Progressive familial intrahepatic cholestasis type 3 (PFIC3) is a rare liver disorder caused by biallelic mutations in the ABCB4 gene, leading to multidrug resistance protein 3 (MDR3) deficiency. PFIC3 often presents with clinical and biochemical features that overlap with Wilson's disease (WD), including hepatic copper accumulation and elevated urinary copper excretion. These similarities contribute to frequent misdiagnosis, resulting in inappropriate chelation therapy and delayed appropriate management. This systematic review examines reported cases of PFIC3 initially misdiagnosed as WD to highlight diagnostic challenges and assess patient outcomes. A comprehensive search across PubMed, ScienceDirect and Google Scholar identified 11 eligible studies involving 16 patients. Most cases were first treated as WD, receiving chelation therapy without clinical improvement. Diagnosis was later revised to PFIC3 following negative ATP7B mutation testing and identification of ABCB4 variants, often via whole-genome sequencing. Upon switching to ursodeoxycholic acid (UDCA), most patients experienced clinical stabilization. The findings underscore the need for heightened awareness of PFIC3 as a differential diagnosis in atypical WD cases, especially when ceruloplasmin is normal and Kayser-Fleischer (KF) rings are absent. Early genetic testing is essential to avoid mismanagement. Further observational studies are warranted to estimate misdiagnosis frequency and guide diagnostic protocols.

VEXAS syndrome (vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic) is an X-linked, systemic, haemato-inflammatory syndrome caused by somatic mutations in the UBA1 gene. No standardized treatment guidelines exist, but evidence is emerging that treatment with hypomethylating agents (HMAs) can induce improvement of the inflammatory symptoms, reverse cytopenia, enable weaning of the corticosteroids and, in some cases, induce molecular remission. We systematically reviewed the literature to evaluate the experience with HMA in the management of VEXAS, following 2020 PRISMA guidelines across three databases. A total of 30 citations reporting 166 patients with genetically confirmed VEXAS syndrome treated with HMA were included. All but three patients were males (98%), and the median age was 71 years (range, 29-86). Inflammatory symptoms (96% of patients), constitutional symptoms (84%) and cytopenia (89%) were most frequently reported, and 81% had a concomitant diagnosis of myelodysplastic syndromes (MDS). Most mutations were substitutions of p.Met41 (81%). Overall inflammatory response was achieved in 59% of treated patients with most complete response (52%). Responses were documented in patients with and without concomitant MDS. Any haematological response was achieved in 74% of the patients, and eradication of the UBA1 clone (to a variant allele fraction of <2%) was detected in 51% of cases in whom mutation testing was performed (n = 32/63). Toxicity was the most frequent reason for discontinuation of therapy. This review highlights HMAs as a feasible option in the management of VEXAS syndrome. Prospective studies are needed to identify predictors for response or resistance and optimal regimens.

Herbal medicinal products (HMPs) prepared from the fruit of Silybum marianum are mainly used in the treatment of liver diseases and dyspeptic symptoms. Despite its long-standing use and good clinical safety profile, the safety assessment of such HMPs requires special attention to toxicological aspects that are difficult to detect clinically, especially genotoxic effects. The genotoxic potential of certain S. marianum extracts has been evaluated previously; however, some in vitro assays gave inconsistent results, which is why an EU list entry was not recommended by the Herbal Medicinal Product Committee (HMPC) of the European regulatory agency EMA. To provide a more comprehensive dataset for the evaluation of the genotoxic potential, six dry extracts covering the entire polarity range of extraction solvents were chosen in accordance with the "bracketing and matrixing" approach recommended by the HMPC. These extracts were subjected to the bacterial reverse mutation test, as specified in the OECD test guideline 471, which was performed both as a pre-incubation and plate incorporation assay. When testing up to 5 mg per plate or up to the solubility or cytotoxicity limits, none of the extracts showed signs of mutagenicity, suggesting that extracts of S. marianum fruit have no genotoxic potential. Within reasonable limits, these results may be extrapolated to other extracts based on the aforementioned "bracketing and matrixing" approach.

Neisseria gonorrhoeae is the bacterial agent responsible for gonorrhea, a common sexually transmitted infection. The emergence of Neisseria gonorrhoeae multidrug-resistant (MDR) strains a presents a critical public health threat, especially due to its contribution to antimicrobial resistance (AMR) and treatment failure. Currently, resistance profiling of N. gonorrhoeae relies on phenotypic methods such as minimum inhibitory concentration (MIC) testing and identification of known resistance mutations. These are with limited application of genome-wide approaches to understand resistance evolution. The lack of genomic epidemiology data among Low- and Middle-Income Countries (LMICs) regions such as Kenya, hampers effective AMR tracking and designation of evidence-based, targeted treatments. This study aims to investigate the genetic diversity, population structure, and recombination dynamics of MDR N. gonorrhoeae isolates from Kenya using whole-genome SNP analysis. A total of 92 genomes (72 FASTQ reads and 20 assembled genomes) were retrieved from NCBI. De novo assembly, identification of AMR genes and variant calling were conducted, followed by Principal Component Analysis (PCA), Neighbor-Net clustering, nucleotide diversity (π), and linkage disequilibrium (LD) decay analysis to assess population structure and recombination patterns. Our results revealed a predominant genetic cluster with several divergent outlier strains, indicating moderate population differentiation. Despite the lack of strong geographic separation in the overall genomic structure, significant regional differences were observed in antimicrobial resistance gene burden. Western Kenyan regions (Kisumu and Kombewa) exhibited higher AMR gene counts despite genetic similarity to isolates from other regions, suggesting that local antibiotic selection pressures, rather than population isolation, are driving the accumulation of resistance determinants. Across the genome, nucleotide diversity was variable with distinct recombination hotspots. A rapid LD decay within the first 1000 bp suggested a high overall recombination rate. These results indicate that recombination plays a pivotal role in shaping genetic variability and AMR evolution in N. gonorrhoeae populations. The absence of strong geographic structure further implies that transmission dynamics, rather than regional isolation, drive the spread of resistance. In conclusion, genome-wide SNP analysis offers valuable insights into the genetic diversity and populations structure of MDR N. gonorrhoeae in Kenya. These findings support the integration of genomic surveillance into national strategies for antimicrobial resistance control. This study reveals how genetic analysis can guide better strategies to track and control drug-resistant gonorrhea in Kenya.

A 34-year-old man from Guinea-Bissau presented with a rapidly growing tumor on his right foot over 6 months, arising from a long-standing pigmented papule present for more than 20 years. Clinical examination revealed an exophytic black tumor with ipsilateral inguinal lymphadenopathy. Histopathological analysis showed an epithelioid and fusiform melanocytic proliferation in the dermis with abundant melanin. Immunohistochemical staining demonstrated diffuse positivity for Sox-10, Melan-A, HMB45, and PRAME. A diagnosis of blue nevus–like melanoma (BNM) was confirmed. BRAF V600 mutation testing on a cutaneous metastasis was negative. Staging imaging revealed widespread metastases involving lymph nodes, brain, lungs, liver, stomach, duodenum, colon, adrenal glands, pancreas, gallbladder, and bones. Despite 4 cycles of ipilimumab/nivolumab immunotherapy, the disease progressed, and the patient died 4 months after diagnosis. BNM is a rare melanoma variant, often arising from a pre-existing blue nevus, with aggressive potential and frequent lymph node metastasis. The molecular heterogeneity of BNM and the limited therapeutic options underscore the need for further research into targeted therapies and prognostic biomarkers for this rare melanoma subtype.

Sodium hypochlorite is a well-established biocide extensively used in both living and industrial environments; however, exposure to it or improper handling can harm the human body and the environment. Thus, safer alternatives to sodium hypochlorite are required. Therefore, the objective of this research was to assess the genotoxicity, cardiotoxicity, 2-week repeated-dose toxicity studies (oral, dermal), and acute irritation/corrosion of sodium dodecanoyloxybenzenesulfonate (LOBS), a new alternative biocide in various biological systems. LOBS did not show genotoxic or cardiotoxic effects in the bacterial reverse mutation test, chromosomal aberration test, or hERG assay. Both males and females showed an increase in U-Na levels (p < 0.01), while an increase in focus/foci, thickness in the stomach, and discoloration was observed in the thymus of males during the 2-week repeated oral-dose toxicity study. In the 2-week repeated dermal-dose toxicity study, no significant adverse effects were observed. While mild erythema and edema were observed in all animals during the acute dermal irritation/corrosion study, all the animals recovered. During the acute eye irritation/corrosion study, all animals showed mild corneal clouding, conjunctival redness, and conjunctival chemosis; however, all animals recovered. Moreover, LOBS was classified into category 3 (mild irritant) and category 2A (irritant) in the acute dermal and eye irritation/corrosion studies. Therefore, LOBS should be further developed as a potential alternative biocide to sodium hypochlorite.

Liquid biopsy and tissue biopsy for the detection of EGFR mutations in patients with stage III non-small-cell lung cancer: an observational real-world study.

Polycythemia vera (PV) is a myeloproliferative neoplasm. The presence of JAK2 mutations is a major diagnostic criterion for PV. PV is linked to chronic inflammation and an increased risk of thrombosis, and inflammation plays a significant part in the pathophysiology of PV. Testing for JAK2 mutations is expensive and is not available in all laboratories. Simple inflammatory markers, including the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and systemic immune-inflammation index (SII), are evaluated in this study as potential diagnostic markers for differentiating patients with PV from other patients with polycythemia. We conducted a retrospective study of the clinical and laboratory data from 281 patients with polycythemia (110 with PV and 181 with secondary polycythemia (SP)) who attended Ghaem Hospital. The diagnosis of PV was established based on the World Health Organization criteria. Individuals who did not meet the criteria were classified as having SP. The median NLR, PLR, and SII in the PV group were considerably elevated compared to the SP group (NLR: 5.00 vs. 1.86, PLR: 261.3 vs. 94.0, SII: 2432.9 vs. 368.8, p < 0.001 for all). The receiver operating characteristic analysis revealed that NLR, PLR, and SII were highly effective in differentiating PV patients from the SP group. Each of these tests showed sensitivities and specificities over 85% and an area under the curve of more than 0.9. SII, NLR, and PLR were all higher in PV than SP, suggesting that these biomarkers, particularly SII, might be helpful in the diagnosis of PV.

The most common genetic cause associated with thrombophilia is Leiden mutation (G1691A) of the coagulation factor V (F5) gene. Data collected anonymously from 355 unrelated Greeks examined for the above mutation were analyzed. Bioinformatic investigation was conducted for factor V, including phylogenetic analysis, genetic network analysis, and 3D modeling of wild-type and Leiden protein. Analysis confirmed the importance of F5 Leiden in thrombosis and the significance of a positive family history of thrombosis. In silico analysis of the F5 Leiden revealed a linkage between mammalian species and a potential functional interaction of F5 with 25 other genes, several of which have been associated with cardiovascular diseases. 3D modeling revealed that the Leiden mutation confers an H-bond network alteration in the functional region of FV, resulting in a hypercoagulable state. This study highlighted the great value of a positive family history of thrombosis and the importance of testing for this common mutation as prevention strategy components for thrombophilia, including thrombotic brain aneurysms. Bioinformatic analysis indicated the importance of the Leiden mutation in protein structure and function. The findings of this study increase understanding of molecular and clinical features of thrombophilia and may eventually lead to better prevention of thrombosis.

The most common genetic cause associated with thrombophilia is Leiden mutation (G1691A) of the coagulation factor V (FV) (F5) gene. Data collected anonymously from 355 unrelated Greeks examined for the above mutation were analyzed. Bioinformatic investigation was conducted for factor V, including phylogenetic analysis, genetic network analysis, and 3D modelling of wild-type and Leiden protein. Analysis confirmed the importance of F5 Leiden in thrombosis and the significance of a positive family history of thrombosis. In silico analysis of the F5 Leiden revealed a linkage between mammalian species and a potential functional interaction of F5 with 25 other genes, several of which have been associated with cardiovascular diseases. 3D modelling revealed that the Leiden mutation confers an H-bond network alteration in the functional region of FV, resulting in a hypercoagulable state. This study highlighted the great value of a positive family history of thrombosis and the importance of testing for this common mutation as a prevention strategy component for thrombophilia, including thrombotic brain aneurysms. Bioinformatic analysis indicated the importance of the Leiden mutation in protein structure and function. The findings of this study increase understanding of molecular and clinical features of thrombophilia and may eventually lead to better prevention of thrombosis.

OA03.02: A Nationwide Study on Driver Mutation Testing and Targeted Therapies in NSCLC patients with comorbid Interstitial Pneumonia

Mutations can take many different forms, such as chromosomal rearrangements, variations in the number of chromosomes, and gene (point) mutations. These changes can be found in both mammalian and bacterial cells, frequently by microscopic examination or growth needs. To ensure safety, it is crucial to evaluate a chemical's mutagenic potential. One popular technique in toxicology is the Ames test. This bacterial gene mutation assay, created by Bruce N. Ames in 1975, is robust, dependable, and reasonably priced. To discover reverse mutations that restore their capacity to generate critical amino acids, it uses strains of Salmonella typhimurium or Escherichia coli that have been modified to be amino acid-dependent. The plate incorporation assay and the preincubation assay are the two main methods used in the Ames test. While the preincubation assay necessitates a preparatory incubation step before plating, the plate integration method mixes the test substance with bacteria and an S9 metabolic activation system before plating. By assessing the development of revertant colonies, these assays shed light on the mutagenic qualities of substances. Standardized procedures have been set up by international regulatory organizations like the Organization for Economic Co-operation and Development (OECD) and International Council for Harmonization (ICH) to guarantee uniformity in mutagenicity evaluations. The test makes use of particular bacterial strains that are intended to identify various changes, such as frameshift and base-pair substitution mutations. The Ames test continues to be a fundamental component of genetic toxicology, aiding in the assessment of industrial chemicals, medicines, and environmental agents by offering a trustworthy screening method for chemical mutagenicity.

This chapter describes a protocol modification of the bacterial reverse mutation test (Ames test). It is based on the same principle as the Ames test in Petri dishes, but uses a liquid, low volume microplate version of the fluctuation method. Clear strengths are the low compound requirement, and the increased throughput as compared to the traditional format in Petri dishes. The liquid system allows for processing several replicates at once with the possibility of using pipetting robots and has an easy colorimetric read-out instead of counting revertant colonies. The Ames II and MPF™ also use less S9 and produce less hazardous waste due to the low volume multi-well format.

The incidence of chronic lymphocytic leukemia (CLL) is rising in Korea; however, clinical management often diverges from international guidelines due to limited clinical experience, restricted access to diagnostics, and delayed reimbursement for novel agents. This study aimed to assess Korean hematologists' awareness, clinical practices, and perceived barriers in the management of CLL. A nationwide, web-based survey was conducted between May 29 and June 19, 2023, targeting hematologists registered with the Korean Society of Hematology who were actively treating patients with CLL. The 15-item questionnaire addressed clinical experience, treatment approaches, use of bruton tyrosine kinase inhibitor (BTKi), awareness and application of prognostic tools, access to molecular diagnostics, reimbursement priorities, and perceived need for national clinical guidelines. A total of 89 hematologists completed the survey. Patient caseloads varied, with 41.6% of physicians managing 10-19 patients over the prior six months. Treatment patterns were heterogeneous, and undertreatment was commonly attributed to patient refusal (33.7%) and advanced age (10.1%). Despite reimbursement limitations, 15.7% reported prescribing BTKis in ≥ 40% of their patients, although adverse events and intolerance were frequently cited challenges. Awareness of prognostic indices was high, yet their implementation was inconsistent. Access to essential molecular diagnostics was suboptimal; 53% lacking Immunoglobulin heavy-chain variable region gene (IGHV) mutation testing and 43% lacked measurable residual disease (MRD) assessment, whereas TP53 testing was broadly available. Nearly half of the respondents prioritized second-generation BTKis for first-line reimbursement, and 94.4% endorsed the need for Korean-specific clinical guidelines. This survey highlights gaps between international recommendations and Korean real-world practice, emphasizing the need for improved diagnostic availability, timely reimbursement of targeted agents, and development of Korea-specific clinical guidelines tailored to the Korean context.

Fault concealment in complex software programs and the difficulty of generating test cases to detect such faults present significant challenges in software testing. To resolve these challenges, this paper suggests a novel method that integrates mutation testing, fuzzy clustering, convolutional neural networks (CNN), and particle swarm optimization (PSO) to efficiently generate test cases that cover multiple paths with numerous faults (mutant branches). Initially, mutation-based paths are classified using fuzzy clustering based on their coverage difficulty and similarity. A multi-feature CNN model (MF_CNNpro) is then constructed and trained on the paths of each cluster. Finally, the predicted particles from the MF_CNNpro model are used as the initial population for PSO, which evolves to generate the test cases. The proposed method is evaluated on six test programs, and the results demonstrate that it significantly improves clustering separation and reduces clustering compactness. By selecting only the cluster center paths to construct the MF_CNNpro model, training and prediction costs are effectively reduced. Moreover, the use of MF_CNNpro and PSO to select representative individuals as the initial population greatly enhances the evolutionary efficiency of PSO. The proposed method outperforms traditional approaches in clustering, prediction, and test data generation. Specifically, the SC clustering method improves cluster separation (SP) by 0.021, reduces compactness (CP) by 0.054, and decreases clustering rate (CR) by 4.97%, thereby enhancing clustering precision. The MF_CNNpro model improves the IA metric by 38.2% and reduces the U-Statistic and MSE by 83.0% and 97.9%, respectively, optimizing prediction performance. The MF_CNNpro+PPSOpro method increases the path coverage success rate from 47.9% to 97.4% (a 103.3% improvement), reduces the number of iterations by 84.1%, and decreases execution time by 95.6%, significantly improving generation efficiency.

Introduction and aimLung cancer is the most common cancer and the leading cause of cancer deaths worldwide; its incidence in the female population is increasing. The objective of this study was to describe the epidemiological, clinical, paraclinical, therapeutic and evolutionary profile of female patients with non-small cell lung cancer (NSCLC) and to research the predictive factors of better survival.MethodsA retrospective cohort study was carried out among women with NSCLC, diagnosed between January 2019 and December 2023, at the National Institute of Oncology, Rabat, Morocco. The diagnosis was proven by pathological analysis of biopsy. Statistical analysis was carried out using Jamovi software version 2.5.3.ResultsWe assessed 146 women (18.3%) among 805 cases with NSCLC; the average age was 63.5 ± 12.3 years. The percentage of non-smokers, passive smokers and active smokers was 87.7%, 7.5% and 4.8%, respectively. Cough (56.2%), chest pain (21.2%) and dyspnea (19.9%) were the major clinical signs. The common histological types were adenocarcinoma (90.4%), squamous cell carcinoma (6.8%) and other types (2.8%). The primary tumor site was in the right lung in 78 patients (53.4%), on the left in 65 cases (44.5%) and in three cases (2.1%) bilateral. Most patients, 87% (127 cases), were diagnosed at stage IV according to the TNM classification 8th edition. The epithelial growth factor receptor (EGFR) mutation test performed in 79 female patients (54% of cases) was positive in 28 patients (35.4%). Mutations were predominant on exon 19 (24 patients or 86%), followed by the L858 mutation on exon 21 (four patients or 14%). The anaplastic lymphoma kinase mutation, which was sought in 65 patients (44% of cases), was positive in six patients (9.2%). The search for the reactive oxygen species (ROS1) mutation was performed in seven patients (4.7%) and was positive in one case (14.2%). Programmed death-ligand 1 (PD-L1) expression was affected in 67 patients (45.8%); it was <50% in 57 patients representing 85% of cases and ≥50% was noted in 10 patients (15%).106 stage IV female patients (72.6%) had received first-line palliative systemic treatment, including 89 cases (84%) who had received platinum-based chemotherapy (including 11 EGFR mutated) and 17 EGFR mutated (16%) received anti-EGFR treatment.In our series, the median overall survival (OS) was 17 months (12.23) and the median progression-free survival (PFS) was 11 months (6.12). Median OS of EGFR-mutated female patients treated in first-line with EGFR TKIs (tyrosine kinase inhibitors) was similar compared with that of EGFR-mutated patients treated with platinum-based chemotherapy (25 months), and median PFS in these patients was slightly longer but statistically non-significant compared with that of patients treated with platinum-based chemotherapy (9.5 months vs. 6.5 months, p=0.22).ConclusionThe population of women with NSCLC studied was characterized by non-smoking, adenocarcinoma histological type dominance, and a high percentage of EGFR mutations. The median PFS of patients treated with EGFR TKIs was slightly longer than that of patients who received platinum-based chemotherapy. These results encourage us to perform EGFR mutation testing in all patients diagnosed with adenocarcinoma so that they can receive targeted therapy with EGFR TKIs.

COVID-19 has invaded human community worldwide threatening public health and the world economy. The confounder data about the pathogenesis of the disease progression has aroused prognostic theories concerning the implications of demographic data and the genetic variability of the disease prognosis. A total of 120 participants; 68 males and 52 females, aged from 15 to 82 years, among them 90 COVID-19-infected patients, and 30 COVID-19-free participants were enrolled in the current study. Blood samples were drawn from 120 participants and then analyzed for D-dimer levels using fluorescence immunoassay. The 120 DNA extracted samples were amplified by PCR using Factor V-specific primers. The PCR products were sequenced and examined for putative mutations in the Factor V gene. A significantly high mutation rate in Factor V (rs6687813) 82%, p&lt; 0.001 was demonstrated in the study population. The mutations were significantly correlated with extremely high D-dimer levels of 2975 ng/ml versus 986 ng/ml in the wild type (r=0.228, p&lt;0.012). Hospitalized COVID-19 patients expressed significantly higher levels of D-dimer compared to non-hospitalized patients 3353 ng/ml vs 513 ng/ml p&lt;0.001, respectively. The age had an odds rate of 1.12 on the acquisition of COVID-19 infection as the ages above 35 were more vulnerable to acquiring the infection than younger ages (r=0.21, p&lt;0.022). In conclusion, the study provides robust evidence about the direct association between the FVL mutation and elevated levels of D-dimer in COVID-19 patients and the implementation of the FVL mutation test as a prognostic marker is recommended for COVID-19 patients.

Metastatic castration-resistant prostate cancer (mCRPC) remains atherapeutic challenge. Poly(adenosine diphosphate-ribose) polymerase inhibitors (PARPi) combined with new hormonal agents (NHA) offer novel treatment options. This review summarizes the current status of PARPi +NHA combination therapy in mCRPC. Summary of relevant phaseII andIII trials on PARPi +NHA and the G‑BA (Gemeinsame Bundesausschuss) decision as well as the current S3guideline recommendations. PARPi +NHA demonstrated improved efficacy compared to NHA alone in an all-comers population that received prior androgen deprivation therapy (ADT) or docetaxel therapy. In particular the subgroup of patients with homologous recombination repair (HRR) and breast cancer (BRCA)1/2 mutations had the best outcomes. Olaparib +abiraterone, talazoparib +enzalutamide, and niraparib +abiraterone are approved combinations, expanding treatment options in mCRPC. PARPi +NHA represent asignificant advance in mCRPC therapy. Molecular genetic testing for HRR mutations, especially BRCA1/2, is crucial for treatment planning.

Accurately distinguishing between benign and malignant thyroid nodules(TNs) in the context of Hashimoto's thyroiditis (HT) is challenging. This study aimed to explore the diagnostic efficacy of artificial intelligence (AI) models constructed using radiomics and deep learning (DL) features extracted from the ultrasound images of TNs in the setting of HT. The study also aimed to quantitatively compare the diagnostic performance of these models against that of fine-needle aspiration (FNA) cytology combined with BRAFV600E gene mutation testing so as to establish a superior diagnostic paradigm for TNs in HT. We analyzed the clinical data and preoperative ultrasound images of 1585 patients with HT admitted to 8 hospitals in China between 1 January 2018, and 30 December 2023. Radiomics features were extracted from each manually annotated and standardized region of interest in the images. The DL features based on various convolutional neural network models were also extracted, including 11 pre-trained DL models based on real images and the image features extracted using a combined DL-radiomics (DLR) approach. Least absolute shrinkage and selection operator regression was used to select features with nonzero coefficients. Further, eight machine learning methods were employed to construct prediction models, namely the DLR models, incorporating both DL and radiomics features. The importance of features in contributing to the model was prioritized using the SHapley Additive exPlanations (SHAP) method for interpretation. The predictive performance of the models was evaluated using the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity. Two rounds of reader studies were conducted (first round: independent reading; second round: guided reading) to validate the clinical application value of the DLR model, with results compared with FNA biopsy outcomes. A total of 1561 radiomics features and 256 DL features were extracted from the original images. The DLR model, leveraging the ResNet152 neural network, could effectively differentiate between benign and malignant nodules in the context of HT. The AUC, accuracy, sensitivity, and specificity of the logistic regression (LR) model, based on these features, were 0.917 [95% confidence interval (CI): 0.838-0.988], 85.7%, 81.8%, and 86.6%, respectively, in the validation cohort and 0.827(95% CI: 0.777-0.876), 93.9%, 80.9%, and 96.3%, respectively, in the external test cohort. The SHAP summary plot illustrated how feature values influenced their impact on the model, whereas the SHAP force plot showed the integrated impact of features on individual responses. Gradient-weighted class activation mapping (Grad-CAM)-generated heatmaps from DL models visually highlighted high-risk areas in HT. The DLR model outperformed five junior-level physicians in terms of diagnostic accuracy, sensitivity, and specificity in the validation cohort. The diagnostic performance of all clinicians was significantly enhanced with the assistance of the DLR model, with no statistical difference detected compared with the FNA biopsy results. The DLR model combined with the SHAP and Grad-CAM method can improve the diagnostic performance of radiologists in identifying benign and malignant TNs in the context of HT. The diagnostic efficacy of this visualization model is comparable to that of FNA cytology combined with gene mutation testing. The DLR model can enhance the diagnostic ability of radiologists in differentiating between benign and malignant TNs in the context of HT, thereby minimizing unnecessary biopsies. Additionally, it can aid clinicians in making personalized decisions regarding the necessity of biopsy or even surgery by providing intuitive visual explanations.

Ensuring correct execution of programs running on today’s parallel systems becomes difficult when memory is shared across several processing units. Memory consistency models (MCMs) were defined to provide a contract between different levels of the hardware-software stack to specify shared memory access orderings for correct implementations. However, instruction set architecture (ISA) MCMs traditionally only reason about the program-visible impacts of shared memory accesses for user-facing program instructions. In virtual memory systems though, there are additional hardware and operating system (OS) level shared memory accesses that can occur to facilitate address translation and may impact the program execution. Memory transistency models (MTMs) were thus coined to define a superset of MCMs that additionally account for underlying virtual memory operations. However, MTM implementations are complex as they are managed across the hardware and OS, making them difficult to specify. In such cases, empirical testing is the usual approach for effective validation of a specification. However, empirical MTM testing is challenging due to the complex ordering relationships between hardware, OS, and user-level operations, as well as the inability to explicitly run and control virtual memory operations with user-level programs. While many MCM testing tools exist and have been used to uncover system bugs, there is no existing work on developing systematic MTM testing techniques, nor has there been any analysis of how effective such techniques are. In this work, we introduce TEMpesT, the first full-system framework that uses ISA-agnostic techniques for performing fuzz testing for MTMs across both hardware and OS levels. TEMpesT uses enhanced litmus tests (ELTs) with novel targeted user-level MTM fuzzing techniques to induce virtual memory operations and coax out corner case behaviors. We used TEMpesT to validate the formal MTM specification \({\tt x86t{_{-}\!}elt}\) on an Intel x86 system running Linux. Our results show TEMpesT’s techniques are able to induce high outcome variety, resulting in 94% of all possible outcomes being observed with just 10,000 iterations of each ELT synthesized for \({\tt x86t{_{-}\!}elt}\) , outperforming prior methodology by 5.8×. This article also introduces MTM mutation tests that we used to evaluate TEMpesT’s fuzzing techniques and demonstrate effective MTM bug detection.