We focus on the problem of Gallbladder Cancer (GBC) detection from Ultrasound (US) images. The problem presents unique challenges to modern Deep Neural Network (DNN) techniques due to low image quality arising from noise, textures, and viewpoint variations. Tackling such challenges would necessitate precise localization performance by the DNN to identify the discerning features for the downstream malignancy prediction. While several techniques have been proposed in the recent years for the problem, all of these methods employ complex custom architectures. Inspired by the success of foundational models for natural image tasks, along with the use of adapters to fine-tune such models for the custom tasks, we investigate the merit of one such design, ViT-Adapter, for the GBC detection problem. We observe that ViT-Adapter relies pre-dominantly on a primitive CNN-based spatial prior module to inject the localization information via cross-attention, which is inefficient for our problem due to the small pathology sizes, and variability in their appearances due to non-regular structure of the malignancy. In response, we propose, LQ-Adapter, a modified Adapter design for ViT, which improves localization information by leveraging learnable content queries over the basic spatial prior module. Our method surpasses existing approaches, enhancing the mean IoU (mIoU) scores by 5.4%, 5.8%, and 2.7% over ViT-Adapters, DINO, and FocalNet-DINO, respectively on the US image-based GBC detection dataset, and establishing a new state-of-the-art (SOTA). Additionally, we validate the applicability and effectiveness of LQ-Adapter on the Kvasir-Seg dataset for polyp detection from colonoscopy images. Superior performance of our design on this problem as well showcases its capability to handle diverse medical imaging tasks across different datasets. Source code and trained models are publicly released.

In recent years, automated Gallbladder Cancer (GBC) detection has gained the attention of researchers. Current state-of-the-art (SOTA) methodologies relying on ultra-sound sonography (US) images exhibit limited generalization, emphasizing the need for transformative approaches. We observe that individual US frames may lack sufficient information to capture disease manifestation. This study advocates for a paradigm shift towards video-based GBC detection, leveraging the inherent advantages of spatiotemporal representations. Employing the Masked Autoencoder (MAE) for representation learning, we address shortcomings in conventional image-based methods. We propose a novel design called FocusMAE to systematically bias the selection of masking tokens from high-information regions, fostering a more refined representation of malignancy. Additionally, we contribute the most extensive US video dataset for GBC detection. We also note that, this is the first study on US video-based GBC detection. We validate the proposed methods on the curated dataset, and report a new SOTA accuracy of 96.4% for the GBC detection problem, against an accuracy of 84% by current Image-based SOTA – GBCNet and RadFormer, and 94.7% by Video-based SOTA – AdaMAE. We further demonstrate the generality of the proposed FocusMAE on a public CT-based Covid detection dataset, reporting an improvement in accuracy by 3.3% over current baselines. Project page with source code, trained models, and data is available at: https://gbc-iitd.github.io/focusmae.

An accurate prediction of El Niño Southern Oscillation (ENSO) holds the key to produce skillful seasonal weather predictions across the globe. All the statistical and dynamical ENSO models developed in the past four decades face a common problem, spring predictability barrier, which is the sudden drop in the skill of the ENSO prediction when the forecast is initiated before the onset of boreal summer. Recent studies suggest that data-driven machine learning models can overcome the spring predictability barrier. We show that using a convolutional long short-term memory (ConvLSTM) network, the monthly mean Nino3.4 index can be skillfully predicted up to one year ahead. The model was also able to predict strong El Niño cases, such as 1997–1998 and 2015–2016 a year ahead. Our results suggest that the proposed ConvLSTM model has significant skill in multiseasonal weather predictions.

The synoptic-scale (3–7 days) variability is a dominant contributor to the Indian summer monsoon (ISM) seasonal precipitation. An accurate prediction of ISM precipitation by dynamical or statistical model remains a challenge. Here, we show that the sea level pressure (SLP) can be used as a proxy to predict the active-break cycle as well as the genesis of low-pressure systems (LPSs), using a deep learning model, namely convolutional long short-term memory (ConvLSTM) networks. The deep learning model is able to reliably predict the daily SLP anomalies over central India and the Bay of Bengal at a lead time of 7 days. As the fluctuations in SLP drive the changes in the strength of the atmospheric circulation, the prediction of SLP anomalies is useful in predicting the intensity of ISM. A comparison of the ConvLSTM predicted SLP with the forecast of a conventional numerical weather prediction model indicates that the deep learning model possesses better skill in capturing the synoptic-scale SLP fluctuations over central India and Bay of Bengal.