), long short-term memory (LSTM), and Transformers. In addition, leveraging the benefits of recent non-contrastive pretraining approaches in 2D, we enhanced the performance and data efficiency of 2.5D techniques even further. We demonstrate the effectiveness of architectures and associated pretraining on a task of predicting progression to wet age-related macular degeneration (AMD) within a six-month period on two large longitudinal OCT datasets.
The lack of reliable biomarkers makes predicting the conversion from intermediate to neovascular age-related macular degeneration (iAMD, nAMD) a challenging task. We develop a Deep Learning (DL) model to predict the future risk of conversion of an eye from iAMD to nAMD from its current OCT scan. Although eye clinics generate vast amounts of longitudinal OCT scans to monitor AMD progression, only a small subset can be manually labeled for supervised DL. To address this issue, we propose Morph-SSL, a novel Self-supervised Learning (SSL) method for longitudinal data. It uses pairs of unlabelled OCT scans from different visits and involves morphing the scan from the previous visit to the next. The Decoder predicts the transformation for morphing and ensures a smooth feature manifold that can generate intermediate scans between visits through linear interpolation. Next, the Morph-SSL trained features are input to a Classifier which is trained in a supervised manner to model the cumulative probability distribution of the time to conversion with a sigmoidal function. Morph-SSL was trained on unlabelled scans of 399 eyes (3570 visits). The Classifier was evaluated with a five-fold cross-validation on 2418 scans from 343 eyes with clinical labels of the conversion date. The Morph-SSL features achieved an AUC of 0.766 in predicting the conversion to nAMD within the next 6 months, outperforming the same network when trained end-to-end from scratch or pre-trained with popular SSL methods. Automated prediction of the future risk of nAMD onset can enable timely treatment and individualized AMD management.
Robust forecasting of the future anatomical changes inflicted by an ongoing disease is an extremely challenging task that is out of grasp even for experienced healthcare professionals. Such a capability, however, is of great importance since it can improve patient management by providing information on the speed of disease progression already at the admission stage, or it can enrich the clinical trials with fast progressors and avoid the need for control arms by the means of digital twins. In this work, we develop a deep learning method that models the evolution of age-related disease by processing a single medical scan and providing a segmentation of the target anatomy at a requested future point in time. Our method represents a time-invariant physical process and solves a large-scale problem of modeling temporal pixel-level changes utilizing NeuralODEs. In addition, we demonstrate the approaches to incorporate the prior domain-specific constraints into our method and define temporal Dice loss for learning temporal objectives. To evaluate the applicability of our approach across different age-related diseases and imaging modalities, we developed and tested the proposed method on the datasets with 967 retinal OCT volumes of 100 patients with Geographic Atrophy, and 2823 brain MRI volumes of 633 patients with Alzheimer's Disease. For Geographic Atrophy, the proposed method outperformed the related baseline models in the atrophy growth prediction. For Alzheimer's Disease, the proposed method demonstrated remarkable performance in predicting the brain ventricle changes induced by the disease, achieving the state-of-the-art result on TADPOLE challenge.
Recent contrastive learning methods achieved state-of-the-art in low label regimes. However, the training requires large batch sizes and heavy augmentations to create multiple views of an image. With non-contrastive methods, the negatives are implicitly incorporated in the loss, allowing different images and modalities as pairs. Although the meta-information (i.e., age, sex) in medical imaging is abundant, the annotations are noisy and prone to class imbalance. In this work, we exploited already existing temporal information (different visits from a patient) in a longitudinal optical coherence tomography (OCT) dataset using temporally informed non-contrastive loss (TINC) without increasing complexity and need for negative pairs. Moreover, our novel pair-forming scheme can avoid heavy augmentations and implicitly incorporates the temporal information in the pairs. Finally, these representations learned from the pretraining are more successful in predicting disease progression where the temporal information is crucial for the downstream task. More specifically, our model outperforms existing models in predicting the risk of conversion within a time frame from intermediate age-related macular degeneration (AMD) to the late wet-AMD stage.
Mammograms are commonly employed in the large scale screening of breast cancer which is primarily characterized by the presence of malignant masses. However, automated image-level detection of malignancy is a challenging task given the small size of the mass regions and difficulty in discriminating between malignant, benign mass and healthy dense fibro-glandular tissue. To address these issues, we explore a two-stage Multiple Instance Learning (MIL) framework. A Convolutional Neural Network (CNN) is trained in the first stage to extract local candidate patches in the mammograms that may contain either a benign or malignant mass. The second stage employs a MIL strategy for an image level benign vs. malignant classification. A global image-level feature is computed as a weighted average of patch-level features learned using a CNN. Our method performed well on the task of localization of masses with an average Precision/Recall of 0.76/0.80 and acheived an average AUC of 0.91 on the imagelevel classification task using a five-fold cross-validation on the INbreast dataset. Restricting the MIL only to the candidate patches extracted in Stage 1 led to a significant improvement in classification performance in comparison to a dense extraction of patches from the entire mammogram.
Chest radiographs are primarily employed for the screening of cardio, thoracic and pulmonary conditions. Machine learning based automated solutions are being developed to reduce the burden of routine screening on Radiologists, allowing them to focus on critical cases. While recent efforts demonstrate the use of ensemble of deep convolutional neural networks(CNN), they do not take disease comorbidity into consideration, thus lowering their screening performance. To address this issue, we propose a Graph Neural Network (GNN) based solution to obtain ensemble predictions which models the dependencies between different diseases. A comprehensive evaluation of the proposed method demonstrated its potential by improving the performance over standard ensembling technique across a wide range of ensemble constructions. The best performance was achieved using the GNN ensemble of DenseNet121 with an average AUC of 0.821 across thirteen disease comorbidities.
Chest radiographs are primarily employed for the screening of pulmonary and cardio-/thoracic conditions. Being undertaken at primary healthcare centers, they require the presence of an on-premise reporting Radiologist, which is a challenge in low and middle income countries. This has inspired the development of machine learning based automation of the screening process. While recent efforts demonstrate a performance benchmark using an ensemble of deep convolutional neural networks (CNN), our systematic search over multiple standard CNN architectures identified single candidate CNN models whose classification performances were found to be at par with ensembles. Over 63 experiments spanning 400 hours, executed on a 11:3 FP32 TensorTFLOPS compute system, we found the Xception and ResNet-18 architectures to be consistent performers in identifying co-existing disease conditions with an average AUC of 0.87 across nine pathologies. We conclude on the reliability of the models by assessing their saliency maps generated using the randomized input sampling for explanation (RISE) method and qualitatively validating them against manual annotations locally sourced from an experienced Radiologist. We also draw a critical note on the limitations of the publicly available CheXpert dataset primarily on account of disparity in class distribution in training vs. testing sets, and unavailability of sufficient samples for few classes, which hampers quantitative reporting due to sample insufficiency.
Automated Computer Aided diagnostic tools can be used for the early detection of glaucoma to prevent irreversible vision loss. In this work, we present a Multi-task Convolutional Neural Network (CNN) that jointly segments the Optic Disc (OD), Optic Cup (OC) and predicts the presence of glaucoma in color fundus images. The CNN utilizes a combination of image appearance features and structural features obtained from the OD-OC segmentation to obtain a robust prediction. The use of fewer network parameters and the sharing of the CNN features for multiple related tasks ensures the good generalizability of the architecture, allowing it to be trained on small training sets. The cross-testing performance of the proposed method on an independent validation set acquired using a different camera and image resolution was found to be good with an average dice score of 0.92 for OD, 0.84 for OC and AUC of 0.95 on the task of glaucoma classification illustrating its potential as a mass screening tool for the early detection of glaucoma.