Tag: statistics

Spine J. 2021 Nov 1:S1529-9430(21)00980-3. doi: 10.1016/j.spinee.2021.10.020. Online ahead of print.

ABSTRACT

BACKGROUND CONTEXT: Computer-aided diagnosis with artificial intelligence (AI) has been used clinically, and ground truth generalizability is important for AI performance in medical image analyses. The AI model was trained on one specific group of older adults (aged≧60) has not yet been shown to work equally well in a younger adult group (aged 18-59).

PURPOSE: To compare the performance of the developed AI model with ensemble method trained with the ground truth for those aged 60 years or older in identifying vertebral fractures (VFs) on plain lateral radiographs of spine (PLRS) between younger and older adult populations.

STUDY DESIGN/SETTING: Retrospective analysis of PLRS in a single medical institution PATIENT SAMPLE OUTCOME MEASURES: Accuracy, sensitivity, specificity, and interobserver reliability (kappa value) were used to compare diagnostic performance of the AI model and subspecialists’ consensus between the two groups.

METHODS: Between January 2016 and December 2018, the ground truth of 941 patients (one PLRS per person) aged 60 years and older with 1101 VFs and 6358 normal vertebrae was used to set up the AI model. The framework of the developed AI model includes: object detection with You Only Look Once Version 3 (YOLOv3) at T0-L5 levels in the PLRS, data pre-preprocessing with image-size and quality processing, and AI ensemble model (ResNet34, DenseNet121, and DenseNet201) for identifying or grading VFs. The reported overall accuracy, sensitivity and specificity were 92%, 91% and 93%, respectively, and external validation was also performed. Thereafter, patients diagnosed as VFs and treated in our institution during October 2019 to August 2020 were the study group regardless of age. In total, 258 patients (339 VFs and 1725 normal vertebrae) in the older adult population (mean age 78 ±10.4; range, 60-106) were enrolled. In the younger adult population (mean age 36 ± 9.43; range, 20-49), 106 patients (120 VFs and 728 normal vertebrae) were enrolled. After identification and grading of VFs based on the Genant method with consensus between two subspecialists’, VFs in each PLRS with human labels were defined as the testing dataset. The corresponding CT or MRI scan was used for labeling in the PLRS. The bootstrap method was applied to the testing dataset.

RESULTS: The model for clinical application, Digital Imaging and Communications in Medicine (DICOM) format, is uploaded directly (available at: http://140.113.114.104/vght_demo/svf-model (grading) and http://14 0.113.114.104/vght demo/svf-model2 (labeling). Overall accuracy, sensitivity and specificity in the older adult population were 93.36% (95% CI 93.34%-93.38%), 88.97% (95% CI 88.59%-88.99%) and 94.26% (95% CI 94.23%-94.29%), respectively. Overall accuracy, sensitivity and specificity in the younger adult population were 93.75% (95% CI 93.7%-93.8%), 65.00% (95% CI 64.33%-65.67%) and 98.49% (95% CI 98.45%-98.52%), respectively. Accuracy reached 100% in VFs grading once the VFs were labeled accurately. The unique pattern of limbus-like VFs, 43 (35.8%) were investigated only in the younger adult population. If limbus-like VFs from the dataset were not included, the accuracy increased from 93.75% (95% CI 93.70%-93.80%) to 95.78% (95% CI 95.73%-95.82%), sensitivity increased from 65.00% (95% CI 64.33%-65.67%) to 70.13% (95% CI 68.98%-71.27%) and specificity remained unchanged at 98.49% (95% CI 98.45%-98.52%), respectively. The main causes of false negative results in older adults were patients’ lung markings, diaphragm or bowel airs (37%, n=14) followed by type I fracture (29%, n=11). The main causes of false negatives in younger adults were limbus-like VFs (45%, n=19), followed by type I fracture (26%, n=11). The overall kappa between AI discrimination and subspecialists’ consensus in the older and younger adult populations were 0.77 (95% CI, 0.733-0.805) and 0.72 (95% CI, 0.6524-0.80), respectively.

CONCLUSIONS: The developed VF-identifying AI ensemble model based on ground truth of older adults achieved better performance in identifying VFs in older adults and non-fractured thoracic and lumbar vertebrae in the younger adults. Different age distribution may have potential disease diversity and implicate the effect of ground truth generalizability on the AI model performance.

PMID:34737066 | DOI:10.1016/j.spinee.2021.10.020

J Acad Nutr Diet. 2021 Nov 1:S2212-2672(21)01434-9. doi: 10.1016/j.jand.2021.10.025. Online ahead of print.

ABSTRACT

BACKGROUND: Some preliminary studies indicate that components in coffee may have anticarcinogenic effects. However, the association between coffee-drinking habits and the risk of NHL remain controversial.

OBJECTIVE: To examine the relationship between coffee intake and non-Hodgkin’s lymphoma (NHL) incidence in a large prospective study of postmenopausal US women.

DESIGN AND PARTICIPANTS/SETTING: The participants included 74,935 women from the Women’s Health Initiative Observational Study (WHI-OS) who were recruited from 1993 through 1998. Information about coffee-drinking habits was collected at baseline via self-administered questionnaires.

MAIN OUTCOME MEASURES: Newly diagnosed NHL was validated by medical records and pathology records. Separate analyses were performed for the following three subtypes of NHL: diffuse large B-cell lymphoma (DLBCL (n=244)), follicular lymphoma (FL (n=166)), and chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL (n=64)).

STATISTICAL ANALYSES PERFORMED: Age-adjusted and multivariable-adjusted Cox proportional hazards models were used to determine associations of coffee intake (specifically, the total amount of coffee consumed daily, coffee types, and coffee preparation methods) with risk of NHL.

RESULTS: A total of 851 women developed NHL during a median 18.34 years of follow-up (range, 0.01 to 24.30 years; SD ± 6.63 years). Overall, no associations were observed between coffee intake and risk of NHL regardless of the total amount of daily coffee intake (P-value for trend test = 0.90), caffeinated (P-value=0.55) or decaffeinated coffee intake (P-value=0.78), and filtered or unfiltered coffee intake (P-value=0.91) after controlling for sociodemographic factors, lifestyle risk factors, and clinical risk factors/current medical conditions. No significant associations were observed between coffee intake with specific subtypes of NHL. A statistically significant interaction was found between alcohol intake, coffee intake, and incident NHL (P-value for interaction=0.02) based on the adjusted analysis. Specifically, among women who frequently consumed alcohol (>7 drinks/week), those who had moderate coffee intake (2-3 cups coffee/day) had a significantly reduced risk of developing NHL (HR:0.61, 95%CI: 0.36-0.98), compared to those who did not drink coffee.

CONCLUSION: The findings from this study do not support an association between coffee consumption and NHL risk, irrespective of the total amount of daily coffee intake, coffee types, or coffee preparation methods.

PMID:34737090 | DOI:10.1016/j.jand.2021.10.025

Photodiagnosis Photodyn Ther. 2021 Nov 1:102611. doi: 10.1016/j.pdpdt.2021.102611. Online ahead of print.

ABSTRACT

PURPOSE: To compare the optical coherence tomography angiography (OCT-A) of the macula, and optic disc in primary open-angle glaucoma (POAG), pseudoexfoliation glaucoma (PXG), and normal eyes.

METHODS: In this observational, cross-sectional study, Sixty-five eyes with POAG, 61 eyes of age, and mean deviation (MD) from standard automated perimetry matched PXG patients, and 45 normal control eyes underwent OCT-A using AngioVue, and optic disc-associated vessel density (VD), macular-associated VD, Foveal avascular zone (FAZ) area, FAZ perimeter (PERIM), and vessel density within a 300 μm wide region of the FAZ (FD) were compared between groups.

RESULTS: Peripapillary OCT-A parameters were significantly different among normal, and glaucomatous eyes with the highest values in the control eyes, but none of the peripapillary OCT-A parameters except inside disc VD was statistically significantly different between the PXG eyes, and POAG eyes. Correlation analysis revealed significant correlation between mean retinal nerve fiber layer (RNFL) and peripapillary VD in control (r=0.427, P=0.006), PXG (r=0.82, P<0.001), and POAG (r=0.79, P<0.001) eyes. PXG eyes exhibited significantly lower superficial, and deep macular vessel densities in parafoveal and perifoveal regions compared with POAG eyes after adjustment (p<= 0.05). Overall, moderate /severe PXG eyes had lower superficial, and deep vessel densities in parafoveal and perifoveal regions compared to mild PXG.

CONCLUSION: Peripapillary VD (inside disc), and macular vessel densities (parafovea and perifovea) demonstrate a significant difference in age and glaucoma severity-matched POAG and PXG eyes. Moderate/advance glaucomatous eyes, exhibit significant damage to the superficial and deep macular vascular structures.

PMID:34737059 | DOI:10.1016/j.pdpdt.2021.102611

Ann Allergy Asthma Immunol. 2021 Nov 1:S1081-1206(21)01170-4. doi: 10.1016/j.anai.2021.10.031. Online ahead of print.

ABSTRACT

BACKGROUND: The prevalence of sesame allergy is increasing; strict avoidance is the mainstay of therapy. Lately, sesame oral immunotherapy was presented as an alternative treatment, with a high rate of success. Therefore, data about the natural history and the clinical characteristics of patients with persistent sesame allergy are important for the management of sesame allergic patients.

OBJECTIVE: To describe the natural history of sesame allergic patients and the clinical characteristics of patients with spontaneous resolution of sesame allergy compared to patients with persistent sesame allergy.

METHODS: In our retrospective study, electronic health records of patients with sesame allergy diagnosis were reviewed for demographic and clinical data. Statistical analysis of clinical characteristics of patients with spontaneous resolution compared to persistent sesame allergy was performed.

RESULTS: One hundred and ninety patients with sesame allergy were followed for 3.86±4.43 years. Sixty one (32.1%) of them had spontaneous resolution of sesame allergy. Patients with mild, early (before the age of 10 months) first sesame allergic reaction, with smaller than 7 mm skin prick test and without concomitant tree nuts allergy had better resolution rate (56.1%).

CONCLUSION: Sesame allergy spontaneously resolved in about one third of our patients and in more than half of the patients with mild first reaction (grade 1) at a young age (<10 months), with small skin prick test (<7mm) and no concomitant tree nuts allergy. Larger prospective studies, with longer follow up period are needed to better characterize the sesame allergic patients with persistent allergy, who may need oral immunotherapy.

PMID:34737038 | DOI:10.1016/j.anai.2021.10.031

Photodiagnosis Photodyn Ther. 2021 Nov 1:102613. doi: 10.1016/j.pdpdt.2021.102613. Online ahead of print.

ABSTRACT

AIM: To evaluate the shear bond strength (SBS) of caries affected dentin (CAD) bonded to resin-modified glass ionomer cement (RMGIC) after being disinfected with different methods of disinfection and their effect on microleakage scores.

MATERIAL AND METHODS: Based on criteria of ICDAS seventy-five teeth were selected and disinfected. All samples were grounded and a cavity prepared followed by random allocation of samples into different disinfection groups i.e., methylene blue photosensitizer (MBP), Silver diamine fluoride (SDF), Ozone (O₃); curcumin photosensitizer (CP) and chlorhexidine (CHX) (n=15). All samples were restored with resin modified glass ionomer (RMGIC) incrementally. Ten specimens from each group were placed in universal testing machine (UTM) to calculate bond failure and failure type. Five samples from each group were evaluated for microleakage scores. SBS and microleakage scores was examined using analysis of variance (ANOVA) with a Post-hoc test. For all tests p-value, less than 0.5 was considered statistically significant.

RESULTS: The highest SBS was observed in CAD disinfected with CP bonded to RMGIC (16.42 ±1.10 MPa). Similarly, the lowest bond values were demonstrated by CAD disinfected with MBP (9.21 ± 0.22 MPa) bonded to RMGIC respectively. CAD disinfected with CHX (Control) bonded to RMGIC demonstrated the lowest microleakage scores and showed a significant difference compared to other experimental groups (p< 0.05).

CONCLUSION: Curcumin along with O₃ has the potential to be used as a disinfectant in CAD as it improves SBS to RMGIC. CHX demonstrates low microleakage scores with decrease bond integrity.

PMID:34737058 | DOI:10.1016/j.pdpdt.2021.102613

Appetite. 2021 Nov 1:105787. doi: 10.1016/j.appet.2021.105787. Online ahead of print.

ABSTRACT

This cross-sectional study estimated associations of eating contexts (including regularity of meals, places where they occur, and if they take place with attention and in company) with food consumption by degree of industrial processing and overweight indicators in a sample of Brazilian adolescents (14-19 years old) enrolled in 29 public schools in Juiz de Fora, MG (n = 805). We used an exploratory questionnaire, which was submitted to cluster analysis. Three clusters were identified: cluster 1 (n = 572), “appropriate eating contexts at breakfast, lunch, and dinner”; cluster 2 (n = 139), “inappropriate eating context at breakfast”; and cluster 3 (n = 94) “inappropriate eating context at dinner”. The evaluation of food consumption involved two 24-h dietary recalls, whose items were analyzed according to the NOVA classification system. Linear regression models were performed, using cluster 1 as reference. Clusters 2 and 3 were associated with an increase in the energy fraction from ultra-processed foods [respectively, β = 2.55% (IC 95%: 0.50; 5.05) and β = 4.18% (IC 95%: 1.21; 7.14)]; and cluster 2 was associated with a reduction in the energy fraction from unprocessed or minimally processed foods and culinary preparations [β = -3.61% (IC 95%: -6.40; -0.82)]. Additionally, clusters 2 and 3 were associated with an increase of body mass index for age [respectively, β = 0.23 z-score (IC 95%: 0.01; 0.46) and β = 0.27 z-score (IC 95%: 0.02; 0.54)]; and cluster 2 was associated with an increase of body fat [β = 1.21% (IC 95%: 0.23; 2.64)]. In conclusion, inappropriate eating contexts at breakfast and dinner were associated with higher ultra-processed food intake, higher body mass index and higher body fat percentage.

PMID:34737031 | DOI:10.1016/j.appet.2021.105787

Kidney Int. 2021 Nov 1:S0085-2538(21)01020-6. doi: 10.1016/j.kint.2021.09.024. Online ahead of print.

ABSTRACT

Registries are essential for health infrastructure planning, benchmarking, continuous quality improvement, hypothesis generation and real world trials. To date, data from these registries have predominantly been analysed in isolated “silos”, hampering efforts to analyse “big data” at the international level, with wide-ranging benefits including enhanced statistical power, an ability to conduct international comparisons, and greater capacity to study rare diseases. This review serves as a valuable resource to clinicians, researchers, and policymakers by comprehensively describing kidney failure registries active in 2021, before proposing approaches for inter-registry research under current conditions and solutions to enhance global capacity for data collaboration. We identified 79 kidney failure registries spanning 77 countries worldwide. International Society of Nephrology exemplar initiatives, including the “Global Kidney Health Atlas” and “Sharing Expertise to support the set-up of Renal Registries (SharE-RR)” continue to raise awareness regarding international healthcare disparities and support the development of universal kidney disease registries. Current barriers to inter-registry collaboration include under-representation of lower income countries, poor syntactic and semantic interoperability, absence of clear consensus guidelines for healthcare data sharing, and limited researcher incentives. This review represents a call to action for international stakeholders to enact systemic change that will harmonise the current fragmented approaches to kidney failure registry data collection and research.

PMID:34736973 | DOI:10.1016/j.kint.2021.09.024

Int J Cardiol. 2021 Nov 1:S0167-5273(21)01766-6. doi: 10.1016/j.ijcard.2021.10.157. Online ahead of print.

ABSTRACT

BACKGROUND: Cardiac sarcoidosis (CS) is clinically diagnosed in 5% of patients with sarcoidosis but imaging studies suggest higher prevalence. We evaluated the prevalence, clinical manifestations, and cardiovascular outcomes of CS, diagnosed through ¹⁸F-Fluorodeoxyglucose positron emission tomography ([¹⁸F]FDG-PET), in a southern European population.

METHODS: Retrospective single-centre study of patients screened for sarcoidosis with [¹⁸F]FDG-PET. Subjects with histological confirmation were divided in two groups, CS or extracardiac sarcoidosis, according to Heart Rhythm Society’s criteria. Primary endpoint was defined as the composite of heart failure hospitalizations, uncontrolled arrythmias, pacemaker implantation, and cardiovascular (CV) mortality. Secondary outcomes included each component and all-cause mortality.

RESULTS: From 128 patients with biopsy-proven extracardiac sarcoidosis, 10.2% had probable CS, 54% without symptoms of cardiac involvement. Ten patients had suggestive [¹⁸F]FDG uptake patterns, three subjects had an indicative cardiac magnetic resonance (CMR). Patients with probable CS had significantly higher prevalence of coronary and valvular disease, heart failure, and atrial fibrillation compared with those without cardiac involvement. During a mean follow-up of 4.0 SD2.7 years, the primary outcome occurred more frequently in patients with probable CS (53.8% vs. 3.5%; HR 25.45; 95% CI 5.27-122.9; p < 0.01) as well as heart failure hospitalizations (46.2% vs. 0.9%), uncontrolled arrhythmias (23.1% vs. 1.7%) and pacemaker implantation (23.1% vs. 0.9%) (p < 0.01 for all). All-cause mortality was three-fold higher in probable CS, despite the absence of statistical significance (15% vs. 5%, p = 0.15).

CONCLUSIONS: Among patients with biopsy-proven sarcoidosis, cardiac involvement detected by [¹⁸F]FDG-PET or CMR is associated with a higher risk of CV events, irrespective of symptoms.

PMID:34736982 | DOI:10.1016/j.ijcard.2021.10.157

Gastrointest Endosc. 2021 Nov 1:S0016-5107(21)01763-6. doi: 10.1016/j.gie.2021.10.026. Online ahead of print.

ABSTRACT

BACKGROUND AND AIMS: With the increasing incidence of small gastrointestinal stromal tumors (GISTs), endoscopic full-thickness resection (EFR) and cap-assisted EFR (EFR-C) have been suggested as 2 effective resection methods. We aimed to compare the outcomes of EFR and EFR-C for the treatment of small (≤ 1.5 cm) gastric GISTs.

METHODS: This retrospective study included 67 patients who underwent EFR and 46 patients who underwent EFR-C at Nanjing Drum Tower Hospital. Clinicopathological features, adverse events and outcomes were compared between the 2 groups. Univariate and multivariate linear and logistic regressions were used to analyze the effects of the procedure on the therapeutic outcomes of patients and adjusted for covariates in multivariate analysis.

RESULTS: The tumor size in the EFR group tended to be larger (P = 0.005). The resection time in the EFR-C group was shorter than that in the EFR group (38.3 ± 20.7 min vs 15.0 ± 11.8 min, P < 0.001), which retained statistical significance with adjustment for the covariates (adjusted mean difference: 22.2; 95% CI, 15.0-29.4, P < 0.001). The R0 resection rate of the EFR group was 94.0%, and for the EFR-C group, it was 97.8% (P = 0.355). The EFR-C group was superior to the EFR group in terms of perioperative therapeutic outcomes, adverse events and postoperative recovery. No recurrence occurred in the EFR and EFR-C groups.

CONCLUSION: EFR-C was found to be the preferable technique for small (≤1.5 cm) gastric GISTs with shorter operation times, lower adverse events, faster postoperative recovery, and shorter hospitalization times as compared with those with EFR.

PMID:34736933 | DOI:10.1016/j.gie.2021.10.026

Prog Biophys Mol Biol. 2021 Nov 1:S0079-6107(21)00125-5. doi: 10.1016/j.pbiomolbio.2021.10.003. Online ahead of print.

ABSTRACT

This response aims to expand on some of the issues raised by Keith Baverstock’s The Gene: An Appraisal, especially on the evolution and nature of knowledge in living things. In contrast to the simple associationism envisaged in “genetic information”, it emphasises the dynamic complexity and changeability of most natural environments, and, therefore, predictability based on underlying statistical structures. That seems to be the basis of the “cognitive” functions increasingly being reported about cellular, as well as more evolved, functions, and of the autonomous agency of organisms thriving creatively in complex environments.

PMID:34736965 | DOI:10.1016/j.pbiomolbio.2021.10.003