Colon and lung cancer classification from multi-modal images using resilient and efficient neural network architectures.

Automatic classification of colon and lung cancer images is crucial for early detection and accurate diagnostics. However, there is room for improvement to enhance accuracy, ensuring better diagnostic precision. This study introduces two novel dense architectures (D1 and D2) and emphasizes their effectiveness in classifying colon and lung cancer from diverse images. It also highlights their resilience, efficiency, and superior performance across multiple datasets. These architectures were tested on various types of datasets, including NCT-CRC-HE-100K (set of 100,000 non-overlapping image patches from hematoxylin and eosin (H&E) stained histological images of human colorectal cancer (CRC) and normal tissue), CRC-VAL-HE-7K (set of 7180 image patches from N = 50 patients with colorectal adenocarcinoma, no overlap with patients in NCT-CRC-HE-100K), LC25000 (Lung and Colon Cancer Histopathological Image), and IQ-OTHNCCD (Iraq-Oncology Teaching Hospital/National Center for Cancer Diseases), showcasing their effectiveness in classifying colon and lung cancers from histopathological and Computed Tomography (CT) scan images. This underscores the multi-modal image classification capability of the proposed models. Moreover, the study addresses imbalanced datasets, particularly in CRC-VAL-HE-7K and IQ-OTHNCCD, with a specific focus on model resilience and robustness. To assess overall performance, the study conducted experiments in different scenarios. The D1 model achieved an impressive 99.80 % accuracy on the NCT-CRC-HE-100K dataset, with a Jaccard Index (J) of 0.8371, a Matthew's Correlation Coefficient (MCC) of 0.9073, a Cohen's Kappa (Kp) of 0.9057, and a Critical Success Index (CSI) of 0.8213. When subjected to 10-fold cross-validation on LC25000, the D1 model averaged (avg) 99.96 % accuracy (avg J, MCC, Kp, and CSI of 0.9993, 0.9987, 0.9853, and 0.9990), surpassing recent reported performances. Furthermore, the ensemble of D1 and D2 reached 93 % accuracy (J, MCC, Kp, and CSI of 0.7556, 0.8839, 0.8796, and 0.7140) on the IQ-OTHNCCD dataset, exceeding recent benchmarks and aligning with other reported results. Efficiency evaluations were conducted in various scenarios. For instance, training on only 10 % of LC25000 resulted in high accuracy rates of 99.19 % (J, MCC, Kp, and CSI of 0.9840, 0.9898, 0.9898, and 0.9837) (D1) and 99.30 % (J, MCC, Kp, and CSI of 0.9863, 0.9913, 0.9913, and 0.9861) (D2). In NCT-CRC-HE-100K, D2 achieved an impressive 99.53 % accuracy (J, MCC, Kp, and CSI of 0.9906, 0.9946, 0.9946, and 0.9906) with training on only 30 % of the dataset and testing on the remaining 70 %. When tested on CRC-VAL-HE-7K, D1 and D2 achieved 95 % accuracy (J, MCC, Kp, and CSI of 0.8845, 0.9455, 0.9452, and 0.8745) and 96 % accuracy (J, MCC, Kp, and CSI of 0.8926, 0.9504, 0.9503, and 0.8798), respectively, outperforming previously reported results and aligning closely with others. Lastly, training D2 on just 10 % of NCT-CRC-HE-100K and testing on CRC-VAL-HE-7K resulted in significant outperformance of InceptionV3, Xception, and DenseNet201 benchmarks, achieving an accuracy rate of 82.98 % (J, MCC, Kp, and CSI of 0.7227, 0.8095, 0.8081, and 0.6671). Finally, using explainable AI algorithms such as Grad-CAM, Grad-CAM++, Score-CAM, and Faster Score-CAM, along with their emphasized versions, we visualized the features from the last layer of DenseNet201 for histopathological as well as CT-scan image samples. The proposed dense models, with their multi-modality, robustness, and efficiency in cancer image classification, hold the promise of significant advancements in medical diagnostics. They have the potential to revolutionize early cancer detection and improve healthcare accessibility worldwide.

Pterostilbene Protects against Osteoarthritis through NLRP3 Inflammasome Inactivation and Improves Gut Microbiota as Evidenced by In Vivo and In Vitro Studies.

Osteoarthritis (OA) is a persistent inflammatory disease, and long-term clinical treatment often leads to side effects. In this study, we evaluated pterostilbene (PT), a natural anti-inflammatory substance, for its protective effects and safety during prolonged use on OA. Results showed that PT alleviated the loss of chondrocytes and widened the narrow joint space in an octacalcium phosphate (OCP)-induced OA mouse model (n = 3). In vitro experiments demonstrate that PT reduced NLRP3 inflammation activation (relative protein expression: C: 1 ± 0.09, lipopolysaccharide (LPS): 1.14 ± 0.07, PT: 0.91 ± 0.07, LPS + PT: 0.68 ± 0.04) and the release of inflammatory cytokines through NF-κB signaling inactivation (relative protein expression: C: 1 ± 0.03, LPS: 3.49 ± 0.02, PT: 0.66 ± 0.08, LPS + PT: 2.78 ± 0.05), ultimately preventing cartilage catabolism. Interestingly, PT also altered gut microbiota by reducing inflammation-associated flora and increasing the abundance of healthy bacteria in OA groups. Collectively, these results suggest that the PT can be considered as a protective strategy for OA.

Flipped classroom teaching model with video instruction improves skills in local anesthesia training.

Background/purpose: Local anesthesia (LA) training is an essential clinical skill in dental education. However, the traditional teaching method of student-to-student injection has ethical concerns. This study investigated whether a flipped classroom (FC) teaching model with instructional videos improves students' skills in administering LA. Materials and methods: Fourth-year dental students in 2017 (traditional teaching, n = 70) and 2018 (FC model, n = 79) were assessed for their ability to perform an inferior alveolar nerve block and lingual nerve block. The FC group watched pre-recorded videos prior to a hands-on training session. Skills were evaluated using a 24-item checklist converted to letter grades. Perceptions of the FC approach were also surveyed. Results: The FC group showed statistically significantly higher final grades than the traditional teaching group (P < 0.05). Most FC students agreed that the videos improved clarity and learning objectives. Conclusion: The FC teaching model with procedural video instruction improved skills and confidence in administering LA over traditional teaching methods. Videos can be a beneficial supplement in pre-clinical dental training.

The Use of Proton and Carbon Ion Radiation Therapy for Sarcomas.

The unique physical and biological characteristics of proton and carbon ions allow for improved sparing of normal tissues, decreased integral dose to the body, and increased biological effect through high linear energy transfer. These properties are particularly useful for sarcomas given their histology, wide array of locations, and age of diagnosis. This review summarizes the literature and describes the clinical situations in which these heavy particles have advantages for treating sarcomas.

The role of 18F-FDG PET in minimizing variability in gross tumor volume delineation of soft tissue sarcomas.

BACKGROUND: Accurate gross tumor volume (GTV) delineation is a critical step in radiation therapy treatment planning. However, it is reader dependent and thus susceptible to intra- and inter-reader variability. GTV delineation of soft tissue sarcoma (STS) often relies on CT and MR images. PURPOSE: This study investigates the potential role of 18F-FDG PET in reducing intra- and inter-reader variability thereby improving reproducibility of GTV delineation in STS, without incurring additional costs or radiation exposure. MATERIALS AND METHODS: Three readers performed independent GTV delineation of 61 patients with STS using first CT and MR followed by CT, MR, and 18F-FDG PET images. Each reader performed a total of six delineation trials, three trials per imaging modality group. Dice Similarity Coefficient (DSC) score and Hausdorff distance (HD) were used to assess both intra- and inter-reader variability using generated simultaneous truth and performance level estimation (STAPLE) GTVs as ground truth. Statistical analysis was performed using a Wilcoxon signed-ranked test. RESULTS: There was a statistically significant decrease in both intra- and inter-reader variability in GTV delineation using CT, MR 18F-FDG PET images vs. CT and MR images. This was translated by an increase in the DSC score and a decrease in the HD for GTVs drawn from CT, MR and 18F-FDG PET images vs. GTVs drawn from CT and MR for all readers and across all three trials. CONCLUSION: Incorporation of 18F-FDG PET into CT and MR images decreased intra- and inter-reader variability and subsequently increased reproducibility of GTV delineation in STS.

Use of reconstruction plates prebent on three-dimensional models to reduce the complications of mandibular reconstruction.

Background/purpose: Though the gold standard method for mandible reconstruction of the defect from segmental mandibulectomy is by osseous flap or graft, using reconstruction plates is still indicated in some cases. Traditionally, the plate is bended immediately after the segmental mandibulectomy by freehand. However, it's difficult to fit well to the original position of mandible, which may result in more complications. This study therefore aimed to investigate whether using prebent plates on computer-aided 3D printing models could reduce the complication rate. Materials and methods: Patients who received mandible reconstruction by reconstruction plate from 2018 to 2022 were enrolled and evaluated in this study. The data, including demographics, indications for surgery, pre-existed preoperative and postoperative therapies, classification of defects, and postoperative outcomes were collected and analyzed. Results: A total of 52 patients were enrolled in our study. The prebent group exhibited a significantly lower complication rate than that of the immediately bent group (P = 0.012). Other risk factors of plate complications included postoperative adjuvant radiotherapy (P = 0.017) and previous surgery (P = 0.047). The complication-free survival rate was also better in the prebent group in a 3-year follow-up period (P = 0.012). Conclusion: Prebent plates on computer-aided printing models proved to be an effective approach to reduce the complications for mandibular reconstruction in segmental mandibulectomy.

Pannexin-3 stabilizes the transcription factor Bcl6 in a channel-independent manner to protect against vascular oxidative stress.

Targeted degradation regulates the activity of the transcriptional repressor Bcl6 and its ability to suppress oxidative stress and inflammation. Here, we report that abundance of endothelial Bcl6 is determined by its interaction with Golgi-localized pannexin 3 (Panx3) and that Bcl6 transcriptional activity protects against vascular oxidative stress. Consistent with data from obese, hypertensive humans, mice with an endothelial cell-specific deficiency in Panx3 had spontaneous systemic hypertension without obvious changes in channel function, as assessed by Ca2+ handling, ATP amounts, or Golgi luminal pH. Panx3 bound to Bcl6, and its absence reduced Bcl6 protein abundance, suggesting that the interaction with Panx3 stabilized Bcl6 by preventing its degradation. Panx3 deficiency was associated with increased expression of the gene encoding the H2O2-producing enzyme Nox4, which is normally repressed by Bcl6, resulting in H2O2-induced oxidative damage in the vasculature. Catalase rescued impaired vasodilation in mice lacking endothelial Panx3. Administration of a newly developed peptide to inhibit the Panx3-Bcl6 interaction recapitulated the increase in Nox4 expression and in blood pressure seen in mice with endothelial Panx3 deficiency. Panx3-Bcl6-Nox4 dysregulation occurred in obesity-related hypertension, but not when hypertension was induced in the absence of obesity. Our findings provide insight into a channel-independent role of Panx3 wherein its interaction with Bcl6 determines vascular oxidative state, particularly under the adverse conditions of obesity.

The prevalence of myopia remains stable under tighter COVID-19 social restriction in preschoolers receiving a school-based eyecare program.

PURPOSE: This study investigated the impact of different levels of COVID-19 social restrictions (social distancing in 2020, large-scale home confinement in 2021) on myopia prevalence and behaviours in a preschool population with school-based eyecare programme. METHODS: Repeated cross-sectional surveys were conducted between August and December in 2019, 2020 and 2021. Children aged 5-6 years received ocular examinations, and questionnaires were answered by caregivers before the day of the examination. The main outcome measures were the changes in after-school time spent on homework, screen-based devices and outdoors. Secondary outcome was the change in myopia prevalence (spherical equivalent [SE] ≤ -0.5 D in either eye after cycloplegia). RESULTS: A total of 9997 preschoolers were included in the analysis. Under tighter restrictions, more preschoolers spent ≥1 h/day on screen-based devices (42.8% in 2019, 45.2% in 2020, 48.9% in 2021, p < 0.001), and fewer preschoolers spent ≥30 min/day on after-school outdoor activities (49.5% in 2019, 46.0% in 2020, 41.0% in 2021, p < 0.001) on weekdays. A similar trend was found on weekends. While more preschoolers spent ≥2 h/day on screen-based devices (35.3% in 2019, 38.5% in 2020, 43.0% in 2021, p < 0.001), fewer preschoolers spent ≥2 h/day on outdoor activities (41.7% in 2019, 41.7% in 2020, 34.0% in 2021, p < 0.001). The mean SE and myopia prevalence were stable (9.1% in 2019, 10.3% in 2020, 9.4% in 2021, p = 0.707). CONCLUSION: Our study showed dose-dependent effect of social restrictions on near-work and outdoor behaviours at home. The prevalence of myopia did not increase significantly with short-term cessation of school-based eyecare programmes.

Effects of aging and diabetes on the deformation mechanisms and molecular structural characteristics of collagen fibrils under daily activity.

Crosslinking plays an important role in collagen-based tissues since it affects mechanical behavior and tissue metabolism. Aging and diabetes affect the type and density of crosslinking, effectively altering tissue properties. However, most studies focus on these effects under large stress rather than daily activities. We focus on the deformation mechanisms and structural change at the binding sites for integrins, proteoglycans, and collagenase in collagen fibrils using a fully atomistic model. We show that high-connectivity enzymatic crosslinking (our "HC" model, representing normal tissues) and advanced-glycation end-products (our "Glucosepane" model, which increase in diabetes) result in uniform deformation under daily activity, but low-connectivity enzymatic crosslinking (our "LC" model, representing aging tissues) does not. In particular, the HC model displays more sliding, which may explain the ability of healthy tissues to absorb more strain energy. In contrast, AGEs induce instability in the structures near the binding sites, which would affect the tissue metabolism of the collagen molecule. Our results provide important insights into the molecular mechanisms of collagen and a possible explanation for the role of crosslinking in tissues undergoing daily activity.

Immunohistochemical Stain-Aided Annotation Accelerates Machine Learning and Deep Learning Model Development in the Pathologic Diagnosis of Nasopharyngeal Carcinoma.

Nasopharyngeal carcinoma (NPC) is an epithelial cancer originating in the nasopharynx epithelium. Nevertheless, annotating pathology slides remains a bottleneck in the development of AI-driven pathology models and applications. In the present study, we aim to demonstrate the feasibility of using immunohistochemistry (IHC) for annotation by non-pathologists and to develop an efficient model for distinguishing NPC without the time-consuming involvement of pathologists. For this study, we gathered NPC slides from 251 different patients, comprising hematoxylin and eosin (H&E) slides, pan-cytokeratin (Pan-CK) IHC slides, and Epstein-Barr virus-encoded small RNA (EBER) slides. The annotation of NPC regions in the H&E slides was carried out by a non-pathologist trainee who had access to corresponding Pan-CK IHC slides, both with and without EBER slides. The training process utilized ResNeXt, a deep neural network featuring a residual and inception architecture. In the validation set, NPC exhibited an AUC of 0.896, with a sensitivity of 0.919 and a specificity of 0.878. This study represents a significant breakthrough: the successful application of deep convolutional neural networks to identify NPC without the need for expert pathologist annotations. Our results underscore the potential of laboratory techniques to substantially reduce the workload of pathologists.

Novel five nucleotide deletion in dysferlin leads to autosomal recessive limb-girdle muscular dystrophy.

Muscular dystrophy (MD) is a genetic disorder that causes progressive muscle weakness and degeneration. Limb-girdle muscular dystrophy (LGMD) is a type of MD that mainly causes muscle atrophy within the shoulder and pelvic girdles. LGMD is classified into autosomal dominant (LGMD-D) and autosomal recessive (LGMD-R) inheritance patterns. Mutations in the Dysferlin gene (DYSF) are common causes of LGMD-R. However, genetic screening of DYSF mutations is rare in Taiwan. Herein, we identified a novel c.2867_2871del ACCAG deletion and a previously reported c.937+1G>A mutation in DYSF from a Taiwanese family with LGMD. The primary symptoms of both siblings were difficulty climbing stairs, walking on the toes, and gradually worsening weakness in the proximal muscles and increased creatine kinase level. Through pedigree analysis and sequencing, two siblings from this family were found to have compound heterozygous DYSF mutations (c. 937+1G>A and c. 2867_2871del ACCAG) within the separated alleles. These mutations induced early stop codons; if translated, truncated DYSF proteins will be expressed. Or, the mRNA products of these two mutations will merit the nonsense-mediated decay, might result in no dysferlin protein expressed. To our knowledge, this is the first report of a novel c.2867_2871del ACCAG deletion in DYSF. Further research is required to examine the effects of the novel DYSF mutation in Taiwanese patients with LGMD.

Pilot Study: Short Term Impact of Radiation Therapy on Bone Mineral Density and Bone Metabolism.

Despite the risk of complications, high dose radiation therapy is increasingly utilized in the management of selected bone malignancies. In this study, we investigate the impact of moderate to high dose radiation (over 50 Gy) on bone metabolism and structure. Between 2015 and 2018, patients with a primary malignant bone tumor of the sacrum that were either treated with high dose definitive radiation only or a combination of moderate to high dose radiation and surgery were prospectively enrolled at a single institution. Quantitative CTs were performed before and after radiation to determine changes in volumetric bone mineral density (BMD) of the irradiated and non-irradiated spine. Bone histomorphometry was performed on biopsies of the irradiated sacrum and the non-irradiated iliac crest of surgical patients using a quadruple tetracycline labeling protocol. In total, 9 patients were enrolled. Two patients received radiation only (median dose 78.3 Gy) and 7 patients received a combination of preoperative radiation (median dose 50.4 Gy), followed by surgery. Volumetric BMD of the non-irradiated lumbar spine did not change significantly after radiation, while the BMD of the irradiated sacrum did (pre-radiation median: 108.0 mg/cm3 (IQR 91.8-167.1); post-radiation median: 75.3 mg/cm3 (IQR 57.1-110.2); p = 0.010). The cancellous bone of the non-irradiated iliac crest had a stable bone formation rate, while the irradiated sacrum showed a significant decrease in bone formation rate [pre-radiation median: 0.005 mm3/mm2/year (IQR 0.003-0.009), post-radiation median: 0.001 mm3/mm2/year (IQR 0.001-0.001); p = 0.043]. Similar effects were seen in the cancellous and endocortical envelopes. This pilot study shows a decrease of volumetric BMD and bone formation rate after high-dose radiation therapy. Further studies with larger cohorts and other endpoints are needed to get more insight into the effect of radiation on bone. Level of evidence: IV.

Purinergic P2Y2 receptor-induced activation of endothelial TRPV4 channels mediates lung ischemia-reperfusion injury.

Lung ischemia-reperfusion injury (IRI), characterized by inflammation, vascular permeability, and lung edema, is the major cause of primary graft dysfunction after lung transplantation. Here, we investigated the cellular mechanisms underlying lung IR-induced activation of endothelial TRPV4 channels, which play a central role in lung edema and dysfunction after IR. In a left lung hilar-ligation model of IRI in mice, we found that lung IRI increased the efflux of ATP through pannexin 1 (Panx1) channels at the endothelial cell (EC) membrane. Elevated extracellular ATP activated Ca2+ influx through endothelial TRPV4 channels downstream of purinergic P2Y2 receptor (P2Y2R) signaling. P2Y2R-dependent activation of TRPV4 channels was also observed in human and mouse pulmonary microvascular endothelium in ex vivo and in vitro models of IR. Endothelium-specific deletion of P2Y2R, TRPV4, or Panx1 in mice substantially prevented lung IRI-induced activation of endothelial TRPV4 channels and lung edema, inflammation, and dysfunction. These results identify endothelial P2Y2R as a mediator of the pathological sequelae of IRI in the lung and show that disruption of the endothelial Panx1-P2Y2R-TRPV4 signaling pathway could be a promising therapeutic strategy for preventing lung IRI after transplantation.

Gamma-Glutamyltransferase Is a Predictor for Future Changes of Diabetogenic Factors in Aged Chinese-A Four-Year Follow-Up Study.

Glucose homeostasis in the body is determined by four diabetes factors (DFs): insulin resistance (IR), glucose effectiveness (GE), and the two phases of insulin secretion-first phase (FPIS) and second phase (SPIS). Previous research points to a correlation between elevated levels of gamma-glutamyl transferase (γGT) and an increased risk of type 2 diabetes. This study investigates the relationship between γGT and the four DFs in older Chinese individuals. This study involved 2644 men and 2598 women, all of whom were relatively healthy Chinese individuals aged 60 years or more. The DFs were calculated using formulas developed by our research, based on demographic data and factors related to metabolic syndrome. Pearson's correlation was utilized to assess the relationship between γGT and the four DFs. The findings suggested a positive correlation between γGT and IR, FPIS, and SPIS, but a negative correlation with GE in men. Among women, only SPIS and GE were significantly correlated with γGT. The factors showed varying degrees of correlation, listed in descending order as follows: GE, SPIS, FPIS, and IR. This study confirms a significant correlation between γGT and DFs in this population, highlighting the noteworthy role of GE.

Therapeutic Effects of Treating COVID-19 Vaccine-Induced Anti-TIF1-γ-Positive Dermatomyositis.

An increase in skin-related autoimmune disorders has been reported as an adverse effect of coronavirus disease 2019 (COVID-19) vaccines. We present the case of a 90-year-old Taiwanese female who was newly diagnosed with anti-transcription intermediary factor 1-gamma (anti-TIF1-γ)-positive dermatomyositis (DM) after receiving a second dose of the AstraZeneca COVID-19 vaccine. Under treatment with prednisolone and monoclonal antibody therapy of abatacept, her skin lesions improved, and her muscle power increased. The serum creatinine phosphokinase level decreased from 4858 to 220 U/L, and the anti-TIF1-γ antibody titer decreased from 202 to 99. Flow cytometry data showed an increase in T cells, while NK cells, B cells (CD19), and plasma blasts all decreased. These findings suggest that standard DM treatment might be beneficial to patients with COVID-19 vaccine-induced DM.

Interactive effects of CDKN2B-AS1 gene polymorphism and habitual risk factors on oral cancer.

Oral squamous cell carcinoma (OSCC) is a common malignant disease associated with a high mortality rate and heterogeneous disease aetiology. Cyclin dependent kinase inhibitor 2B antisense RNA 1 (CDKN2B-AS1), is a long noncoding RNA that has been shown to act as a scaffold, sponge, or signal hub to promote carcinogenesis. Here, we attempted to assess the effect of CDKN2B-AS1 single-nucleotide polymorphisms (SNPs) on the susceptibility to OSCC. Five CDKN2B-AS1 SNPs, including rs564398, rs1333048, rs1537373, rs2151280 and rs8181047, were analysed in 1060 OSCC cases and 1183 cancer-free controls. No significant association of these five SNPs with the risk of developing OSCC was detected between the case and control group. However, while examining the clinical characteristics, patients bearing at least one minor allele of rs1333048 (CA and CC) were more inclined to develop late-stage (stage III/IV, adjusted OR, 1.480; 95% CI, 1.129-1.940; p = 0.005) and large-size (greater than 2 cm in the greatest dimension, adjusted OR, 1.347; 95% CI, 1.028-1.765; p = 0.031) tumours, as compared with those homologous for the major allele (AA). Further stratification analyses demonstrated that this genetic correlation with the advanced stage of disease was observed only in habitual betel quid chewers (adjusted OR, 1.480; 95% CI, 1.076-2.035; p = 0.016) or cigarette smokers (adjusted OR, 1.531; 95% CI, 1.136-2.063; p = 0.005) but not in patients who were not exposed to these major habitual risks. These data reveal an interactive effect of CDKN2B-AS1 rs1333048 with habitual exposure to behavioural risks on the progression of oral cancer.