Researchers extracted leptospirosis surveillance data from the Weekly Epidemiological Reports of the Epidemiology Unit and used Department of Census and Statistics population estimates to calculate district-level incidence rates per 100,000 population. They examined temporal trends and spatial distribution with R version 4.3.1. The available abstract does not report confidence intervals, measures of variability, or other uncertainty estimates, and it does not describe checks for changes in reporting or diagnostic practices. It therefore supports a descriptive assessment of reported cases and calculated rates rather than a causal explanation for the observed patterns.

The national totals included two major surges: 7,295 reported cases in 2008 and a peak of 8,323 in 2020. Ratnapura, Kalutara and Galle frequently exceeded 800–1,200 reported cases annually during peak transmission years, while the southwestern region remained the main area of reported transmission. District-level incidence also did not simply follow population size: Kegalle and Matale frequently had higher rates per 100,000 than the more populous districts of Colombo and Gampaha. The pattern shifted during the 2011 outbreak, when Kurunegala recorded the highest case count among all districts.

Eight consistently lower-burden districts were Jaffna, Kilinochchi, Mannar, Mullaitivu, Vavuniya, Batticaloa, Ampara and Trincomalee, each reporting comparatively low case numbers and incidence rates throughout the study period. The abstract concludes that leptospirosis remained a persistent public-health challenge characterized by spatial heterogeneity and periodic national surges. It identifies district-level surveillance and population-adjusted analysis as important for locating high-burden areas and says that stronger surveillance, expanded diagnostic capacity and occupational and environmental risk-reduction strategies in high-burden districts are needed to improve control. Because only the abstract and early article version were supplied, the full methods, results and limitations of the study could not be independently assessed.

Subject of Research: District-level leptospirosis surveillance trends in Sri Lanka from 2006 to 2022

Article Title: District–level epidemiological trends of leptospirosis in Sri Lanka, 2006–2022

Article References: Kumari, P., & Dabrera, T. (2026). District–level epidemiological trends of leptospirosis in Sri Lanka, 2006–2022. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14495-w

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14495-w

Keywords: leptospirosis, Sri Lanka, epidemiological surveillance, district-level analysis, incidence rates, spatial variation, temporal trends, public health, District, level, epidemiological, trends

Among the 369 respondents, 366, or 99.2%, reported using AI-based applications. Of these respondents, ChatGPT was the most frequently used tool at 98.1%, followed by Google Gemini at 35.0%, DeepSeek at 20.2% and Microsoft Copilot at 6.3%. The main uses were subject-matter reference, reported by 89.3% of AI users, general information retrieval by 73.0%, assignment assistance by 50.8% and writing improvement by 39.6%. Self-rated proficiency was highest for ChatGPT, with 92.4% of its users rating their proficiency as good or higher. In an analysis of 360 respondents, 77.5% reported using Medscape, 37.2% used UpToDate and 15.3% used WebMD or WebMD Symptom Checker.

Most students agreed that AI could enhance learning in preclinical subjects, with 85.1% expressing agreement, compared with 77.8% for paraclinical subjects and 68.6% for clinical subjects. The questionnaire contained 12 Likert-scale items covering curriculum integration, physician roles and patient care. After adjustment for multiple comparisons, agreement with including AI in the medical curriculum differed significantly across year groups, according to a Kruskal–Wallis test (H = 29.00, Holm-adjusted p < 0.001). The abstract also reports differences in attitudes toward AI for selected items between preclinical and clinical students, with higher agreement among some earlier-year groups for curriculum support.

These findings support consideration of structured, ethically grounded AI education that addresses students’ needs for safe, effective and responsible use of AI. The study is limited by its cross-sectional design, self-reported online responses and focus on undergraduates at a single Sri Lankan medical faculty. It describes reported use, proficiency and attitudes at one point in time; it cannot establish longitudinal change, causal effects or the effectiveness of any educational intervention. The authors’ conclusion is therefore a consideration for curriculum development rather than evidence that a particular training programme improves outcomes.

Subject of Research: Medical students’ use and perceptions of artificial intelligence

Article Title: Artificial intelligence use, perceptions, and educational needs among medical undergraduates at a Sri Lankan medical faculty: a cross-sectional survey

Article References: Andadola, A. M. S. U., Walallawita, W. K. M. M. M., Perera, P. D. V. M., & Mendis, S. N. J. K. (2026). Artificial intelligence use, perceptions, and educational needs among medical undergraduates at a Sri Lankan medical faculty: a cross-sectional survey. BMC Medical Education. https://doi.org/10.1186/s12909-026-10496-0

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10496-0

Keywords: artificial intelligence, medical education, medical undergraduates, Sri Lanka, AI literacy, student perceptions, curriculum integration, ChatGPT, health information technology, Artificial, intelligence, perceptions

Participants were selected through simple random sampling, and data were collected using a researcher-developed, self-administered questionnaire. Knowledge was categorized as high, moderate or poor. The investigators examined associations between knowledge levels and demographic and professional characteristics using chi-square and Fisher’s tests.

Age was significantly associated with knowledge levels (p = 0.006). Participation in in-service programs (p = 0.042), availability of ward guidelines (p = 0.006) and reading scientific literature (p = 0.014) were also associated with knowledge. No significant associations were reported for gender, education level or professional experience.

These findings represent measured statistical associations and do not show that any of the associated factors caused the differences in knowledge. The available evidence comes from the abstract of a descriptive, cross-sectional study and is limited to 317 nurses at three teaching hospitals. The design did not measure changes in knowledge over time or establish whether training, guidelines or literature reading produced higher scores.

The abstract did not report whether the significant factors were independently associated in a multivariable analysis, nor did it provide details about the questionnaire’s content, response rate or full statistical model. The evidence also did not assess whether knowledge translated into clinical practice or affected patient outcomes. The authors proposed structured training and access to updated guidelines as targeted interventions, but these were recommendations rather than evaluated interventions or demonstrated improvements.

Subject of Research: Stoma-care knowledge and associated factors among surgical nurses in Sri Lanka

Article Title: Knowledge and associated factors related to stoma care among surgical nurses in Sri Lanka

Article References: Madushan De Silva, D. K., Tharushi Uthpala, U. H., Hirun Udasara Dharmarathne, E. A., Dona Sandushani Kaushalya, K. A., & Asurakkody, T. A. (2026). Knowledge and associated factors related to stoma care among surgical nurses in Sri Lanka. BMC Health Services Research. https://doi.org/10.1186/s12913-026-15661-1

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15661-1

Keywords: stoma care, surgical nursing, Sri Lanka, nurses, cross-sectional study, knowledge levels, in-service training, ward guidelines, scientific literature, Knowledge, associated, factors

The analysis simulated NAFLD progression from 2007 to 2042 using one-year cycles and nine disease stages, ranging from early non-alcoholic steatohepatitis to liver transplantation. It incorporated obesity trends, age variations and diabetes comorbidity. Epidemiological inputs for prevalence and incidence came from published Sri Lankan community-based cohorts, while transition probabilities were adapted from published stage-specific matrices. Direct costs were based on local specialist consensus and private-sector hospital tariffs and were expressed in 2024 Sri Lankan rupees and US dollars.

The model estimated average annual per-patient costs in 2024 ranging from USD 82 at the early stage of non-alcoholic steatohepatitis to USD 377 at advanced stages. The study examined disease progression, healthcare costs and productivity losses, although the supplied abstract does not provide separate productivity-loss estimates. A deterministic sensitivity analysis used an alternative published transition matrix. The simulations, cost aggregation and analyses were performed in Microsoft Excel, with R used to cross-check matrix computations.

These findings are model estimates rather than direct measurements of disease prevalence or spending. The projections depend on the selected transition probabilities, epidemiological inputs, cost assumptions and the modeled effect of diabetes comorbidity. The abstract reports the model’s scope and principal results but does not describe the magnitude or direction of changes under the alternative transition matrix, and it does not present a comparison with other countries or health conditions. The authors concluded that the estimated burden supports a need for prevention and early intervention strategies, but the supplied evidence does not report outcomes from any such interventions.

Subject of Research: Modeled long-term economic burden of non-alcoholic fatty liver disease in Sri Lanka

Article Title: The economic burden of non-alcoholic fatty liver disease in Sri Lanka: a 35-year Markov model analysis

Article References: Niriella, M. A., de Silva, T., Fathima, H., Martinus, K., Wijesingha, I. P., Dassanayake, A. S., De Silva, A. P., & de Silva, H. J. (2026). The economic burden of non-alcoholic fatty liver disease in Sri Lanka: a 35-year Markov model analysis. BMC Health Services Research. https://doi.org/10.1186/s12913-026-15677-7

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15677-7

Keywords: NAFLD, Sri Lanka, economic burden, Markov model, healthcare costs, productivity losses, diabetes comorbidity, disease progression, health economics, economic, burden, non-alcoholic

The researchers built a kinetic-based quasi-equilibrium infection-and-recovery model with data-driven components and control events. A Sugeno-type fuzzy inference system estimated per-capita vector density from climatic variables and past dengue incidence, while a Gamma distribution represented human contact patterns. Weekly reported dengue cases, rainfall, temperature and humidity were incorporated with specified time lags for model variables. The model also represented two socio-economic events: the large-scale cleaning and fogging campaign conducted in 2018 following the 2017 outbreak, and reduced mobility during COVID-19 restrictions in 2020. The abstract does not provide enough methodological detail to determine whether mobility itself was entered with a time lag.

In the case study using data from the CMC area of Sri Lanka, simulated dengue incidence patterns closely matched the reported data across 2016–2024, with the stated accuracy levels varying by year. Comparisons with a traditional kinetic susceptible-infected-recovered model showed negligible differences in estimated infected-host populations. The researchers interpreted this as qualitative and quantitative agreement between the two modeling approaches. The model also reflected seasonal trends associated with climatic data and social behavior, according to the abstract.

The study presents a framework for combining climatic, vector-related and human-behavioral information in dengue transmission modeling. Its authors propose that such integration could assist transmission prediction and planning of control strategies, particularly where vaccine coverage is limited. That application remains a proposed use rather than evidence of improved real-world control outcomes. The reported evaluation is geographically bounded to one Sri Lankan area, and the abstract does not describe validation in other locations, uncertainty intervals, performance against additional modeling methods or the effects of individual interventions separately from the combined model. It also identifies more detailed information on human movement and behavior as a possible direction for improving predictive accuracy.

Subject of Research: Kinetic modeling of dengue transmission using climate, incidence and socio-economic data

Article Title: A robust kinetic theory-based model for dengue transmission integrating climatic and socio-economic dynamics

Article References: Hansika, S., Erandi, H., Perera, S., Attanayake, C., & Amadi, M. (2026). A robust kinetic theory-based model for dengue transmission integrating climatic and socio-economic dynamics. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14492-z

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14492-z

Keywords: dengue transmission, kinetic model, mathematical modeling, Sri Lanka, climate variables, vector density, case study, COVID-19 mobility, vector control, robust, kinetic, theory-based

Both the mitochondrial and nuclear trees strongly supported Sri Lankan Schistura as a monophyletic group, with 100% bootstrap support. This pattern is consistent with the island’s loaches having descended from one colonization event rather than representing several independent arrivals from South or Southeast Asia. The study did not include comparable molecular samples from peninsular India, where the ancestral population of the Sri Lankan radiation would have originated, so that broader relationship remains less fully resolved.

A tree-based molecular species delimitation analysis divided the mitochondrial dataset into eight molecular operational taxonomic units, or MOTUs, designated L1 through L8. A second, distance-based method generally recovered similar groupings but identified nine units: it split L7 into four, combined L2 with L3, and combined L4 with L5. These differences are not decisive evidence for a fixed number of species.

Both methods depend on sampling density and the number of haplotypes detected, and Sri Lanka’s steep terrain can make genetic divergence within a lineage appear unusually low while exaggerating separation between lineages. The eight MOTUs are therefore best treated as provisional evolutionary lineages whose limits may be refined with additional sampling and more informative nuclear data. The lineages have a pronounced geographic structure.

L1 is restricted to downstream sites in the Mahaweli basin at elevations of 1,155–1,182 metres above sea level, whereas L3 occurs across 17 localities in that basin between 243 and 1,266 metres, including the headwaters of the Menik, Kumbukkan and Gal rivers. L2 occupies the headwaters of the Maa, Attanagalu and Kelani rivers and is the sister lineage of L3. L4 occupies lowlands of the Kalu and Bentara basins, while L5 corresponds to the recognized Schistura scripta and L6 to S. madhavai.

In the second major clade, L7 is a mid-elevation lineage found at 125–648 metres in the Kalu, Gin, Nilwala and Walawe basins, while the higher-elevation L8 occurs at 571–835 metres, including specimens from the Kalu and Walawe headwaters. Genetic isolation between river basins was generally strong. Haplotype networks showed that, except within L3, sampled mitochondrial haplotypes were not shared across watersheds, even between adjacent basins.

L3 was the exception, with haplotypes occurring across the Mahaweli, Menik, Gal, Kumbukkan and Kelani systems. One L3 specimen from Makandawa in the Kelani basin, only about 10 kilometres from the Kelani–Mahaweli watershed, also grouped with the widespread highland lineage. The authors considered headwater capture a possible explanation for such transfers, but the available sampling identified only this isolated case and could not establish the precise mechanism.

The evidence therefore supports limited movement between some headwater systems rather than broad mixing throughout the island’s wet zone. Body measurements and coloration showed that genetic separation is only partly mirrored by visible differences. Principal component analysis of 39 size-corrected measurements and counts separated L1, L3, L5 and L8 clearly, whereas L2, L4, L6 and L7 overlapped substantially in morphological space.

Multiple correspondence analysis of 15 categorical traits also distinguished the lineages statistically, especially through anus position, body scaling, caudal-fin bars and the shape and width of dorsal bars. Yet some striking traits occurred in only one population of a genetic lineage, and populations assigned to the same lineage could differ noticeably. Altitude was associated with several conspicuous traits.

Highland lineages L1, L3 and L8 had larger average body sizes, generally possessed 8½ branched dorsal-fin rays and often had a rudimentary or distinct pelvic axillary lobe. The same lineages tended to have body scales that were distinct only behind the dorsal-fin origin. Lowland lineages more often had 7½ branched dorsal-fin rays, and most lacked the pelvic lobe.

Body-scalation patterns were not uniform across all lowland lineages: L2, L4, L6 and L7 showed inconsistent distributions of distinct scales. Standard length was positively correlated with elevation, while the number of body bars declined as fish became larger. The researchers suggested that reduced competition in species-poor mountain streams may contribute to larger body size, but the comparative study did not experimentally test that explanation.

The three recognized species remain valid. Genetic distances and re-examination of the rediscovered syntypes of S. notostigma support its separation from S. scripta and S. madhavai. S. scripta differed by more than 3.1% in uncorrected cox1 distance from the other lineages, including 6.0–6.9% from S. madhavai, while S. madhavai exceeded 4.8% from the other lineages.

The broader sampling also showed that both named species are more widely distributed than their original descriptions implied. In contrast, the apparent diagnostic value of some local features is limited: the distinctive body-bar and lateral-line patterns of S. scripta at its type locality were not consistently present in other populations assigned to the same genetic lineage. The type locality of S. notostigma remains uncertain because its original description gave only “Ceylon.”

The phylogeographic pattern suggests that Sri Lanka’s highlands were colonized more than once and may have lost older fish communities during dry climatic episodes. L3, the most widespread montane lineage, produced significantly negative results in both Tajima’s D and Fu and Li’s F neutrality tests, a pattern consistent with a recent expansion after population decline. Its presence in both the physiographically separate central mountains and Knuckles Massif strengthens the interpretation that the highlands were colonized recently, possibly after a climate-driven extinction as recent as the Last Glacial Maximum.

L1 and L8 appear to represent independent highland colonizations from southern basins, and L1 may instead be a relict that survived a previous highland desiccation. The southwestern lowlands contain more lineages in a smaller area, consistent with their role as refugia during dry periods. Overall, the study portrays a rapidly diversifying loach radiation whose evolutionary history is only partly visible in anatomy.

Sri Lankan Schistura apparently colonized the island relatively recently, spread through its moist southwestern quadrant to elevations of about 1,450 metres and separated into eight provisional mitochondrial lineages. Yet the traits traditionally used in loach taxonomy often failed to track those genetic boundaries, underscoring the risk of recognizing species from a single locality or a small series of specimens. The authors called for wider sampling and genome-scale nuclear markers, which could distinguish recent lineage sorting from historical gene flow and clarify how many species the eight groups ultimately represent.

The study itself is a research assessment rather than a formal revision of Sri Lankan Schistura taxonomy, so its MOTUs should not automatically be read as eight formally named species.

Subject of Research: Genetic and morphological diversification of Sri Lankan loaches in the genus Schistura

Article Title: The diversification of the loach genus Schistura McClelland, 1838 ( Teleostei , Nemacheilidae ) in Sri Lanka

Article References: Sudasinghe, H., Wijesooriya, K., Ranasinghe, T., Pethiyagoda, R., & Meegaskumbura, M. (2025). The diversification of the loach genus Schistura McClelland, 1838 (Teleostei, Nemacheilidae) in Sri Lanka. Zoosystematics and Evolution, 101(4), 2133-2159. https://doi.org/10.3897/zse.101.165370

Image Credits: AI Generated

DOI: 10.3897/zse.101.165370

Keywords: Schistura loaches, Sri Lanka, evolutionary lineages, mitochondrial DNA, nuclear DNA, phylogeography, cryptic diversity, freshwater fish, wet-zone rivers, highland colonization, diversification, loach

The specimen, measuring an impressive 2580 millimeters in total length, was caught by a surface tuna gillnet operated from a multiday fishing boat off the western coastal waters of Sri Lanka on 27 October 2021. The fish was landed at Beruwala Fishery Harbor and subsequently preserved in 5% formaldehyde before being deposited in the National Faunal Repository at the Colombo National Museum, Sri Lanka. This single, fresh, complete specimen now represents the first detailed morphological description of R. russellii from Sri Lankan waters.

Oarfishes, belonging to the family Regalecidae, are among the most extraordinary creatures in the ocean. Characterized by an extremely slender, laterally compressed, and very elongated ribbon-like body, they hold the distinction of being the longest bony fish in the world. The family is classified into just two genera, Regalecus and Agrostichthys, with three known species: Agrostichthys parkeri, Regalecus glesne, and Regalecus russellii. These oceanic fishes occur from surface waters to greater depths in temperate and tropical waters, yet they are exceptionally rare in nature, making every confirmed sighting scientifically significant.

The limited number of sightings, incomplete notes on encounters, scarce voucher specimens, and frequent misidentifications have all contributed to inadequate knowledge about the diversity, taxonomy, ecology, and distribution of oarfish species. This scarcity of information makes the Sri Lankan specimen particularly valuable, as it allows researchers to conduct a thorough morphological analysis that can be compared with confirmed records from other ocean basins.

The research team employed rigorous morphological methodologies to document the specimen. Measurements were collected to 1 millimeter precision using calibrated tape and Vernier calipers, while the specimen’s weight was measured to the nearest gram using a digital weighing balance. The fish was radiographed by dividing it into six overlapping sections using a static X-ray machine at the Faculty of Veterinary Medicine and Animal Science, University of Peradeniya. The X-radiograph images of all six sections were then processed to create a complete image, enabling researchers to count vertebrae and examine the dorsal fin ray arrangement.

The morphological examination revealed fascinating details about the specimen. The dorsal fin contained 337 total soft dorsal rays, with two distinctive dorsal crests: the first crest bearing five rays united by fin membranes and the second crest consisting of a single ray free from other rays. The body was long, deeply compressed, and ribbon-like, tapering towards the tail without scales but covered with dermal tubercles with circular bases. These tubercles became more prominent towards the tail end and lower sides of the body. The head was small, measuring just 9.1% of total length, with an almost vertical profile prior to the dorsal fin origin. The mouth was small, vertical, and protractible, with teeth entirely absent.

The coloration of the fresh specimen was equally striking. The body displayed a silvery-grey base with amorphous oblique dark streaks and blotches distributed across the body, most distinct in the first third. The forehead, mouth, preopercle, opercle, and lower margin of the head appeared greyish black. Perhaps most notably, the upper and lower body margins exhibited a crimson red luster, especially in the first half of the body, while the dorsal and ventral fins were entirely dark crimson red. When preserved, the body coloration shifted to yellowish brown with the same dark streaks and blotches.

The morphological characters of the Sri Lankan specimen matched well with key identification features of R. russellii previously reported from the eastern and western Pacific Ocean and the Atlantic Ocean. This consistency across ocean basins provides important confirmation of the species’ identity and extends its known range into the northern tropical Indian Ocean. The vertebral count of 114 and the gill raker arrangement on the first gill arch, numbering 55 in total with 12 on the upper portion and 43 on the lower, further supported the identification.

The significance of this record extends beyond simple range extension. Regalecus russellii was originally described as Gymnetrus russelii by Cuvier in 1816 from Vizagapatam, now Visakhapatnam, on the eastern coast of India in the Bay of Bengal. The first documentation of a Regalecus species from the Indian Ocean actually dates back to 1788, when Patrick Russell recorded a specimen from Visakhapatnam, though that juvenile specimen was not retained after drawing. Despite numerous scattered records of Regalecus from the Indian Ocean over the centuries, only two specimens had been identified to species level as R. russellii prior to this report, and no single specimen from the Indian Ocean has been accepted as a neotype for the species.

The two Regalecus species exhibit contrasting distribution patterns across the world’s oceans. In the Pacific Ocean, numerous records of R. russellii are documented from the North Pacific, while all records of R. glesne are concentrated in the South, with the exception of one R. russellii record from northern New Guinea. In the Atlantic Ocean, R. glesne has been documented exclusively in the North Atlantic and Mediterranean, while both species occur in the South Atlantic. The two species also show distinct habitat preferences related to water temperature and ocean currents: R. glesne is typically associated with colder waters, whereas R. russellii is commonly found in warmer regions. The Sri Lankan record, coming from the warm tropical waters of the northern Indian Ocean, aligns perfectly with these thermal preferences.

Interestingly, the spawning grounds of both Regalecus species occur at the extreme western end of the Indian Ocean, on the southern coast of South Africa, with R. glesne spawning also documented at the eastern end of the Indian Ocean on the coast of Western Australia. According to Roberts, the adults and early life stages of Regalecus found on the south coast of South Africa could be part of populations from the South Atlantic. The presence of a mature specimen in Sri Lankan waters raises intriguing questions about whether resident populations exist in the northern Indian Ocean or whether individuals occasionally venture far from known population centers.

The association of Regalecus with upwelling zones is particularly noteworthy. These areas feature surface cold water rich in biological productivity, which can serve as feeding grounds with optimum environmental conditions close to surface waters. The present specimen was found in shallow waters off the western coast of Sri Lanka, an area that overlaps with the coastal upwelling region. Numerous mesopelagic fish species, including members of the family Trachipteridae, have been recorded from the western and southern coastal upwelling zones of Sri Lankan waters during the summer monsoon, suggesting these productive areas attract deep-water species towards the surface.

Most records of Regalecus species are accidentally caught in fishing gear during operations and are often discarded without proper documentation. There have been many reports of oarfish landings by fishermen at various sites along the Indian and Sri Lankan coastal waters, but many of these accounts have not been published in peer-reviewed publications or made accessible to the scientific community. Advancing technology, along with the widespread accessibility of cameras and smartphone cameras, has significantly improved the frequency of recordings in recent years within the Indian Ocean region.

Citizen science programs and knowledge dissemination platforms, including social media, apps, and websites, play a crucial role in enhancing public awareness and encouraging the recording of rare or poorly known marine species. Such programs are essential to understanding marine biodiversity and supporting conservation strategies, particularly in understudied tropical waters. The Sri Lankan oarfish record demonstrates how collaboration between fishermen, researchers, and natural history institutions can yield significant scientific discoveries, even in an era of advanced marine research technology.

Seasonal patterns in Regalecus strandings and accidental catch records have been observed in certain areas, such as the Gulf of Mexico coast of Florida, though these events have not been clearly linked to environmental factors. In contrast, stranding and accidental catch reports from Australian waters have shown a clear association with storm events. The few records of Regalecus from the Indian Ocean remain inadequate to observe any patterns of stranding or accidental catch, but the species recorded are closely associated with upwelling zones of the Indian Ocean. As more records accumulate from this vast and understudied region, researchers may begin to identify environmental drivers behind these encounters.

This first confirmed record of Regalecus russellii from Sri Lankan waters represents a milestone in Indian Ocean marine biology. The detailed morphological description, complete with radiographic analysis and comprehensive meristic data, provides a reference point for future identifications in the region. The specimen, now safely preserved in the National Faunal Repository at the Colombo National Museum, stands as a testament to the scientific value of even a single well-documented specimen. As technology continues to advance and more researchers and citizens turn their attention to the ocean’s mysteries, the veil surrounding these magnificent ribbon-like giants of the deep may finally begin to lift, revealing insights into their biology, distribution, and the ecological roles they play in the vast tropical Indian Ocean.

Subject of Research: First record of oarfish, Regalecus russellii (Actinopterygii, Lampriformes, Regalecidae), from Sri Lankan waters, Indian Ocean

Article Title: First record of oarfish, Regalecus russellii ( Actinopterygii , Lampriformes , Regalecidae ), from Sri Lankan waters, Indian Ocean

Article References: Rathnasuriya, M. I. G., & Hapuarachchi, T. H. I. S. (2025). First record of oarfish, Regalecus russellii (Actinopterygii, Lampriformes, Regalecidae), from Sri Lankan waters, Indian Ocean. Acta Ichthyologica et Piscatoria, 55, 145-150. https://doi.org/10.3897/aiep.55.148496

Image Credits: AI Generated

DOI: 10.3897/aiep.55.148496

Keywords: oarfish, Regalecus russellii, Sri Lanka, Indian Ocean, marine biology, deep sea, rare species, fish taxonomy, upwelling zones, marine biodiversity, First, record

Tuberculous meningitis is the most severe form of central nervous system tuberculosis. Mycobacterium tuberculosis reaches the meninges through the bloodstream and sets up a chronic inflammatory process concentrated at the base of the brain, where the basal cisterns become filled with thick, gelatinous exudate. This basal meningitis can compress cranial nerves, obstruct cerebrospinal fluid flow and damage blood vessels, producing hydrocephalus, infarction and a wide range of neurological deficits. Even with modern anti-tuberculous drugs and corticosteroids, mortality and disability remain high, particularly when diagnosis is delayed.

Paradoxical reactions add a cruel twist to this already difficult disease. In a paradoxical reaction, a patient initially improves on anti-tuberculous therapy, then deteriorates neurologically despite the drugs continuing to kill the bacteria. The prevailing explanation is immunological: as anti-tuberculous drugs reduce the bacterial load, they also strip away antigens that had been dampening the immune response, allowing a rebound of cell-mediated immunity directed at residual mycobacterial antigens. Tumor necrosis factor-alpha and other inflammatory mediators surge, granulomatous inflammation flares, and lesions that had been shrinking can enlarge or appear at new sites. Paradoxical worsening is well documented in HIV-negative patients and in HIV-positive patients after starting antiretroviral therapy, a related phenomenon known as immune reconstitution inflammatory syndrome.

The Sri Lankan team, led by Shanindra de Alwis of the University of Sri Jayewardenepura with colleagues from the University of Colombo and the Institute of Neurology at the National Hospital of Sri Lanka, focused on one of the most destructive forms this rebound can take. Optochiasmatic arachnoiditis involves the optic nerves and the chiasmatic cisterns, the fluid-filled spaces surrounding the optic chiasm where fibers from both eyes cross. Inflammation and fibrosis in this region entrap the optic pathways, causing progressive, often irreversible loss of vision. Because the optic chiasm sits directly above the suprasellar cistern and within reach of the basal inflammatory exudate of tuberculous meningitis, it is uniquely vulnerable when paradoxical inflammation flares at the base of the brain.

All three patients in the series followed the same ominous trajectory. Each initially improved after starting anti-tuberculous therapy, consistent with the drugs doing their job. Then, during the course of treatment, they developed new neurological deterioration and radiological changes consistent with a paradoxical reaction. Neuroimaging revealed suprasellar tuberculomas, granulomatous masses in the region above the pituitary fossa, together with basal cisternal involvement affecting the optic pathways in every case. The pattern of optochiasmatic arachnoiditis was thus visible on magnetic resonance imaging as inflammatory disease encasing the structures that carry visual signals from the eyes to the brain.

Managing these patients proved extraordinarily difficult. Treatment required prolonged high-dose corticosteroid therapy to suppress the runaway inflammatory response, an approach that carries its own substantial risks when maintained for months. In selected cases, the team added adjunctive immunomodulatory therapy, including thalidomide, a drug with potent anti-tumor necrosis factor-alpha activity that has been used experimentally in severe central nervous system tuberculosis and in refractory paradoxical reactions. Thalidomide’s mechanism is relevant here: by inhibiting tumor necrosis factor-alpha, it aims to blunt the very cytokine driving the granulomatous flare, offering a targeted way to calm inflammation that corticosteroids alone could not control.

The outcomes were grim. Despite intensive immunosuppressive therapy, two patients died from complications related to prolonged immunosuppression. The surviving patient experienced persistent visual deficits, a lasting reminder that even when life is saved, the optic pathways damaged by optochiasmatic arachnoiditis may never fully recover. The authors emphasize that optochiasmatic arachnoiditis represents a rare but catastrophic paradoxical complication of tuberculous meningitis in HIV-negative patients, a population in whom paradoxical reactions are less commonly anticipated than in patients with HIV who begin antiretroviral therapy.

The series carries important lessons for clinicians worldwide. First, deterioration during anti-tuberculous therapy is not always treatment failure. Distinguishing paradoxical reactions from drug resistance, inadequate dosing or alternative diagnoses is essential, because the management strategies differ fundamentally: paradoxical worsening calls for intensified immunosuppression, not for changing anti-tuberculous drugs. Second, new visual symptoms in a patient being treated for tuberculous meningitis should prompt urgent neuroimaging of the suprasellar and chiasmatic regions, because early recognition of optochiasmatic arachnoiditis may allow intervention before irreversible fibrosis entombs the optic nerves. Third, the fatal complications of prolonged immunosuppression in two patients highlight the double-edged nature of the available therapies, which must suppress a destructive immune flare without leaving the patient defenseless against infection and other steroid-related harms.

The immunopathology underlying paradoxical reactions remains incompletely understood, and the authors argue that improved understanding is essential to guide targeted adjunctive therapies and improve outcomes. The interplay between mycobacterial antigen load, cell-mediated immunity, tumor necrosis factor-alpha and other cytokines determines whether residual disease smolders quietly or erupts into granulomatous inflammation around the optic chiasm. Better biomarkers to predict which patients will flare, and immunomodulatory agents that can be deployed with precision rather than through prolonged blunt immunosuppression, are the goals the field must pursue. For now, this Sri Lankan case series stands as a stark warning: the very success of anti-tuberculous therapy can ignite an inflammatory backlash capable of stealing a patient’s sight, and in the most severe cases, their life.

Published as an open access case report in BMC Infectious Diseases, the study adds to the growing recognition that central nervous system tuberculosis is not merely an infectious disease but an immunological battleground. As tuberculosis continues to affect millions globally, and as drug-resistant strains prolong treatment courses and expand the window for paradoxical events, awareness of optochiasmatic arachnoiditis among neurologists, infectious disease specialists and ophthalmologists could make the difference between timely intervention and permanent blindness. The Sri Lankan experience shows that vigilance during the treatment of tuberculous meningitis must extend beyond bacterial cure to the immune consequences of that cure itself.

Subject of Research: Health & Medicine

Article Title: Optochiasmatic arachnoiditis as a paradoxical reaction in tuberculous meningitis: a case series from Sri Lanka

Article References: de Alwis, S., Chandrasiri, D., Ruwanpathirana, P., Fernando, A., & Senanayake, B. (2026). Optochiasmatic arachnoiditis as a paradoxical reaction in tuberculous meningitis: a case series from Sri Lanka. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14471-4

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14471-4

Keywords: tuberculous meningitis, optochiasmatic arachnoiditis, paradoxical reaction, central nervous system tuberculosis, thalidomide, corticosteroids, Sri Lanka, BMC Infectious Diseases, Optochiasmatic, arachnoiditis, paradoxical, reaction

The research team, led by scientists surveying more than one hundred forest localities across the island, assembled the first phylogeny of Sri Lankan goblin spiders. The family Oonopidae currently includes over 1,978 described species in 115 genera worldwide, and Sri Lanka harbours 45 recognised species across 13 genera, of which 38 are found nowhere else on Earth. To understand how these species relate to one another and to their global relatives, the researchers extracted DNA from ethanol-preserved leg tissue and amplified two nuclear ribosomal markers, 18S and 28S rRNA, the same genetic workhorses used in the earlier global phylogeny of the family published by Busschere and colleagues in 2014.

The fieldwork itself was a painstaking exercise in micro-hunting. Specimens were collected from Sri Lanka’s major climatic and physiographic zones, including the Central Mountain Range, the Knuckles Range, the low country wet zone and the dry zone. Leaf litter was sifted and left overnight in Winkler extractors, devices that coax out the small arthropods hiding within decaying vegetation. Preserved specimens were examined under stereomicroscopes and identified to genus using recent international revisionary works. The resulting dataset, comprising 43 Sri Lankan taxa, was then integrated into a global framework of 154 taxa representing 40 genera, with outgroup species from the families Dysderidae, Orsolobidae and Segestriidae providing evolutionary context.

The laboratory protocol demanded considerable ingenuity. Initial attempts to amplify fragments of 16S, CO1 and H3 genes were only partially successful, so the researchers settled on four primer pairs targeting three overlapping regions of 18S and a partial fragment of 28S. Polymerase chain reactions were run with carefully calibrated annealing temperatures, and older specimens required a specialised multiplex PCR kit to recover usable DNA. Purified products were Sanger sequenced in both directions, and the final concatenated alignment stretched to 2,782 base pairs. Maximum likelihood searches in IQ-TREE, using the best-fit GTR+F+I+G4 model of nucleotide substitution and 1,000 ultrafast bootstrap replicates, produced the phylogenetic trees at the heart of the study.

The headline finding is that the family Oonopidae is monophyletic, meaning all goblin spiders descend from a single common ancestor, which confirms earlier molecular work. But within that framework, several familiar genera turned out to be evolutionary patchworks rather than natural groups. The genera Brignolia, Opopaea, Aschnaoonops, Paradysderina, Unicorn and Silhouettella were all recovered as paraphyletic in at least some analyses, meaning their members are scattered across the tree rather than clustered together. This has direct taxonomic consequences: these genera, defined largely by morphological features such as the dorsal depression on the male palp, will need to be redefined or split.

The Brignolia and Opopaea story is particularly tangled. The study recovered a well-supported clade containing both genera, but neither was monophyletic within it. Brignolia ondaatjei emerged as the sister species of the Sri Lankan Opopaea spinosiscorona, while other Brignolia species grouped in distinct subclusters. Intriguingly, all Sri Lankan Brignolia and Opopaea species together were sister to Opopaea apicalis from the Galapagos, yet the Sri Lankan Opopaea formed a group entirely separate from their Australian, South American and African namesakes. The researchers conclude that both genera require morphological redefinition to match their evolutionary reality.

Not all results were disruptive. The genus Xestaspis, included in a goblin spider phylogeny for the first time, formed a well-supported clade of four Sri Lankan species, sister to the Australian Prethopalpus tropicus. Orchestina, the soft-bodied canopy jumpers with enlarged fourth femora, was confirmed as monophyletic, and the three Sri Lankan species formed a distinct clade sister to Orchestina saaristoi from the Congo. The genera Pelicinus and Escaphiella also emerged as natural groups. Meanwhile, the finding that most soft-bodied genera sit at the base of the oonopid tree supports the hypothesis that reduced body sclerotisation, the degree to which the body is hardened by plates, is an ancestral trait in the family, with heavily armoured species representing a derived condition.

Perhaps the most striking pattern is the sheer scale of microendemism. Of the 45 Sri Lankan species, 38 are endemic, and most are known from only one, two or three sites among the more than hundred localities surveyed. Brignolia ondaatjei is known only from two sites in the Badulla District at elevations between 990 and 1,280 metres. Xestaspis nuwaraeliya appears only above 1,700 metres in the Nuwara Eliya District. Aprusia vankhedei is restricted to two secondary forest patches within Kandy City itself. Some species, such as Aprusia rawanaellensis and Brignolia meemure, are known solely from their type localities. The phylogeny shows that endemic clades such as Aprusia, Brignolia and Xestaspis consist of closely related assemblages of multiple endemic species, strong evidence of recent, within-island speciation rather than repeated colonisation from the mainland.

The Sri Lankan taxa fall into five distinct clades within the global tree, reflecting a combination of multiple colonisation events and localised radiations. The researchers suggest that both ecological isolation and in situ speciation have driven this diversity, though they caution that it is premature to draw firm conclusions about origins because the Indian fauna remains poorly sampled. Goblin spiders, as short-range endemics with highly restricted distributions, may also serve as important indicators for monitoring the effects of climate change, agroforestry, urbanisation and pollution on forest habitats, since species confined to a single forest patch have nowhere to go when that patch degrades.

The study is a reminder that biodiversity hotspots operate at every scale, from charismatic megafauna down to spiders the size of a sesame seed hiding in the leaf litter. It also underscores the unfinished business of taxonomy: molecular trees are increasingly showing that genera erected on the basis of a handful of physical traits do not always correspond to true evolutionary lineages. With broader sampling of Indian and Southeast Asian taxa, and additional molecular markers, the researchers hope to fully resolve the origins and diversification of this ecologically important spider family. For now, Sri Lanka’s goblin spiders stand as a vivid example of how a small island, over deep time, can generate an extraordinary wealth of life in the smallest of packages.

Subject of Research: Molecular phylogenetics and microendemism of Sri Lankan goblin spiders (Oonopidae)

Article Title: Phylogenetic placement of Sri Lankan goblin spiders ( Araneae , Oonopidae ): integrating new taxa and taxonomic implications

Article References: Ranasinghe, U. G. S. L., & Benjamin, S. P. (2025). Phylogenetic placement of Sri Lankan goblin spiders (Araneae, Oonopidae): integrating new taxa and taxonomic implications. Evolutionary Systematics, 9(2), 259-270. https://doi.org/10.3897/evolsyst.9.168533

Image Credits: AI Generated

DOI: 10.3897/evolsyst.9.168533

Keywords: goblin spiders, Oonopidae, Sri Lanka, phylogenetics, microendemism, 18S rRNA, 28S rRNA, taxonomy, Brignolia, Opopaea, Orchestina, speciation

The stakes could hardly be higher. Flooding has caused average annual losses of US$240 million in Sri Lanka and affected more than 10 million people over the past three decades, making it the most frequent and destructive disaster among the 25,000 extreme weather events recorded between 1975 and 2022. Germanwatch’s Climate Risk Index once ranked the island nation fourth among the ten most affected countries globally. Yet the traditional monitoring systems still in use rely on manual observations and static threshold alarms that offer no predictive analytics at all. When a river crests faster than an observer can file a report, the warning arrives late—or never.

The research team attacked this gap with a deliberately safety-oriented design philosophy. Rather than chasing headline accuracy, they optimized their models for recall: the proportion of genuine flood conditions the system manages to catch. In flood forecasting, the two types of error carry wildly asymmetric costs. A false alarm wastes resources and erodes trust, but a missed flood—a false negative—can delay evacuations and cost lives. The team therefore built their ensemble around a simple logical-OR rule: if either of the two models flags a positive flood risk, the system sounds the warning. Only when both models agree the river is safe does the dashboard stay quiet.

The two voters bring complementary architectures to the problem. The first is a Long Short-Term Memory network, a form of recurrent neural network whose internal memory gates allow it to learn the sequence and delay between rainfall and the resulting river response. The LSTM was configured to forecast the next day’s numerical water level at the Kaliodai River using five predictor variables: catchment rainfall, downstream rainfall, reservoir water level, reservoir discharge rate, and the current river stage. Its stacked layers of 64 and 32 hidden units were trained with the Adam optimizer on roughly five years of daily observations—1,873 entries spanning March 2020 to April 2025—obtained from Sri Lanka’s Department of Irrigation and Department of Meteorology.

The second model, Extreme Gradient Boosting, or XGBoost, is an ensemble of decision trees built sequentially, where each new tree corrects the errors of its predecessors. Instead of forecasting a number, XGBoost classifies the next day’s flood risk directly as positive or negative. Because genuine flood days are vanishingly rare in the historical record, the training data was severely imbalanced, and the team applied the Synthetic Minority Oversampling Technique to generate synthetic examples of the minority flood class for the XGBoost classifier. A gain-based feature importance analysis revealed that the current Kaliodai River water level dominated the classifier’s decisions, followed by the reservoir water level, while reservoir discharge contributed almost nothing—its information apparently already embedded in the correlated river stage.

Labeling the data required an empirical threshold. The team examined official flood-warning records for the Gal Oya Basin and found that documented flood conditions corresponded to river levels of at least 3.0 meters, a value reviewed by a civil engineer familiar with local conditions. Every daily observation at or above that mark was labeled positive. The historical record contained just nine such positive observations, forming two continuous flood episodes: January 11 to 15, 2024, and November 25 to 28, 2024. The models were trained on 80 percent of the data using a strict chronological split, with no random shuffling, so that future observations never leaked into training.

The evaluation results tell a nuanced story. On the chronological holdout, the rule-based ensemble with a one-day input window achieved 98.73 percent overall accuracy and, critically, 100 percent positive-class recall—correctly classifying all four positive-labeled observations in the test set, which occurred consecutively from November 25 to 28, 2024, representing one continuous flood episode. Positive-class precision was 50.00 percent, reflecting four false alarms, and the F1-score was 66.67 percent. The team calculated exact binomial confidence intervals to quantify the uncertainty: the 95 percent interval for recall spanned 39.8 to 100 percent, a sobering reminder that four samples from a single episode cannot establish population-level performance. The authors are explicit that the perfect recall is a sample-level result, not proof of generalized flood detection across independent events.

A benchmark comparison sharpened the interpretation. A simple current-water-level threshold—predicting a flood whenever the river already stood at or above 3.0 meters—achieved 99.37 percent accuracy, 75 percent precision, and 75 percent recall on the same test set. The benchmark missed the flood onset on November 25, 2024, when the previous day’s level was 2.9 meters, just below the threshold. The ensemble caught that rising episode, suggesting that the models’ additional hydrological information and the safety-first fusion rule provided genuine sensitivity to the onset of flood conditions, at the cost of extra false alarms. Notably, the LSTM alone achieved 98.72 percent accuracy with a 21-day window but zero recall for flood cases—a vivid demonstration that accuracy alone is a treacherous metric for imbalanced safety-critical datasets.

The field deployment demonstrated operational feasibility rather than predictive validation. The IoT node pairs a JSN SR-04T waterproof ultrasonic sensor, which measures river level from ultrasonic pulse travel time, with an XS-RS01 tipping-bucket rain gauge whose reed-switch tips each correspond to 0.3 millimeters of rainfall. An ESP32 microcontroller aggregates the readings and transmits them via NB-IoT or Wi-Fi to a ThingsBoard cloud platform, powered by a 12-volt solar panel and rechargeable battery for autonomous off-grid operation. During the ten-day observation window, the system correctly classified all days as negative, consistent with real conditions, and the LSTM’s water-level forecasts tracked observed rises during heavy rainfall, with a percentage error of 10.77 percent on November 29, 2025. The dashboard even captured dam spilling effects downstream during dry days.

The authors caution that no flood occurred during the deployment period, so flood-event detection under real conditions remains unproven, and future work must include longer-term field validation, sequence-aware imbalance handling for recurrent networks, rolling-origin time-series validation, and explainable AI techniques such as Shapley additive exPlanations. Still, the framework offers a template for smaller nations with abundant rivers and limited monitoring infrastructure: a localized, automated, scalable system that can be centrally monitored and expanded to other basins. As climate change intensifies extreme rainfall across South Asia, the marriage of cheap solar-powered sensors with safety-first machine learning may prove one of the most consequential tools in the flood-warning arsenal.

Subject of Research: AI flood forecasting

Article Title: A recall-optimized LSTM–XGBoost ensemble model for river flood forecasting using multivariate time-series data—case study in Sri Lanka

Article References: Sankalpa, T. H. S., Sandaruwan, G. H. C., Saja, A. M. A., Sajith, N. T. M., & Ishqy, M. F. M. (2026). A recall-optimized LSTM–XGBoost ensemble model for river flood forecasting using multivariate time-series data—case study in Sri Lanka. Frontiers in Artificial Intelligence, 9, Article 1903150. https://doi.org/10.3389/frai.2026.1903150

Image Credits: AI Generated

DOI: 10.3389/frai.2026.1903150

Keywords: flood forecasting, machine learning, LSTM, XGBoost, Internet of Things, Sri Lanka, early warning systems, hydrology, recall-optimized, ensemble, model, river