Medical Image Retrieval via Nearest Neighbor Search on Pre-trained Image Features.

Nearest neighbor search, also known as NNS, is a technique used to locate the points in a high-dimensional space closest to a given query point. This technique has multiple applications in medicine, such as searching large medical imaging databases, disease classification, and diagnosis. However, when the number of points is significantly large, the brute-force approach for finding the nearest neighbor becomes computationally infeasible. Therefore, various approaches have been developed to make the search faster and more efficient to support the applications. With a focus on medical imaging, this paper proposes DenseLinkSearch (DLS), an effective and efficient algorithm that searches and retrieves the relevant images from heterogeneous sources of medical images. Towards this, given a medical database, the proposed algorithm builds an index that consists of pre-computed links of each point in the database. The search algorithm utilizes the index to efficiently traverse the database in search of the nearest neighbor. We also explore the role of medical image feature representation in content-based medical image retrieval tasks. We propose a Transformer-based feature representation technique that outperformed the existing pre-trained Transformer-based approaches on benchmark medical image retrieval datasets. We extensively tested the proposed NNS approach and compared the performance with state-of-the-art NNS approaches on benchmark datasets and our created medical image datasets. The proposed approach outperformed the existing approaches in terms of retrieving accurate neighbors and retrieval speed. In comparison to the existing approximate NNS approaches, our proposed DLS approach outperformed them in terms of lower average time per query and ≥ 99% R@10 on 11 out of 13 benchmark datasets. We also found that the proposed medical feature representation approach is better for representing medical images compared to the existing pre-trained image models. The proposed feature extraction strategy obtained an improvement of 9.37%, 7.0%, and 13.33% in terms of P@5, P@10, and P@20, respectively, in comparison to the best-performing pre-trained image model. The source code and datasets of our experiments are available at https://github.com/deepaknlp/DLS.

Scrub Typhus in Children: A Prospective Observational Study in a Tertiary Care Hospital in Eastern India.

Background Scrub typhus is a reemerging, acute, undifferentiating febrile illness and one of the most neglected tropical diseases, calling for an in-depth investigation into its clinical diversity, complications, and mortality, which drives us to carry out this research work. Methods Over a year, prospective observational research was carried out after gaining parental consent and institutional ethical clearance, 206 children of either gender aged between one month and 12 years who had been hospitalized with a fever for at least five days and subsequently tested positive for Orientia​​​​ tsutsugamushi were included in the study. Basic demographic information, clinical characteristics, laboratory findings, complications, related coinfections, and results were gathered and analyzed. A P-value of 0.05 was set as the statistical benchmark. Results The current study found that boys outnumbered girls. The ratio of boys to girls was 1.22:1, and the average age was 5.18 years. All had a fever (100%), and the other most frequently occurring clinical signs and symptoms were abdominal pain (16.99%), vomiting (22.33%), hepatosplenomegaly (49.51%), facial puffiness (39.32%), edema (27.18%), lymphadenopathy (19.90%), eschar (19.90%), macular-erythematous rash (17.96%), cough (21.84%), conjunctival congestion (25.24%), and headache (13.59%). Anemia (81.55%), leucocytosis (20.39%), leucopenia (6.8%), thrombocytopenia (49.51%), thrombocytosis (2.43%), and elevated serum levels of alanine aminotransferase (ALT, 57.28%) and aspartate aminotransferase (AST, 63.59%) were characteristic laboratory results. The coinfections were dengue, enteric fever, urinary tract infections, and malaria. Children who also had dengue were more likely to develop thrombocytopenia, which was statistically significant (P-value = 0.008). With doxycycline medication, early defervescence of fever occurred earlier than with azithromycin, and it was statistically significant (P-value = 0.000). The complications were hepatitis (63.59%), lower respiratory tract infections (LRTIs, 22.82%), scrub typhus meningoencephalitis (STME, 3.88%), acute kidney injury (AKI, 2.91%), myocarditis (1.46%), and acute disseminated encephalomyelitis (ADEM, 0.49%). Except for one who had ADEM, everyone was sent back home after receiving the best care possible. The average duration of hospital stay was 6.89 days. Conclusions Even in the absence of eschar, scrub typhus should be suspected in any febrile child who experiences clinical signs of meningoencephalitis syndrome, capillary leakage, skin rash, conjunctival congestion, LRTI, AKI, lymphadenopathy, hepatosplenomegaly, thrombocytopenia, and liver dysfunction in the post-monsoon season. Strong clinical suspicion and prompt anti-scrub drug administration go a long way in preventing or decreasing the morbidity and mortality of the same.

Knowledge of Breastfeeding Practices Among Mothers Attending a Tertiary Care Setting in East India.

Background Human milk offers a neonate a balanced diet for healthy growth and development, in addition to its myriad of benefits like preventing stunting, protecting against infectious and chronic diseases, and decreasing infant mortality. Objective To assess the knowledge of mothers and other factors that contribute to breastfeeding practices. Methods This is a one-year hospital-based cross-sectional study that included 400 mothers who followed up with the hospital for the healthcare of their child, aged between six and 24 months. A survey was used for data collection. Results Ninety-three percent of the mothers were from the countryside, and 78% of them were under 25 years of age. Eighty-seven percent of mothers worked at home, while 83% of mothers were part of nuclear households. Ninety-nine percent of mothers delivered their neonates at a medical facility, and 77% of mothers did so for the first time. Only 53% of mothers resorted to exclusive breastfeeding (EBF), even though 68% of mothers were aware of its significance. Thirty-six percent of mothers adopted EBF, while only 23% of women were aware that breastfeeding should be started within the first hour of childbirth. Working women (p=0.000), mothers with several children (p=0.000), mothers older than 25 years of age (p=0.002), and mothers with higher education levels than the 10th grade (p=0.000) showed good understanding and practice of breastfeeding, which was statistically significant (p<0.5). Conclusion The levels of breastfeeding awareness and practice among mothers fell short of both national statistics and WHO recommendations. All helpful information about breastfeeding should be shared with the community at large to improve the data currently available.

Effects of Porting Essie Tokenization and Normalization to Solr.

Search for information is now an integral part of healthcare. Searches are enabled by search engines whose objective is to efficiently retrieve the relevant information for the user query. When it comes to retrieving biomedical text and literature, Essie search engine developed at the National Library of Medicine (NLM) performs exceptionally well. However, Essie is a software system developed for NLM that has ceased development and support. On the other hand, Solr is a popular opensource enterprise search engine used by many of the world's largest internet sites, offering continuous developments and improvements along with the state-of-the-art features. In this paper, we present our approach to porting the key features of Essie and developing custom components to be used in Solr. We demonstrate the effectiveness of the added components on three benchmark biomedical datasets. The custom components may aid the community in improving search methods for biomedical text retrieval.

Filariasis presenting as pancytopenia in a child.

A female child hailing from South Asia, India presented with pallor, multiple petechiae and ecchymosis. Based on the clinical picture and demography, the differentials considered were pancytopenia of nutritional origin, acute leukaemia, autoimmune and infective aetiologies. After ruling these out by respective tests, a literature review was done which revealed the possibility of filariasis especially in a patient with eosinophilia which was present in our case. A repeat peripheral blood smear study with a nocturnally drawn sample revealed multiple microfilariae and a diagnosis of filariasis was made. The patient was treated with triple drug therapy of diethylcarbamazine (6 mg/kg), ivermectin (6 µg/kg) and albendazole (400 mg) administered as a single dose. Subsequent haemograms showed improved cell counts. This along with a previous handful of case reports emphasises filariasis as one of the differentials of pancytopenia and should be kept in mind while evaluating for the same, especially in the endemic areas.

Question-driven summarization of answers to consumer health questions.

Automatic summarization of natural language is a widely studied area in computer science, one that is broadly applicable to anyone who needs to understand large quantities of information. In the medical domain, automatic summarization has the potential to make health information more accessible to people without medical expertise. However, to evaluate the quality of summaries generated by summarization algorithms, researchers first require gold standard, human generated summaries. Unfortunately there is no available data for the purpose of assessing summaries that help consumers of health information answer their questions. To address this issue, we present the MEDIQA-Answer Summarization dataset, the first dataset designed for question-driven, consumer-focused summarization. It contains 156 health questions asked by consumers, answers to these questions, and manually generated summaries of these answers. The dataset's unique structure allows it to be used for at least eight different types of summarization evaluations. We also benchmark the performance of baseline and state-of-the-art deep learning approaches on the dataset, demonstrating how it can be used to evaluate automatically generated summaries.

A dataset of clinically generated visual questions and answers about radiology images.

Radiology images are an essential part of clinical decision making and population screening, e.g., for cancer. Automated systems could help clinicians cope with large amounts of images by answering questions about the image contents. An emerging area of artificial intelligence, Visual Question Answering (VQA) in the medical domain explores approaches to this form of clinical decision support. Success of such machine learning tools hinges on availability and design of collections composed of medical images augmented with question-answer pairs directed at the content of the image. We introduce VQA-RAD, the first manually constructed dataset where clinicians asked naturally occurring questions about radiology images and provided reference answers. Manual categorization of images and questions provides insight into clinically relevant tasks and the natural language to phrase them. Evaluating with well-known algorithms, we demonstrate the rich quality of this dataset over other automatically constructed ones. We propose VQA-RAD to encourage the community to design VQA tools with the goals of improving patient care.