Search

The scarcity of high-quality annotated data poses a significant challenge to the application of deep learning in fabric defect tasks, limiting the generalization and segmentation performance of existing models and impeding their capability to address the complexity of various fabric types and defects. To overcome these obstacles, this study introduces an innovative method to infuse specialized knowledge of fabric defects into the Segment Anything Model (SAM), a large-scale visual model. By introducing and training a unique set of fabric defect-related parameters, this approach seamlessly integrates domain-specific knowledge into SAM without the need for extensive modifications to the preexisting model parameters. The revamped SAM model leverages generalized image understanding learned from large-scale natural image datasets while incorporating fabric defect-specific knowledge, ensuring its proficiency in fabric defect segmentation tasks. The experimental results reveal a significant improvement in the model’s segmentation performance, attributable to this novel amalgamation of generic and fabric-specific knowledge. When benchmarking against popular existing segmentation models across three datasets, our proposed model demonstrates a substantial leap in performance. Its impressive results in cross-dataset comparisons and few-shot learning experiments further demonstrate its potential for practical applications in textile quality control.

We sought to quantify plasma high sensitivity cardiac troponin (hs-cTnT) levels, their determinants, and their associations with left ventricular (LV) myocardial deformation in adult survivors of childhood acute leukaemias.One hundred adult survivors (57 males) of childhood acute leukaemias, aged 24.1 ± 4.2 years, and 42 age-matched controls (26 males) were studied. Plasma cTnT was determined using a highly sensitive assay. Genotyping of NAD(P)H oxidase and multidrug resistance protein polymorphisms was performed. Left ventricular function was assessed by conventional, three-dimensional, and speckle tracking echocardiography. The medians (interquartile range) of hs-cTnT in male and female survivors were 4.9 (4.2 to 7.2) ng/L and 1.0 (1.0 to 3.5) ng/L, respectively. Nineteen survivors (13 males, 6 females) (19%) had elevated hs-cTnT (>95(th) centile of controls). Compared to those without elevated hs-TnT levels, these subjects had received larger cumulative anthracycline dose and were more likely to have leukaemic relapse, stem cell transplant, and cardiac irradiation. Their LV systolic and early diastolic myocardial velocities, isovolumic acceleration, and systolic longitudinal strain rate were significantly lower. Survivors having CT/TT at CYBA rs4673 had higher hs-cTnT levels than those with CC genotype. Functionally, increased hs-cTnT levels were associated with worse LV longitudinal systolic strain and systolic and diastolic strain rates.Increased hs-cTnT levels occur in a significant proportion of adult survivors of childhood acute leukaemias and are associated with larger cumulative anthracycline dose received, history of leukaemic relapse, stem cell transplant, and cardiac irradiation, genetic variants in free radical metabolism, and worse LV myocardial deformation.