Ajeenkya DY-Patil University

The detection of tuberculosis lesions in chest X-rays is increasingly critical but remains hindered by sparse annotations and significant class imbalance within medical imaging datasets. This study addresses these limitations by proposing an intrinsic motivation-based exploration framework, leveraging curiosity-driven algorithms to encourage artificial intelligence models to autonomously seek and interpret novel radiological cues of tuberculosis. By introducing learned novelty bonuses as intrinsic rewards, the approach enables deep models to discover rare or unannotated lesion states that would otherwise be overlooked in conventional supervised learning. Experiments conducted on a large-scale chest X-ray dataset with incomplete labels demonstrate significant improvements in exploration coverage, recall of rare lesion regions, intersection over union for lesion detection, and overall F1-score compared to state-of-the-art baselines. Notably, the model achieved higher average precision and substantially reduced annotation gaps without relying on exhaustive manual labeling, indicating robust generalization to challenging and unseen cases. The primary contribution of this work is the advancement of autonomous, curiosity-driven methods that enhance lesion discovery and diagnostic reliability in resource-constrained global healthcare environments.

Automated early detection of pulmonary tuberculosis using chest radiographs is hampered by extreme class imbalance, resulting in poor performance across real-world settings where the clinical and operational cost of missed TB cases outstrips that of false positives. In this study, we address these challenges through the development of a cost-sensitive logistic regression framework that is explicitly tailored to better represent the highly uneven penalties associated with missing TB detection cases in screening scale clinical settings. By using a large chest X-ray dataset, our approach cascades cost-awareness from the training phase to the evaluation phase and learns clinically relevant trade-offs between costs due to false negatives (FNs) and FPs - simultaneously optimizing for multiple metrics including sensitivity, specificity. The results of our experiments show that the proposed method yields a significant improvement in detection sensitivity across risk and cost regimes (by approximately 60 %) at the same level of interpretability and operational scalability required for use in resource-limited or high-burden settings. We believe that these insights can help stakeholders develop more effective and equitable TB case finding in practice, particularly with adaptive risk-cost calibration for AI-driven TB screening tools.

A novel bacterial strain, designated UC10 T , was isolated from cold, high-altitude farmland soil in the Gangotri region of the Western Himalayas, India. The strain is Gram-stain-negative, aerobic, non-spore forming, and non-motile, forming golden colonies that produce a flexirubin-like pigment. Strain UC10 T grows over a broad range of temperatures (5°C–30°C), pH (6–11), and salinities (up to 4% NaCl), with optimal growth at 30°C, pH 7.0, and 1% NaCl. The nearly full-length 16S rRNA gene sequence ( MK743979 ) shares 98.95% similarity with Dyadobacter luticola , followed by 97.68% with Dyadobacter crusticola and 97.40% with Dyadobacter koreensis , and phylogenetic analysis places UC10ᵀ in a distinct clade within the genus Dyadobacter . Whole-genome phylogenetic analyses revealed that UC10 T is closely related to Dyadobacter linearis , D. crusticola , and D. luticola but is clearly distinguished from them by low average nucleotide identity (<81%), digital DNA–DNA hybridization (<24%), and amino acid identity (<80%) values. The genome of UC10 T is 6.93 Mb with a G+C content of 46.5 mol% and encodes multiple cold adaptation-related genes, including cold-shock proteins and fatty acid desaturases. The strain also harbors genes for aromatic compound degradation and demonstrated the ability to grow in minimal medium containing sodium benzoate as the sole carbon source. Additionally, fatty acid and polar lipid profiles of UC10 T revealed unique compositions, further supporting its differentiation. The combined genomic, phenotypic, and chemotaxonomic evidence supports the designation of strain UC10 T as representing a novel species, for which the name Dyadobacter aurulentus sp. nov. is proposed. The type strain is UC10 T (= MCC 4019 T = KCTC 72455 T = JCM 34514 T ).

High-altitude, cold habitats such as the Gangotri region of the Western Himalayas remain underexplored for culturable microbial diversity. Here, we describe Dyadobacter aurulentus sp. nov., a novel cold-adapted species isolated from such an environment. This strain demonstrates unique ecological and metabolic traits, including growth at low temperatures and degradation of aromatic compounds like sodium benzoate. Genomic analysis revealed key cold adaptation features such as cold-shock proteins, fatty acid desaturases, nitrate assimilation pathways, and multidrug resistance genes, supporting survival in nutrient-limited, low-temperature soils. The strain’s distinct chemotaxonomic profile, marked by elevated C 16:0 and unique polar lipids, underscores its ecological specialization. Together, these features point to its potential utility in bioremediation and cold-environment biotechnology. This study broadens our understanding of the adaptive strategies and ecological functions of Dyadobacter spp. in extreme environments, with implications for bioprospecting cold-active enzymes and understanding resistance evolution in high-altitude microbial communities.