Abstract

Purpose : This study aims to develop an artificial intelligence (AI)-based system for accurately grading arrested retinal development using optical coherence tomography (OCT) across various manufacturers. We employ deep learning techniques, specifically Unsupervised Domain Adaptation (UDA), to create a device-agnostic classification model distinguishing normal and abnormal retinal development. Methods : Foveal scans from three OCT manufacturers (TM-OCT1, HH-OCT1, TM-OCT2, TM-OCT3) were collected and annotated from datasets exceeding 20,000 OCT scans. The dataset was divided into training (80%) and testing (20%) sets. We utilised Convolutional Neural Networks (CNN) with ResNet50 backbone, assessing each pair as the source (for supervised training) and target (for unsupervised training). The diagnostic accuracy of the AI models was compared to seven clinician graders with varying experience (1 to 10 years), evaluating sensitivity, specificity, and overall accuracy. Results : The cross-domain binary classification demonstrated exceptional diagnostic accuracy ranging from 87.75% to 96.06%. Sensitivity and specificity metrics further validated the robustness of our AI system (sensitivity: 88.68% to 98.38%, specificity: 78.00% to 98.02%). Notably, the model trained on TM-OCT1 and HH-OCT1 achieved 96.06% accuracy, 98.38% sensitivity, and 89.11% specificity on the HH-OCT1 test set, with an Area Under the Curve (AUC) of 99.4 (95% CI). These outcomes highlight the system's ability to distinguish normal and abnormal retinal development across diverse OCT devices. Conclusions : Our study demonstrates, for the first time, the feasibility of employing a device-agnostic AI system in paediatric OCT interpretation without additional labelled data. The AI system's diagnostic performance is comparable to a clinician with over 10 years of experience, showcasing its potential to reduce inter-examiner variability and enhance clinical care pathways. With integration of paediatric OCT into routine clinical assessment, our AI system serves as a robust foundation for the development of a real-time, frontline diagnostic tool for retinal developmental disorders.

Abstract

Large-scale deployment of fully autonomous vehicles requires a very high degree of robustness to unstructured traffic, weather conditions, and should prevent unsafe mispredictions. While there are several datasets and benchmarks focusing on segmentation for drive scenes, they are not specifically focused on safety and robustness issues. We introduce the IDD-AW dataset, which provides 5000 pairs of high-quality images with pixel-level annotations, captured under rain, fog, low light, and snow in unstructured driving conditions. As compared to other adverse weather datasets, we provide i.) more annotated images, ii.) paired Near-Infrared (NIR) image for each frame, iii.) larger label set with a 4-level label hierarchy to capture unstructured traffic conditions. We benchmark state-of-the-art models for semantic segmentation in IDD-AW. We also propose a new metric called “Safe mean Intersection over Union (Safe mIoU)” for hierarchical datasets which penalizes dangerous mispredictions that are not captured in the traditional definition of mean Intersection over Union (mIoU). The results show that IDD-AW is one of the most challenging datasets to date for these tasks. The dataset and code will be available here: http://iddaw.github.io.

Abstract

Molecular spectroscopy studies the interaction of molecules with electromagnetic radiation, and interpreting the resultant spectra is invaluable for deducing the molecular structures. However, predicting the molecular structure from spectroscopic data is a strenuous task that requires highly specific domain knowledge. DeepSPInN is a deep reinforcement learning method that predicts the molecular structure when given Infrared and 13C Nuclear magnetic resonance spectra by formulating the molecular structure prediction problem as a Markov decision process (MDP) and employs Monte-Carlo tree search to explore and choose the actions in the formulated MDP. On the QM9 dataset, DeepSPInN is able to predict the correct molecular structure for 91.5% of the input spectra in an average time of 77 seconds for molecules with less than 10 heavy atoms. This study is the first of its kind that uses only infrared and 13C nuclear magnetic resonance spectra for molecular structure prediction without referring to any pre-existing spectral databases or molecular fragment knowledge bases, and is a leap forward in automated molecular spectral analysis.

Abstract

The discovery of potential therapeutic agents for life‑threatening diseases has become a significant problem. There is a requirement for fast and accurate methods to identify drug‑like molecules that can be used as potential candidates for novel targets. Existing techniques like high‑throughput screening and virtual screening are time‑consuming and inefficient. Traditional molecule generation pipelines are more efficient than virtual screening but use time‑consuming docking software. Such docking functions can be emulated using Machine Learning models with comparable accuracy and faster execution times. However, we find that when pre‑trained machine learning models are employed in generative pipelines as oracles, they suffer from model degradation in areas where data is scarce. In this study, we propose an active learning‑based model that can be added as a supplement to enhanced molecule generation architectures. The proposed method uses uncertainty sampling on the molecules created by the generator model and dynamically learns as the generator samples molecules from different regions of the chemical space. The proposed framework can generate molecules with high binding affinity with ∼a 70% improvement in runtime compared to the baseline model by labeling only ∼30% of molecules compared to the baseline oracle.

Abstract

Air pollution is a growing concern across the world. Road traffic is one of the major contributors to air pollution in urban areas. One of the approaches to solve this problem is to design a transportation policy that i) avoids extreme pollution in any area, ii) enables short transit times, and iii) makes effective use of the road capacities. Previous work to address this problem named MaxFlow-MCMC algorithm, proposed a novel sampling-based approach for this problem. In this work, we propose a significantly faster extension to the algorithm without compromising on the performance involving the following contributions: (a) We provide the first construction of a Markov Chain to sample a set of k-optimal max flow solutions directly from a planar graph. (b) We simulate traffic on large-scale real-world roadmaps using the SUMO traffic simulator. We observe a significant speed improvement in the range of 22 to 242 times in our experiments while obtaining lesser average pollution.

Abstract

Semantic segmentation provides a pixel-level understanding of an image essential for various scene-understanding vision tasks. However, semantic segmentation models demand significant computational resources during training and inference. These requirements pose a challenge in resource-constraint scenarios. To address this issue, we present a compression algorithm based on differentiable meta-pruning through hypernetwork: MPHyp. Our proposed method MPHyp utilizes hypernetworks that take latent vectors as input and output weight matrices for the segmentation model. L1 sparsification follows the proximal gradient optimizer, updates the latent vectors and introduces sparsity leading to automatic model pruning. The proposed method offers the benefit of achieving controllable compression during the training and significantly reducing the training time. We compare our methodology with a popular pruning approach and demonstrate its efficacy by reducing the number of parameters and floating point operations while maintaining the mean Intersection over Union (mIoU) metric. We conduct experiments on two widely accepted semantic segmentation architectures: UNet and ERFNet. Our experiments and ablation study demonstrate the effectiveness of our proposed methodology by achieving efficient and reasonable segmentation results.

Abstract

Large-scale deployment of fully autonomous vehicles re- quires a very high degree of robustness to unstructured traf- fic, weather conditions, and should prevent unsafe mispre- dictions. While there are several datasets and benchmarks focusing on segmentation for drive scenes, they are not specifically focused on safety and robustness issues. We introduce the IDD-AW dataset, which provides 5000 pairs of high-quality images with pixel-level annotations, cap- tured under rain, fog, low light, and snow in unstructured driving conditions. As compared to other adverse weather datasets, we provide i.) more annotated images, ii.) paired Near-Infrared (NIR) image for each frame, iii.) larger label set with a 4-level label hierarchy to capture unstructured traffic conditions. We benchmark state-of-the-art models for semantic segmentation in IDD-AW. We also propose a new metric called “Safe mean Intersection over Union (Safe mIoU)” for hierarchical datasets which penalizes danger- ous mispredictions that are not captured in the traditional definition of mean Intersection over Union (mIoU). The re- sults show that IDD-AW is one of the most challenging datasets to date for these tasks. The dataset and code will be available here: http://iddaw.github.io

Abstract

The discovery of potential therapeutic agents for life-threatening diseases has become a significant problem. There is a requirement for fast and accurate methods to identify drug-like molecules that can be used as potential candidates for novel targets. Existing techniques like high-throughput screening and virtual screening are time-consuming and inefficient. Traditional molecule generation pipelines are more efficient than virtual screening but use time-consuming docking software. Such docking functions can be emulated using Machine Learning models with comparable accuracy and faster execution times. However, we find that when pre-trained machine learning models are employed in generative pipelines as oracles, they suffer from model degradation in areas where data is scarce. In this study, we propose an active learning-based model that can be added as a supplement to enhanced molecule generation architectures. The proposed method uses uncertainty sampling on the molecules created by the generator model and dynamically learns as the generator samples molecules from different regions of the chemical space. The proposed framework can generate molecules with high binding affinity with ∼ a 70% improvement in runtime compared to the baseline model by labeling only ∼ 30% of molecules compared to the baseline oracle.

Abstract

A significant cause of air pollution in urban areas worldwide is the high volume of road traffic. Long-term exposure to severe pollution can cause serious health issues. One approach towards tackling this problem is to design a pollution-aware traffic routing policy that balances multiple objectives of i) avoiding extreme pollution in any area ii) enabling short transit times, and iii) making effective use of the road capacities. We propose a novel sampling-based approach for this problem. We provide the first construction of a Markov Chain that can sample integer max flow solutions of a planar graph, with theoretical guarantees that the probabilities depend on the aggregate transit length. We designed a traffic policy using diverse samples and simulated traffic on real-world road maps using the SUMO traffic simulator. We observe a considerable decrease in areas with severe pollution when experimented with maps of large cities across the world compared to other approaches.

Abstract

Sparse neural networks are shown to give accurate predictions competitive to denser versions, while also minimizing the number of arithmetic operations performed. However current GPU hardware can only exploit structured sparsity patterns for better efficiency. We propose a framework for generating structured multilevel block sparse neural networks by using the theory of graph products. Our Ramanujan bipartite graph product (RBGP) framework uses products of Ramanujan graphs to obtain the best connectivity for a given level of sparsity. This essentially ensures that the i.) the networks has the structured block sparsity for which runtime efficient algorithms exists, ii.) the model gives high prediction accuracy, due to the better expressive power derived from the connectivity of the graph, iii.) the graph data structure has a succinct representation that can be stored efficiently in memory. We use our framework to design a specific connectivity pattern called RBGP4 which makes efficient use of the memory hierarchy available on GPU. We benchmark our approach on image classification and machine translation tasks with an edge (Jetson Nano 2GB) as well as server (V100) GPUs. When compared with commonly used sparsity patterns like unstructured and block, we obtain significant speedups while achieving the same level of accuracy.