Classification of Mobile-Based Oral Cancer Images Using the Vision Transformer and the Swin Transformer.

Oral cancer, a pervasive and rapidly growing malignant disease, poses a significant global health concern. Early and accurate diagnosis is pivotal for improving patient outcomes. Automatic diagnosis methods based on artificial intelligence have shown promising results in the oral cancer field, but the accuracy still needs to be improved for realistic diagnostic scenarios. Vision Transformers (ViT) have outperformed learning CNN models recently in many computer vision benchmark tasks. This study explores the effectiveness of the Vision Transformer and the Swin Transformer, two cutting-edge variants of the transformer architecture, for the mobile-based oral cancer image classification application. The pre-trained Swin transformer model achieved 88.7% accuracy in the binary classification task, outperforming the ViT model by 2.3%, while the conventional convolutional network model VGG19 and ResNet50 achieved 85.2% and 84.5% accuracy. Our experiments demonstrate that these transformer-based architectures outperform traditional convolutional neural networks in terms of oral cancer image classification, and underscore the potential of the ViT and the Swin Transformer in advancing the state of the art in oral cancer image analysis.

Structural insights into human EMC and its interaction with VDAC.

The endoplasmic reticulum (ER) membrane protein complex (EMC) is a conserved, multi-subunit complex acting as an insertase at the ER membrane. Growing evidence shows that the EMC is also involved in stabilizing and trafficking membrane proteins. However, the structural basis and regulation of its multifunctionality remain elusive. Here, we report cryo-electron microscopy structures of human EMC in apo- and voltage-dependent anion channel (VDAC)-bound states at resolutions of 3.47 Å and 3.32 Å, respectively. We discovered a specific interaction between VDAC proteins and the EMC at mitochondria-ER contact sites, which is conserved from yeast to humans. Moreover, we identified a gating plug located inside the EMC hydrophilic vestibule, the substrate-binding pocket for client insertion. Conformation changes of this gating plug during the apo-to-VDAC-bound transition reveal that the EMC unlikely acts as an insertase in the VDAC1-bound state. Based on the data analysis, the gating plug may regulate EMC functions by modifying the hydrophilic vestibule in different states. Our discovery offers valuable insights into the structural basis of EMC's multifunctionality.

Cryo-EM structure of human sphingomyelin synthase and its mechanistic implications for sphingomyelin synthesis.

Sphingomyelin (SM) has key roles in modulating mammalian membrane properties and serves as an important pool for bioactive molecules. SM biosynthesis is mediated by the sphingomyelin synthase (SMS) family, comprising SMS1, SMS2 and SMS-related (SMSr) members. Although SMS1 and SMS2 exhibit SMS activity, SMSr possesses ceramide phosphoethanolamine synthase activity. Here we determined the cryo-electron microscopic structures of human SMSr in complexes with ceramide, diacylglycerol/phosphoethanolamine and ceramide/phosphoethanolamine (CPE). The structures revealed a hexameric arrangement with a reaction chamber located between the transmembrane helices. Within this structure, a catalytic pentad E-H/D-H-D was identified, situated at the interface between the lipophilic and hydrophilic segments of the reaction chamber. Additionally, the study unveiled the two-step synthesis process catalyzed by SMSr, involving PE-PLC (phosphatidylethanolamine-phospholipase C) hydrolysis and the subsequent transfer of the phosphoethanolamine moiety to ceramide. This research provides insights into the catalytic mechanism of SMSr and expands our understanding of sphingolipid metabolism.

Stoichiometry and architecture of the platelet membrane complex glycoprotein Ib-IX-V.

Glycoprotein (GP) Ib-IX-V is the second most abundant platelet receptor for thrombin and other ligands crucial for hemostasis and thrombosis. Its activity is involved in platelet adhesion to vascular injury sites and thrombin-induced platelet aggregation. GPIb-IX-V is a heteromeric complex composed of four subunits, GPIbα, GPIbß, GPV and GPIX, in a stoichiometric ratio that has been wildly debated. Despite its important physiological roles, the overall structure and molecular arrangement of GPIb-IX-V are not yet fully understood. Here, we purify stable and functional human GPIb-IX-V complex from reconstituted EXPi293F cells in high homogeneity, and perform biochemical and structural characterization of this complex. Single-particle cryo-electron microscopy structure of GPIb-IX-V is determined at ∼11 Å resolution, which unveils the architecture of GPIb-IX-V and its subunit organization. Size-exclusion chromatography-multi-angle static light scattering analysis reveals that GPIb-IX-V contains GPIb-IX and GPV at a 1:1 stoichiometric ratio and surface plasmon resonance assays show that association of GPV leads to slow kinetics of thrombin binding to GPIb-IX-V. Taken together, our results provide the first three-dimensional architecture of the intact GPIb-IX-V complex, which extends our understanding of the structure and functional mechanism of this complex in hemostasis and thrombosis.

Inter-observer agreement among specialists in the diagnosis of oral potentially malignant disorders and oral cancer using store-and-forward technology.

OBJECTIVES: Oral cancer is a leading cause of morbidity and mortality. Screening and mobile Health (mHealth)-based approach facilitates early detection remotely in a resource-limited settings. Recent advances in eHealth technology have enabled remote monitoring and triage to detect oral cancer in its early stages. Although studies have been conducted to evaluate the diagnostic efficacy of remote specialists, to our knowledge, no studies have been conducted to evaluate the consistency of remote specialists. The aim of this study was to evaluate interobserver agreement between specialists through telemedicine systems in real-world settings using store-and-forward technology. MATERIALS AND METHODS: The two remote specialists independently diagnosed clinical images (n=822) from image archives. The onsite specialist diagnosed the same participants using conventional visual examination, which was tabulated. The diagnostic accuracy of two remote specialists was compared with that of the onsite specialist. Images that were confirmed histopathologically were compared with the onsite diagnoses and the two remote specialists. RESULTS: There was moderate agreement (k= 0.682) between two remote specialists and (k= 0.629) between the onsite specialist and two remote specialists in the diagnosis of oral lesions. The sensitivity and specificity of remote specialist 1 were 92.7% and 83.3%, respectively, and those of remote specialist 2 were 95.8% and 60%, respectively, each compared with histopathology. CONCLUSION: The diagnostic accuracy of the two remote specialists was optimal, suggesting that "store and forward" technology and telehealth can be an effective tool for triage and monitoring of patients. CLINICAL RELEVANCE: Telemedicine is a good tool for triage and enables faster patient care in real-world settings.

The cryo-EM structure of the human ERAD retrotranslocation complex.

Endoplasmic reticulum-associated degradation (ERAD) maintains protein homeostasis by retrieving misfolded proteins from the endoplasmic reticulum (ER) lumen into the cytosol for degradation. The retrotranslocation of misfolded proteins across the ER membrane is an energy-consuming process, with the detailed transportation mechanism still needing clarification. We determined the cryo-EM structures of the hetero-decameric complex formed by the Derlin-1 tetramer and the p97 hexamer. It showed an intriguing asymmetric complex and a putative coordinated squeezing movement in Derlin-1 and p97 parts. With the conformational changes of p97 induced by its ATP hydrolysis activities, the Derlin-1 channel could be torn into a "U" shape with a large opening to the lipidic environment, thereby forming an entry for the substrates in the ER membrane. The EM analysis showed that p97 formed a functional protein complex with Derlin-1, revealing the coupling mechanism between the ERAD retrotranslocation and the ATP hydrolysis activities.

Clinical assessment of a low-cost, hand-held, smartphone-attached intraoral imaging probe for 5-aminolevulinic acid photodynamic therapy monitoring and guidance.

Significance: India has one of the highest rates of oral squamous cell carcinoma (OSCC) in the world, with an incidence of 15 per 100,000 and more than 70,000 deaths per year. The problem is exacerbated by a lack of medical infrastructure and routine screening, especially in rural areas. New technologies for oral cancer detection and timely treatment at the point of care are urgently needed. Aim: Our study aimed to use a hand-held smartphone-coupled intraoral imaging device, previously investigated for autofluorescence (auto-FL) diagnostics adapted here for treatment guidance and monitoring photodynamic therapy (PDT) using 5-aminolevulinic acid (ALA)-induced protoporphyrin IX (PpIX) fluorescence (FL). Approach: A total of 12 patients with 14 buccal mucosal lesions having moderately/well-differentiated micro-invasive OSCC lesions (<2 cm diameter and <5 mm depth) were systemically (in oral solution) administered three doses of 20 mg/kg ALA (total 60 mg/kg). Lesion site PpIX and auto-FL were imaged using the multichannel FL and polarized white-light oral cancer imaging probe before/after ALA administration and after light delivery (fractionated, total 100 J/cm2 of 635 nm red LED light). Results: The handheld device was conducive for access to lesion site images in the oral cavity. Segmentation of ratiometric images in which PpIX FL is mapped relative to auto-FL enabled improved demarcation of lesion boundaries relative to PpIX alone. A relative FL (R-value) threshold of 1.4 was found to segment lesion site PpIX production among the patients with mild to severe dysplasia malignancy. The segmented lesion size is well correlated with ultrasound findings. Lesions for which R-value was >1.65 at the time of treatment were associated with successful outcomes. Conclusion: These results indicate the utility of a low-cost, handheld intraoral imaging probe for image-guided PDT and treatment monitoring while also laying the groundwork for an integrated approach, combining cancer screening and treatment with the same hardware.

Inter-observer agreement among specialists in the diagnosis of Oral Potentially Malignant Disorders and Oral Cancer using Store-and-Forward technology.

Oral Cancer is one of the most common causes of morbidity and mortality. Screening and mobile Health (mHealth) based approach facilitates remote early detection of Oral cancer in a resource-constrained settings. The emerging eHealth technology has aided specialist reach to rural areas enabling remote monitoring and triaging to downstage Oral cancer. Though the diagnostic accuracy of the remote specialist has been evaluated, there are no studies evaluating the consistency among the remote specialists, to the best of our knowledge. The purpose of the study was to evaluate the interobserver agreement between the specialists through telemedicine systems in real-world settings using store and forward technology. Two remote specialists independently diagnosed the clinical images from image repositories, and the diagnostic accuracy was compared with onsite specialist and histopathological diagnosis when available. Moderate agreement (k = 0.682) between two remote specialists and (k = 0.629) between the onsite specialist and two remote specialists in diagnosing oral lesions. The sensitivity and specificity of remote specialist 1 were 92.7% and 83.3%, whereas remote specialist 2 was 95.8% and 60%, respectively, compared to histopathology. The store and forward technology and telecare can be effective tools in triaging and surveillance of patients.

Structural insight into the intraflagellar transport complex IFT-A and its assembly in the anterograde IFT train.

Intraflagellar transport (IFT) trains, the polymers composed of two multi-subunit complexes, IFT-A and IFT-B, carry out bidirectional intracellular transport in cilia, vital for cilia biogenesis and signaling. IFT-A plays crucial roles in the ciliary import of membrane proteins and the retrograde cargo trafficking. However, the molecular architecture of IFT-A and the assembly mechanism of the IFT-A into the IFT trains in vivo remains elusive. Here, we report the cryo-electron microscopic structures of the IFT-A complex from protozoa Tetrahymena thermophila. We find that IFT-A complexes present two distinct, elongated and folded states. Remarkably, comparison with the in situ cryo-electron tomography structure of the anterograde IFT train unveils a series of adjustments of the flexible arms in apo IFT-A when incorporated into the anterograde train. Our results provide an atomic-resolution model for the IFT-A complex and valuable insights into the assembly mechanism of anterograde IFT trains.

The structural basis of the pH-homeostasis mediated by the Cl-/HCO3- exchanger, AE2.

The cell maintains its intracellular pH in a narrow physiological range and disrupting the pH-homeostasis could cause dysfunctional metabolic states. Anion exchanger 2 (AE2) works at high cellular pH to catalyze the exchange between the intracellular HCO3- and extracellular Cl-, thereby maintaining the pH-homeostasis. Here, we determine the cryo-EM structures of human AE2 in five major operating states and one transitional hybrid state. Among those states, the AE2 shows the inward-facing, outward-facing, and intermediate conformations, as well as the substrate-binding pockets at two sides of the cell membrane. Furthermore, critical structural features were identified showing an interlock mechanism for interactions among the cytoplasmic N-terminal domain and the transmembrane domain and the self-inhibitory effect of the C-terminal loop. The structural and cell-based functional assay collectively demonstrate the dynamic process of the anion exchange across membranes and provide the structural basis for the pH-sensitive pH-rebalancing activity of AE2.

Interpretable and Reliable Oral Cancer Classifier with Attention Mechanism and Expert Knowledge Embedding via Attention Map.

Convolutional neural networks have demonstrated excellent performance in oral cancer detection and classification. However, the end-to-end learning strategy makes CNNs hard to interpret, and it can be challenging to fully understand the decision-making procedure. Additionally, reliability is also a significant challenge for CNN based approaches. In this study, we proposed a neural network called the attention branch network (ABN), which combines the visual explanation and attention mechanisms to improve the recognition performance and interpret the decision-making simultaneously. We also embedded expert knowledge into the network by having human experts manually edit the attention maps for the attention mechanism. Our experiments have shown that ABN performs better than the original baseline network. By introducing the Squeeze-and-Excitation (SE) blocks to the network, the cross-validation accuracy increased further. Furthermore, we observed that some previously misclassified cases were correctly recognized after updating by manually editing the attention maps. The cross-validation accuracy increased from 0.846 to 0.875 with the ABN (Resnet18 as baseline), 0.877 with SE-ABN, and 0.903 after embedding expert knowledge. The proposed method provides an accurate, interpretable, and reliable oral cancer computer-aided diagnosis system through visual explanation, attention mechanisms, and expert knowledge embedding.

An updated review of feline coronavirus: mind the two biotypes.

Feline coronavirus (FCoV) includes two biotypes: feline infectious peritonitis virus (FIPV) and feline enteric coronavirus (FECV). Although both biotypes can infect cats, their pathogenicities differ. The FIPV biotype is more virulent than the FECV biotype and can cause peritonitis or even death in cats, while most FECV biotypes do not cause lesions. Even pathogenic strains of the FECV biotype can cause only mild enteritis because of their very low virulence. This article reviews recent progress in FCoV research with regard to FCoV etiological characteristics; epidemiology; clinical symptoms and pathological changes; pathogenesis; and current diagnosis, prevention and treatment methods. It is hoped that this review will provide a reference for further research on FCoV and other coronaviruses.

Exploring uncertainty measures in convolutional neural network for semantic segmentation of oral cancer images.

Significance: Oral cancer is one of the most prevalent cancers, especially in middle- and low-income countries such as India. Automatic segmentation of oral cancer images can improve the diagnostic workflow, which is a significant task in oral cancer image analysis. Despite the remarkable success of deep-learning networks in medical segmentation, they rarely provide uncertainty quantification for their output. Aim: We aim to estimate uncertainty in a deep-learning approach to semantic segmentation of oral cancer images and to improve the accuracy and reliability of predictions. Approach: This work introduced a UNet-based Bayesian deep-learning (BDL) model to segment potentially malignant and malignant lesion areas in the oral cavity. The model can quantify uncertainty in predictions. We also developed an efficient model that increased the inference speed, which is almost six times smaller and two times faster (inference speed) than the original UNet. The dataset in this study was collected using our customized screening platform and was annotated by oral oncology specialists. Results: The proposed approach achieved good segmentation performance as well as good uncertainty estimation performance. In the experiments, we observed an improvement in pixel accuracy and mean intersection over union by removing uncertain pixels. This result reflects that the model provided less accurate predictions in uncertain areas that may need more attention and further inspection. The experiments also showed that with some performance compromises, the efficient model reduced computation time and model size, which expands the potential for implementation on portable devices used in resource-limited settings. Conclusions: Our study demonstrates the UNet-based BDL model not only can perform potentially malignant and malignant oral lesion segmentation, but also can provide informative pixel-level uncertainty estimation. With this extra uncertainty information, the accuracy and reliability of the model's prediction can be improved.

Mechanistic insight into allosteric activation of human pyruvate carboxylase by acetyl-CoA.

Pyruvate carboxylase (PC) catalyzes the two-step carboxylation of pyruvate to produce oxaloacetate, playing a key role in the maintenance of metabolic homeostasis in cells. Given its involvement in multiple diseases, PC has been regarded as a potential therapeutic target for obesity, diabetes, and cancer. Albeit acetyl-CoA has been recognized as the allosteric regulator of PC for over 60 years, the underlying mechanism of how acetyl-CoA induces PC activation remains enigmatic. Herein, by using time-resolved cryo-electron microscopy, we have captured the snapshots of PC transitional states during its catalytic cycle. These structures and the biochemical studies reveal that acetyl-CoA stabilizes PC in a catalytically competent conformation, which triggers a cascade of events, including ATP hydrolysis and the long-distance communication between the two reactive centers. These findings provide an integrated picture for PC catalysis and unveil the unique allosteric mechanism of acetyl-CoA in an essential biochemical reaction in all kingdoms of life.

Diagnosis of paediatric tuberculosis by optically detecting two virulence factors on extracellular vesicles in blood samples.

Sensitive and specific blood-based assays for the detection of pulmonary and extrapulmonary tuberculosis would reduce mortality associated with missed diagnoses, particularly in children. Here we report a nanoparticle-enhanced immunoassay read by dark-field microscopy that detects two Mycobacterium tuberculosis virulence factors (the glycolipid lipoarabinomannan and its carrier protein) on the surface of circulating extracellular vesicles. In a cohort study of 147 hospitalized and severely immunosuppressed children living with HIV, the assay detected 58 of the 78 (74%) cases of paediatric tuberculosis, 48 of the 66 (73%) cases that were missed by microbiological assays, and 8 out of 10 (80%) cases undiagnosed during the study. It also distinguished tuberculosis from latent-tuberculosis infections in non-human primates. We adapted the assay to make it portable and operable by a smartphone. With further development, the assay may facilitate the detection of tuberculosis at the point of care, particularly in resource-limited settings.

Field validation of deep learning based Point-of-Care device for early detection of oral malignant and potentially malignant disorders.

Early detection of oral cancer in low-resource settings necessitates a Point-of-Care screening tool that empowers Frontline-Health-Workers (FHW). This study was conducted to validate the accuracy of Convolutional-Neural-Network (CNN) enabled m(mobile)-Health device deployed with FHWs for delineation of suspicious oral lesions (malignant/potentially-malignant disorders). The effectiveness of the device was tested in tertiary-care hospitals and low-resource settings in India. The subjects were screened independently, either by FHWs alone or along with specialists. All the subjects were also remotely evaluated by oral cancer specialist/s. The program screened 5025 subjects (Images: 32,128) with 95% (n = 4728) having telediagnosis. Among the 16% (n = 752) assessed by onsite specialists, 20% (n = 102) underwent biopsy. Simple and complex CNN were integrated into the mobile phone and cloud respectively. The onsite specialist diagnosis showed a high sensitivity (94%), when compared to histology, while telediagnosis showed high accuracy in comparison with onsite specialists (sensitivity: 95%; specificity: 84%). FHWs, however, when compared with telediagnosis, identified suspicious lesions with less sensitivity (60%). Phone integrated, CNN (MobileNet) accurately delineated lesions (n = 1416; sensitivity: 82%) and Cloud-based CNN (VGG19) had higher accuracy (sensitivity: 87%) with tele-diagnosis as reference standard. The results of the study suggest that an automated mHealth-enabled, dual-image system is a useful triaging tool and empowers FHWs for oral cancer screening in low-resource settings.

Interpretable deep learning approach for oral cancer classification using guided attention inference network.

SIGNIFICANCE: Convolutional neural networks (CNNs) show the potential for automated classification of different cancer lesions. However, their lack of interpretability and explainability makes CNNs less than understandable. Furthermore, CNNs may incorrectly concentrate on other areas surrounding the salient object, rather than the network's attention focusing directly on the object to be recognized, as the network has no incentive to focus solely on the correct subjects to be detected. This inhibits the reliability of CNNs, especially for biomedical applications. AIM: Develop a deep learning training approach that could provide understandability to its predictions and directly guide the network to concentrate its attention and accurately delineate cancerous regions of the image. APPROACH: We utilized Selvaraju et al.'s gradient-weighted class activation mapping to inject interpretability and explainability into CNNs. We adopted a two-stage training process with data augmentation techniques and Li et al.'s guided attention inference network (GAIN) to train images captured using our customized mobile oral screening devices. The GAIN architecture consists of three streams of network training: classification stream, attention mining stream, and bounding box stream. By adopting the GAIN training architecture, we jointly optimized the classification and segmentation accuracy of our CNN by treating these attention maps as reliable priors to develop attention maps with more complete and accurate segmentation. RESULTS: The network's attention map will help us to actively understand what the network is focusing on and looking at during its decision-making process. The results also show that the proposed method could guide the trained neural network to highlight and focus its attention on the correct lesion areas in the images when making a decision, rather than focusing its attention on relevant yet incorrect regions. CONCLUSIONS: We demonstrate the effectiveness of our approach for more interpretable and reliable oral potentially malignant lesion and malignant lesion classification.

Bayesian deep learning for reliable oral cancer image classification.

In medical imaging, deep learning-based solutions have achieved state-of-the-art performance. However, reliability restricts the integration of deep learning into practical medical workflows since conventional deep learning frameworks cannot quantitatively assess model uncertainty. In this work, we propose to address this shortcoming by utilizing a Bayesian deep network capable of estimating uncertainty to assess oral cancer image classification reliability. We evaluate the model using a large intraoral cheek mucosa image dataset captured using our customized device from high-risk population to show that meaningful uncertainty information can be produced. In addition, our experiments show improved accuracy by uncertainty-informed referral. The accuracy of retained data reaches roughly 90% when referring either 10% of all cases or referring cases whose uncertainty value is greater than 0.3. The performance can be further improved by referring more patients. The experiments show the model is capable of identifying difficult cases needing further inspection.

Zipper head mechanism of telomere synthesis by human telomerase.

Telomerase, a multi-subunit ribonucleoprotein complex, is a unique reverse transcriptase that catalyzes the processive addition of a repeat sequence to extend the telomere end using a short fragment of its own RNA component as the template. Despite recent structural characterizations of human and Tetrahymena telomerase, it is still a mystery how telomerase repeatedly uses its RNA template to synthesize telomeric DNA. Here, we report the cryo-EM structure of human telomerase holoenzyme bound with telomeric DNA at resolutions of 3.5 Å and 3.9 Å for the catalytic core and biogenesis module, respectively. The structure reveals that a leucine residue Leu980 in telomerase reverse transcriptase (TERT) catalytic subunit functions as a zipper head to limit the length of the short primer-template duplex in the active center. Moreover, our structural and computational analyses suggest that TERT and telomerase RNA (hTR) are organized to harbor a preformed active site that can accommodate short primer-template duplex substrates for catalysis. Furthermore, our findings unveil a double-fingers architecture in TERT that ensures nucleotide addition processivity of human telomerase. We propose that the zipper head Leu980 is a structural determinant for the sequence-based pausing signal of DNA synthesis that coincides with the RNA element-based physical template boundary. Functional analyses unveil that the non-glycine zipper head plays an essential role in both telomerase repeat addition processivity and telomere length homeostasis. In addition, we also demonstrate that this zipper head mechanism is conserved in all eukaryotic telomerases. Together, our study provides an integrated model for telomerase-mediated telomere synthesis.

The structural basis for the phospholipid remodeling by lysophosphatidylcholine acyltransferase 3.

As the major component of cell membranes, phosphatidylcholine (PC) is synthesized de novo in the Kennedy pathway and then undergoes extensive deacylation-reacylation remodeling via Lands' cycle. The re-acylation is catalyzed by lysophosphatidylcholine acyltransferase (LPCAT) and among the four LPCAT members in human, the LPCAT3 preferentially introduces polyunsaturated acyl onto the sn-2 position of lysophosphatidylcholine, thereby modulating the membrane fluidity and membrane protein functions therein. Combining the x-ray crystallography and the cryo-electron microscopy, we determined the structures of LPCAT3 in apo-, acyl donor-bound, and acyl receptor-bound states. A reaction chamber was revealed in the LPCAT3 structure where the lysophosphatidylcholine and arachidonoyl-CoA were positioned in two tunnels connected near to the catalytic center. A side pocket was found expanding the tunnel for the arachidonoyl CoA and holding the main body of arachidonoyl. The structural and functional analysis provides the basis for the re-acylation of lysophosphatidylcholine and the substrate preference during the reactions.