Tag: statistics

Ecotoxicol Environ Saf. 2025 Dec 18;309:119587. doi: 10.1016/j.ecoenv.2025.119587. Online ahead of print.

ABSTRACT

Environmental remediation including mercury (Hg) source closure, cage-farming prohibition, and sediment capping has been implemented at Baihua Reservoir (BHR) since 2007, which strongly influences Hg bioaccumulation in fish. In this study, total Hg (THg) and methylmercury (MeHg) concentrations were determined in abiotic matrices and fish in 2015-2016 and 2018-2019 and compared with data extracted from previous studies referring to 2002-2006 and 2008-2009. Following these remediation measures, the water quality improved, eutrophication declined, and environmental Hg levels dropped dramatically. By 2018-2019, the aqueous THg and MeHg decreased by 82 % and 89 %, and sediment THg and MeHg decreased by 99 % and 86 %, respectively, compared to 2002-2006 levels. However, prohibition of cage fish farming in BHR shifted fish diets toward plankton, slowing their growth rate and lengthening the food chain by one level. Paradoxically, average fish THg during 2008-2019 increased significantly by 73-168 %, whereas the increase in MeHg (33-137 %) was not statistically significant. The relative proportion of MeHg to THg (%MeHg) in fish transiently peaked at 56 ± 23 % in 2015-2016. The bioaccumulation factor for fish Hg was high, while the trophic magnification rates of Hg were lower and remained low compared to global averages, reflecting efficient initial uptake into the food web but constrained transfer efficiency. This unexpected rise in fish Hg is likely due to the dietary shift to Hg-rich plankton, reduced somatic growth dilution in wild fish, and enhanced Hg trophic transfer. Therefore, ongoing monitoring and updated fish consumption advisories are recommended to mitigate potential health risks.

PMID:41418362 | DOI:10.1016/j.ecoenv.2025.119587

Biomed Phys Eng Express. 2025 Dec 19. doi: 10.1088/2057-1976/ae2f65. Online ahead of print.

ABSTRACT

OBJECTIVE: To classify digital mammograms based on radiological findings using morphology and texture descriptors with artificial neural networks (ANN) for breast cancer detection.

APPROACH: The mammography dataset from High Specialty Regional Hospital of Oaxaca (HRAEO) (median patient age (mpa), 48 years [interquartile range (IQR), 41-54 years]) with radiological findings was retrospectively analyzed. All patients underwent breast biopsy and were not previously treated. External testing was performed using mammograms from the National Cancer Institute (INCAN) (mpa: 47 years [IQR, 37-62 years]). The morphology was analyzed using a circularity descriptor (), and the texture was analyzed using the mean height/width ratio of the extrema descriptor (). These results were compared with cancer/benign histopathology, which was binarily classified using ANNs. The F1-score, Cohen’s kappa (K), and area under the ROC curve (AUC) were employed as evaluation metrics, and the Wilcoxon rank-sum test was used for statistical analysis (h = 0, with p > 0.05, was considered as not statistically significant).

MAIN RESULTS: 216 raw mammograms from HRAEO and 33 mammograms from INCAN (95+16 breast cancer and 121+17 benign findings) were included. The best internal testing results were obtained with a one-hidden-layer ANN with 100 neurons, achieving a F1-score of 0.95, K of 0.91, and an AUC of 0.953 (95% confidence interval [CI]: 0.917, 0.977) (h=0, p>0.99). However, the external testing results were significantly lower: 0.38 F1-score, 0.02 K, and 0.509 AUC (95% CI: 0.344, 0.664) (h=0, p=0.14) due to not exactly meeting the inclusion criteria and possible demographic and spectrum bias, or domain-adaptation issues.

SIGNIFICANCE: The proposed morphology () and texture () descriptors show promise for detecting breast cancer in raw mammograms, with radiological findings, in a local context. However, their poor external performance highlights the need for substantial further work before this approach can be deemed suitable for broader diagnostic applications.

PMID:41418324 | DOI:10.1088/2057-1976/ae2f65

JMIR Form Res. 2025 Dec 19;9:e75607. doi: 10.2196/75607.

ABSTRACT

BACKGROUND: Multiple-choice questions (MCQs) are essential in medical education for assessing knowledge and clinical reasoning. Traditional MCQ development involves expert reviews and revisions, which can be time-consuming and subject to bias. Large language models (LLMs) have emerged as potential tools for evaluating MCQ accuracy and efficiency. However, direct comparisons of these models in orthopedic MCQ assessments are limited.

OBJECTIVE: This study compared the performance of ChatGPT and DeepSeek in terms of correctness, response time, and reliability when answering MCQs from an orthopedic examination for medical students.

METHODS: This cross-sectional study included 209 orthopedic MCQs from summative assessments during the 2023-2024 academic year. ChatGPT (including the “Reason” function) and DeepSeek (including the “DeepThink” function) were used to identify the correct answers. Correctness and response times were recorded and compared using a χ2 test and Mann-Whitney U test where appropriate. The two LLMs’ reliability was assessed using the Cohen κ coefficient. The MCQs incorrectly answered by both models were reviewed by orthopedic faculty to identify ambiguities or content issues.

RESULTS: ChatGPT achieved a correctness rate of 80.38% (168/209), while DeepSeek achieved 74.2% (155/209; P=.04). ChatGPT’s Reason function also outperformed DeepSeek’s DeepThink function (177/209, 84.7% vs 168/209, 80.4%; P=.12). The average response time for ChatGPT was 10.40 (SD 13.29) seconds, significantly shorter than DeepSeek’s 34.42 (SD 25.48) seconds (P<.001). Regarding reliability, ChatGPT demonstrated an almost perfect agreement (κ=0.81), whereas DeepSeek showed substantial agreement (κ=0.78). A completely false response was recorded in 7.7% (16/209) of responses for both models.

CONCLUSIONS: ChatGPT outperformed DeepSeek in correctness and response time, demonstrating its efficiency in evaluating orthopedic MCQs. This high reliability suggests its potential for integration into medical assessments. However, our results indicate that some MCQs will require revisions by instructors to improve their clarity. Further studies are needed to evaluate the role of artificial intelligence in other disciplines and to validate other LLMs.

PMID:41418321 | DOI:10.2196/75607

J Med Internet Res. 2025 Dec 19;27:e75337. doi: 10.2196/75337.

ABSTRACT

BACKGROUND: The link between social media use and depressive symptoms remains bidirectional. Findings in this area are often compromised by methodological limitations related to measurement and sample size. As a result, it is challenging to assess dose-response relationships and potential causal pathways.

OBJECTIVE: This study aims to use objective measurement methods to assess the dose-response relationship and potential mechanisms between social media use and depressive symptoms.

METHODS: This study was conducted in 6 universities in 2022. Social media use duration was assessed based on the monitoring of mobile phone systems, and depressive symptoms were evaluated by the Self-Rating Depression Scale. Logistic regression and restricted cubic spline were used to assess the relationship between social media use and depressive symptoms. Mediation analysis was used to elucidate the biological pathways of sleep quality in the abovementioned relationship.

RESULTS: A total of 7401 college students were included in the final analysis, with 4.93% of moderate to severe depressive symptoms. After adjusting for variables such as sociodemographic characteristics and health-related characteristics, there was a significant association between individuals with longer weekly social media use time and depressive symptoms (odds ratio [OR>48h] 1.769, 95% CI 1.303-2.400). Similarly, the association between instant messaging-based social media use duration and depressive symptoms was also significant (OR>24h 1.728, 95% CI 1.225-2.437), while no associations were observed for content-based social use (OR>24h 1.251, 95% CI 0.932-1.680). Restricted cubic splines regression demonstrated a J-type relationship between social media use duration and depressive symptoms. Additionally, sleep quality played a partial mediating role in the relationship between social media use duration and depressive symptoms, with the mediating effect values ranging from 24.10% to 25.25%.

CONCLUSIONS: Prolonged social media use duration might be associated with an increased prevalence of depressive symptoms in emerging adulthood and may increase the odds of depression by affecting sleep quality, suggesting that early prevention and intervention regarding social media use might help to ameliorate depressive symptoms.

PMID:41418287 | DOI:10.2196/75337

Ophthalmic Plast Reconstr Surg. 2025 Dec 19. doi: 10.1097/IOP.0000000000003146. Online ahead of print.

ABSTRACT

PURPOSE: To quantitatively compare the degrees of surgical freedom achieved by a lateral orbitotomy without bone removal versus a lacrimal keyhole-assisted approach, in which the superolateral orbital rim is shaved without full removal.

METHODS: Preoperative orbital computed tomography scans from consecutive lateral orbitotomy patients (January 2023-April 2025) were retrospectively analyzed. Using Brainlab software, we compared the theoretical volume of surgical freedom and horizontal and vertical attack angles between a simulated orbitotomy without bone removal and the keyhole-assisted technique with incremental rim shaving (25% to 100%). Statistical analysis included paired t tests and linear regression.

RESULTS: Twelve patients (mean age: 52.2 years; 58% female) were evaluated. The complete keyhole-assisted approach significantly increased the volume of surgical freedom by 21% (6.2 ± 2.2 mm³ vs. 5.1 ± 2.0 mm³; p = 0.005) and the horizontal and vertical attack angles by 62% (20.4 ± 7.2° vs. 12.6 ± 4.2°; p < 0.001) and 38% (36.4 ± 9.5° vs. 26.4 ± 8.4°; p < 0.001), respectively. Partial rim shaving also yielded significant, progressive gains (p < 0.05). Linear regression showed that each 1 mm of bone shaved improved the horizontal angle by 16%, the vertical angle by 10%, and the volume of surgical freedom by 6%. The size of the rim prominence strongly correlated with surgical freedom (r = 0.74, p = 0.02). No postoperative cosmetic or functional deficits were observed.

CONCLUSIONS: The lacrimal keyhole-assisted orbitotomy significantly enhances surgical freedom for orbital lesions. This minimally invasive technique offers a titratable alternative to marginotomy, allowing tailored bone removal for required maneuverability while preserving orbital rim integrity.

PMID:41418283 | DOI:10.1097/IOP.0000000000003146

JMIR Mhealth Uhealth. 2025 Dec 19;13:e68438. doi: 10.2196/68438.

ABSTRACT

BACKGROUND: Mobile apps are powerful tools for promoting and sustaining healthy behaviors, including supporting diverse recovery pathways from substance use, including alcohol use disorder. Indeed, prior research strongly supports the notion that social connection through mobile platforms, supplemented by an in-person interaction, is vital in helping individuals strengthen their recovery and improve overall well-being and mental health. However, research into the digital footprints of mobile app users, as a strategy to assess app usage experiences in a recovery context, is lacking.

OBJECTIVE: This study utilizes a dataset from The Phoenix app, a social media platform specifically designed for individuals impacted by substance use, including those in or seeking recovery, to identify core uses of the app, including how it is leveraged by members from a thematic and emotional valence context.

METHODS: We applied natural language processing and deep learning methods to analyze a random sample of 19,685 posts. Analyses included the Bidirectional Encoder Representation from Transformers topic modeling tool to generate themes and a Valence Aware Dictionary and Sentiment Reasoner sentiment analysis to approximate emotional tone and mood from posts ranging from highly negative (-0.99) to highly positive (0.99).

RESULTS: After removing duplicate and nonsensical posts, we retained a final sample size of 17,617 posts. Bidirectional Encoder Representation from Transformers topic modeling tool identified 10 topics (coherence score=0.48) within 2 overarching themes: (1) those related to engaging app members through in-person and online interactions (7 topics) and (2) as a forum to discuss more serious topics pertaining to substance use and mental health recovery (3 topics). Overall, the topics revealed a distinct and recurring theme of community support. Valence Aware Dictionary and Sentiment Reasoner sentiment analysis was 0.44 (SD 0.42), indicating highly positive posts, with only 429 (2.4%) being highly negative.

CONCLUSIONS: The study findings broadly show positive uses of The Phoenix app as a tool for social connections and community among people in recovery from substance use. With the high positive sentiment of posts, the app was distinct from other social media platforms (eg, X, Reddit, Facebook), which often feature a mix of highly positive and highly negative posts. Additional research is needed to confirm these results using a larger dataset and with comparative analysis of other recovery forums to contribute to the understanding of social media’s role and function in changing health-related behaviors.

PMID:41418282 | DOI:10.2196/68438

J Pak Med Assoc. 2025 Dec;75(12):1998-20003. doi: 10.47391/JPMA.25-20593.

ABSTRACT

OBJECTIVE: To determine the association between attention span and the academic year among medical students, and to identify the factors affecting the attention span.

METHODS: The cross-sectional study at the Foundation University Medical College, Islamabad, Pakistan, from July to August 2022, and comprised medical students from 1st to 5th academic year. Data was collected using the online Attention Control Scale questionnaire. Data was analysed using SPSS 26.

RESULTS: Of the 285 subjects with mean age 21.1+1.68 years, 195(68.4%) were females and 90(31.6%) were males. The mean score among the males was 50.31±6.37 compared to 48.12±7.12 for the females. Third-year students had the lowest mean score 46.67±7.02, while fourth-year students scored the highest 49.98±6.67. Significantly influenced internal factors were emotions about 154(54.7%), while longer courses 196(68.8%) and nearby noises 118 (41.4%) were additional factors. Finally, 204(71.6%) students struggled while listening to a one-hour class, and 148 (52.28%) preferred problem-based learning and small group discussions as teaching methods.

CONCLUSION: Traditional lectures and long group discussions affected students’ attention span negatively, with external factors, like course length and loud surroundings, and internal factors, like emotions and learner type, having an influence.

PMID:41418262 | DOI:10.47391/JPMA.25-20593

J Pak Med Assoc. 2025 Dec;75(12):1914-1917. doi: 10.47391/JPMA.25-21836.

ABSTRACT

OBJECTIVE: To compare the outcomes of Fenestrating and Reconstituting subtypes of laparoscopic subtotal cholecystectomy in technically difficult cases.

METHODS: The retrospective study was conducted at the Department of Surgery, Khyber Teaching Hospital, Peshawar, Pakistan, from August 1, 2023, to July 31, 2024, and comprised data of patients who underwent Fenestrating or Reconstituting laparoscopic subtotal cholecystectomy between January 1, 2014, and December 31, 2023, for gallstone disease. Postoperative length of hospital stay, frequency of bile leak, bile collections, requirement of endoscopic retrograde cholangiopancreatography and reoperations were the parameters noted and compared. Data was analysed using IBM® SPSS® Statistics 26.0.

RESULTS: Of 1,579 patients, 78(4.9%) had undergone laparoscopic subtotal cholecystectomy. Of them, 6(7.7%) cases were i excluded due to incomplete records, and the final sample comprised 72(92.3%) patients; 44(61.2%) females and 28(38.8%) males with mean age 56.91±16.28 years. Overall, 43(59.7%) patients were in the Fenestrating group and 29(40.3%) in the Reconstituting group. Mean postoperative hospital stay was significantly different between the groups (p<0.05), while the frequency of bile leak, bile collection, requirement of endoscopic retrograde cholangiopancreatography and the rate of reoperations were not significantly different between the groups (p>0.05).

CONCLUSIONS: Compared to the Fenestrating type, the Reconstituting type of laparoscopic subtotal cholecystectomy resulted in reduced postoperative hospital stay.

PMID:41418240 | DOI:10.47391/JPMA.25-21836

J Pak Med Assoc. 2025 Dec;75(12):1898-1902. doi: 10.47391/JPMA.25-21963.

ABSTRACT

OBJECTIVES: To assess the risk factors regarding early cessation of breastfeeding.

METHODS: The cross-sectional study was conducted at the Paediatrics Department of Ayub Medical Teaching Institute, Abbottabad, Pakistan, from June 2022 to February 2023, and comprised mother-child dyads who initiated the breastfeeding process at birth, but failed to continue over the subsequent 24 months. Data was collected using a structured proforma through interviews with mothers. Data was analysed using SPSS 26.

RESULTS: There were 250 mothers with mean age 28.48±5.73 years, while the mean age of 250 children was 14.64±7.53 months. The mean duration of breastfeeding was 5.05±3.99 months. The major risk factors identified were a combination of breastfeed and bottle-feed in the first month of life in 198(79.2%) cases, maternal diet restriction postpartum 186(74.4%) and no counselling about breastfeeding 168(67.2%). Pacifier use (p= 0.002), combined bottle-feed and breastfeed in the first month of life (p=0.001) and mothers wrongly considering formula milk healthier than breastfeed (p=0.000) were significant predictors of shorter duration of breastfeeding.

CONCLUSIONS: All the reasons identified as potential risk factors for early discontinuation of breastfeeding were modifiable. Health education-based interventions are needed for optimising breastfeeding practices.

PMID:41418237 | DOI:10.47391/JPMA.25-21963

J Pak Med Assoc. 2025 Dec;75(12):1883-1890. doi: 10.47391/JPMA.25-23187.

ABSTRACT

OBJECTIVES: To determine the correlation of the Glasgow Coma Scale score and different Neuroimaging Radiological Interpretation System categories in patients with traumatic brain injury.

METHODS: The prospective observational study was conducted at the Shaheed Mohtarma Benazir Bhutto Institute of Trauma, Karachi, from July to December 2023, and comprised patients aged at least 18 years with documented traumatic brain injury. The patient underwent assessment in the emergency department. The mechanism of injury, and Glasgow Coma Scale score at the time of hospital arrival were noted. A non-contrast head computed tomography scan was done to acquire images of sufficient quality for Neuroimaging Radiological Interpretation System classification. Data was analysed using SPSS 23.

RESULTS: Of the 388 patients with mean age 33.44±18.51 years, 319(82.2%) were males and 69(17.8%) were females. Road traffic accidents accounted for 283(72.9%) cases. Glasgow Coma Scale-based severity classification revealed mild traumatic brain injury in 185(47.7%), moderate in 100(25.8%) and severe in 103(26.5%) patients. Neuroimaging Radiological Interpretation System category distribution showed category 2 being the most frequent 165(42.5%), while category 0 was least common 5(1.3%). There was a strong negative correlation between Glasgow Coma Scale scores and Neuroimaging Radiological Interpretation System categories (p=0.001), with the coefficient of determination indicating that 27.4% of Glasgow Coma Scale score variance was explained by Neuroimaging Radiological Interpretation System classification.

CONCLUSIONS: There was a negative correlation between Neuroimaging Radiological Interpretation System category and Glasgow Coma Scale score in patients with traumatic brain injury.

PMID:41418235 | DOI:10.47391/JPMA.25-23187