Multiparametric MR characterization for human epithelial growth factor receptor 2 expression in bladder cancer: an exploratory study.

PURPOSE: To investigate the application value of multiparametric MRI in evaluating the expression status of human epithelial growth factor receptor 2 (HER2) in bladder cancer (BCa). METHODS: From April 2021 to July 2023, preoperative imaging manifestations of 90 patients with pathologically confirmed BCa were retrospectively collected and analyzed. All patients underwent multiparametric MRI including synthetic MRI, DWI, from which the T1, T2, proton density (PD) and apparent diffusion coefficient (ADC) values were obtained. The clinical and imaging characteristics as well as quantitative parameters (T1, T2, PD and ADC values) between HER2-positive and -negative BCa were compared using student t test and chi-square test. The diagnostic efficacy of parameters in predicting HER2 expression status was evaluated by calculating the area under ROC curve (AUC). RESULTS: In total, 76 patients (mean age, 63.59 years ± 12.84 [SD]; 55 men) were included: 51 with HER2-negative and 25 with HER2-positive BCa. HER2-positive group demonstrated significantly higher ADC, T1, and T2 values than HER2-negative group (all P < 0.05). The combination of ADC values and tumor grade yielded the best diagnostic performance in evaluating HER2 expression level with an AUC of 0.864. CONCLUSION: The multiparametric MR characterization can accurately evaluate the HER2 expression status in BCa, which may further guide the determination of individualized anti-HER2 targeted therapy strategies.

Molecular evidence provides new insights into the evolutionary origin of an ancient traditional Chinese medicine, the domesticated "Baizhi".

Introduction: "Baizhi" is a famous herbal medicine in China, and it includes four landraces named as 'Hangbaizhi', 'Chuanbaizhi', 'Qibaizhi', and 'Yubaizhi'. Long-term artificial selection had caused serious degradation of these germplasms. Determining the wild progenitor of the landraces would be benefit for their breed improvements. Previous studies have suggested Angelica dahurica var. dahurica, A. dahurica var. formosana, or A. porphyrocaulis as potential candidates, but the conclusion remains uncertain, and their phylogenetic relationships are still in controversy. Methods: In this study, the genetic variation and phylogenetic analyses of these species and four landraces were conducted on the basis of both the nrITS and plastome datasets. Results: Genetic variation analysis showed that all 8 population of four landraces shared only one ITS haplotype, meanwhile extremely low variation occurred within 6 population at plastid genome level. Both datasets supported the four landraces might be originated from a single wild germplasm. Phylogenetic analyses with both datasets revealed largely consistent topology using Bayesian inference and Maximum likelihood methods. Samples of the four landraces and all wild A. dahurica var. dahurica formed a highly supported monophyletic clade, and then sister to the monophyly clade comprised by samples of A. porphyrocaulis, while four landraces were clustered into one clade, which further clustered with a mixed branches of A. porphyrocaulis and A. dahurica var. dahurica to form sister branches for plastid genomes. Furthermore, the monophyletic A. dahurica var. formosana was far distant from the A. dahurica var. dahurica-"Baizhi" clade in Angelica phylogeny. Such inferences was also supported by the evolutionary patterns of nrITS haplotype network and K2P genetic distances. The outcomes indicated A. dahurica var. dahurica is most likely the original plant of "Baizhi". Discussion: Considering of phylogenetic inference and evolutionary history, the species-level status of A. dahurica var. formosana should be accepted, and the taxonomic level and phylgenetic position of A. porphyrocaulis should be further confirmed. This study preliminarily determined the wild progenitor of "Baizhi" and clarified the phylogenetic relationships among A. dahurica var. dahurica, A. dahurica var. formosana and A. porphyrocaulis, which will provide scientific guidance for wild resources protections and improvement of "Baizhi".

Prediction of breast cancer and axillary positive-node response to neoadjuvant chemotherapy based on multi-parametric magnetic resonance imaging radiomics models.

PURPOSE: Accurate identification of primary breast cancer and axillary positive-node response to neoadjuvant chemotherapy (NAC) is important for determining appropriate surgery strategies. We aimed to develop combining models based on breast multi-parametric magnetic resonance imaging and clinicopathologic characteristics for predicting therapeutic response of primary tumor and axillary positive-node prior to treatment. MATERIALS AND METHODS: A total of 268 breast cancer patients who completed NAC and underwent surgery were enrolled. Radiomics features and clinicopathologic characteristics were analyzed through the analysis of variance and the least absolute shrinkage and selection operator algorithm. Finally, 24 and 28 optimal features were selected to construct machine learning models based on 6 algorithms for predicting each clinical outcome, respectively. The diagnostic performances of models were evaluated in the testing set by the area under the curve (AUC), sensitivity, specificity, and accuracy. RESULTS: Of the 268 patients, 94 (35.1 %) achieved breast cancer pathological complete response (bpCR) and of the 240 patients with clinical positive-node, 120 (50.0 %) achieved axillary lymph node pathological complete response (apCR). The multi-layer perception (MLP) algorithm yielded the best diagnostic performances in predicting apCR with an AUC of 0.825 (95 % CI, 0.764-0.886) and an accuracy of 77.1 %. And MLP also outperformed other models in predicting bpCR with an AUC of 0.852 (95 % CI, 0.798-0.906) and an accuracy of 81.3 %. CONCLUSIONS: Our study established non-invasive combining models to predict the therapeutic response of primary breast cancer and axillary positive-node prior to NAC, which may help to modify preoperative treatment and determine post-NAC surgery strategy.

Predicting One-year Recurrence of HCC based on Gadoxetic Acid-enhanced MRI by Machine Learning Models.

OBJECTIVE: Accurate prediction of recurrence risk after resction in patients with Hepatocellular Carcinoma (HCC) may help to individualize therapy strategies. This study aimed to develop machine learning models based on preoperative clinical factors and multiparameter Magnetic Resonance Imaging (MRI) characteristics to predict the 1-year recurrence after HCC resection. METHODS: Eighty-two patients with single HCC who underwent surgery were retrospectively analyzed. All patients underwent preoperative gadoxetic acidenhanced MRI examination. Preoperative clinical factors and MRI characteristics were collected for feature selection. Least Absolute Shrinkage and Selection Operator (LASSO) was applied to select the optimal features for predicting postoperative 1-year recurrence of HCC. Four machine learning algorithms, Multilayer Perception (MLP), random forest, support vector machine, and k-nearest neighbor, were used to construct the predictive models based on the selected features. A Receiver Operating Characteristic (ROC) curve was used to assess the performance of each model. RESULTS: Among the enrolled patients, 32 patients experienced recurrences within one year, while 50 did not. Tumor size, peritumoral hypointensity, decreasing ratio of liver parenchyma T1 value (ΔT1), and α-fetoprotein (AFP) levels were selected by using LASSO to develop the machine learning models. The area under the curve (AUC) of each model exceeded 0.72. Among the models, the MLP model showed the best performance with an AUC, accuracy, sensitivity, and specificity of 0.813, 0.742, 0.570, and 0.853, respectively. CONCLUSION: Machine learning models can accurately predict postoperative 1-year recurrence in patients with HCC, which may help to provide individualized treatment.

Temporal attention networks for biomedical hypothesis generation.

OBJECTIVES: Hypothesis Generation (HG) is a task that aims to uncover hidden associations between disjoint scientific terms, which influences innovations in prevention, treatment, and overall public health. Several recent studies strive to use Recurrent Neural Network (RNN) to learn evolutional embeddings for HG. However, the complex spatiotemporal dependencies of term-pair relations will be difficult to depict due to the inherent recurrent structure. This paper aims to accurately model the temporal evolution of term-pair relations using only attention mechanisms, for capturing crucial information on inferring the future connectivities. METHODS: This paper proposes a Temporal Attention Networks (TAN) to produce powerful spatiotemporal embeddings for Biomedical Hypothesis Generation. Specifically, we formulate HG problem as a future connectivity prediction task in a temporal attributed graph. Our TAN develops a Temporal Spatial Attention Module (TSAM) to establish temporal dependencies of node-pair (term-pair) embeddings between any two time-steps for smoothing spatiotemporal node-pair embeddings. Meanwhile, a Temporal Difference Attention Module (TDAM) is proposed to sharpen temporal differences of spatiotemporal embeddings for highlighting the historical changes of node-pair relations. As such, TAN can adaptively calibrate spatiotemporal embeddings by considering both continuity and difference of node-pair embeddings. RESULTS: Three real-world biomedical term relationship datasets are constructed from PubMed papers. TAN significantly outperforms the best baseline with 12.03%, 4.59 and 2.34% Micro-F1 Score improvement in Immunotherapy, Virology and Neurology, respectively. Extensive experiments demonstrate that TAN can model complex spatiotemporal dependencies of term-pairs for explicitly capturing the temporal evolution of relation, significantly outperforming existing state-of-the-art methods. CONCLUSION: We proposed a novel TAN to learn spatiotemporal embeddings based on pure attention mechanisms for HG. TAN learns the evolution of relationships by modeling both the continuity and difference of temporal term-pair embeddings. The important spatiotemporal dependencies of term-pair relations are extracted based solely on attention mechanism for generating hypotheses.

Multi-modality radiomics model predicts axillary lymph node metastasis of breast cancer using MRI and mammography.

OBJECTIVES: We aimed to develop a multi-modality model to predict axillary lymph node (ALN) metastasis by combining clinical predictors with radiomic features from magnetic resonance imaging (MRI) and mammography (MMG) in breast cancer. This model might potentially eliminate unnecessary axillary surgery in cases without ALN metastasis, thereby minimizing surgery-related complications. METHODS: We retrospectively enrolled 485 breast cancer patients from two hospitals and extracted radiomics features from tumor and lymph node regions on MRI and MMG images. After feature selection, three random forest models were built using the retained features, respectively. Significant clinical factors were integrated with these radiomics models to construct a multi-modality model. The multi-modality model was compared to radiologists' diagnoses on axillary ultrasound and MRI. It was also used to assist radiologists in making a secondary diagnosis on MRI. RESULTS: The multi-modality model showed superior performance with AUCs of 0.964 in the training cohort, 0.916 in the internal validation cohort, and 0.892 in the external validation cohort. It surpassed single-modality models and radiologists' ALN diagnosis on MRI and axillary ultrasound in all validation cohorts. Additionally, the multi-modality model improved radiologists' MRI-based ALN diagnostic ability, increasing the average accuracy from 70.70 to 78.16% for radiologist A and from 75.42 to 81.38% for radiologist B. CONCLUSION: The multi-modality model can predict ALN metastasis of breast cancer accurately. Moreover, the artificial intelligence (AI) model also assisted the radiologists to improve their diagnostic ability on MRI. CLINICAL RELEVANCE STATEMENT: The multi-modality model based on both MRI and mammography images allows preoperative prediction of axillary lymph node metastasis in breast cancer patients. With the assistance of the model, the diagnostic efficacy of radiologists can be further improved. KEY POINTS: • We developed a novel multi-modality model that combines MRI and mammography radiomics with clinical factors to accurately predict axillary lymph node (ALN) metastasis, which has not been previously reported. • Our multi-modality model outperformed both the radiologists' ALN diagnosis based on MRI and axillary ultrasound, as well as single-modality radiomics models based on MRI or mammography. • The multi-modality model can serve as a potential decision support tool to improve the radiologists' ALN diagnosis on MRI.

Radiomics nomogram for predicting disease-free survival after partial resection or radical cystectomy in patients with bladder cancer.

OBJECTIVES: To create a MRI-derived radiomics nomogram that combined clinicopathological factors and radiomics signature (Rad-score) for predicting disease-free survival (DFS) in patients with bladder cancer (BCa) following partial resection (PR) or radical cystectomy (RC), including lymphadenectomy (LAE). METHODS: Finally, 80 patients with BCa after PR or RC with LAE were enrolled. Patients were randomly split into training (n = 56) and internal validation (n = 24) cohorts. Radiomic features were extracted from T2-weighted, dynamic contrast-enhanced, diffusion-weighted imaging, and apparent diffusion coefficient sequence. The least absolute shrinkage and selection operator (LASSO) Cox regression algorithm was applied to choose the valuable features and construct the Rad-score. The DFS prediction model was built using the Cox proportional hazards model. The relationship between the Rad-score and DFS was assessed using Kaplan-Meier analysis. A radiomics nomogram that combined the Rad-score and clinicopathological factors was created for individualized DFS estimation. RESULTS: In both the training and validation cohorts, the Rad-score was positively correlated with DFS (P < .001). In the validation cohort, the radiomics nomogram combining the Rad-score, tumour pathologic stage (pT stage), and lymphovascular invasion (LVI) achieved better performance in DFS prediction (C-index, 0.807; 95% CI, 0.713-0.901) than either the clinicopathological (C-index, 0.654; 95% CI, 0.467-0.841) or Rad-score-only model (C-index, 0.770; 95% CI, 0.702-0.837). CONCLUSION: The Rad-score was an independent predictor of DFS for patients with BCa after PR or RC with LAE, and the radiomics nomogram that combined the Rad-score, pT stage, and LVI achieved better performance in individual DFS prediction. ADVANCES IN KNOWLEDGE: This study provided a non-invasive and simple method for personalized and accurate prediction of DFS in BCa patients after PR or RC.

Amide Proton Transfer-Weighted MRI and Diffusion-Weighted Imaging in Bladder Cancer: A Complementary Tool to the VI-RADS.

RATIONALE AND OBJECTIVES: To investigate the feasibility of amide proton transfer-weighted (APTw) and diffusion-weighted Magnetic Resonance Imaging (MRI) as a means by which to add value to the Vesical Imaging Reporting and Data System (VI-RADS) for discriminating muscle invasive bladder cancer (MIBC) from nonmuscle invasive bladder cancer (NMIBC). MATERIALS AND METHODS: This prospective study enrolled participants with pathologically confirmed bladder cancer (BCa) who underwent preoperative multiparametric MRI, including APTw and diffusion-weighted MRI, from July 2020 to January 2023. The exclusion criteria were lesions smaller than 10 mm, missing smooth muscle layer in the operation specimen, neoadjuvant therapy before MRI, inadequate image quality, and malignancy other than urothelial neoplasm. Two radiologists independently assigned the VI-RADS score for each participant. Quantitative parameters derived from APTw and diffusion-weighted MRI were obtained by another two radiologists. Receiver operating characteristic (ROC) curve analysis with the area under the ROC curve (AUC) was performed to evaluate the diagnostic performances of quantitative parameters for discriminating BCa detrusor muscle invasion status. RESULTS: A total of 106 participants were enrolled (mean age, 64 ± 12 years [SD]; 90 men): 32 with MIBC and 74 with NMIBC. Lower apparent diffusion coefficient (ADC) values (0.88 × 10-3 mm2/s ± 0.12 vs. 1.08 × 10-3 mm2/s ± 0.25; P < 0.001) and higher APTw values (6.89% [interquartile range {IQR}, 5.05%-12.17%] vs. 3.61% [IQR, 2.23%-6.83%]; P < 0.001) were observed in the MIBC group. Compared to VI-RADS alone, both APTw (P = 0.003) and ADC (P = 0.020) values could improve the diagnostic performance of VI-RADS in differentiating MIBC from NMIBC. The combination of the three yielded the highest diagnostic performance (AUC, 0.93; 95% CI:0.87,0.97) for evaluating muscle invasion status. The addition of the APTw values to the combination of VI-RADS and ADC values notably improved the diagnostic performance for differentiating NMIBC from MIBC (VI-RADS+ADC vs. VI-RADS+APTw+ADC, P = 0.046). CONCLUSION: MRI parameters derived from APTw and diffusion-weighted MRI can be used to accurately assess muscle invasion status in BCa and provide additional value to VI-RADS.

High power density output and durability of microbial fuel cells enabled by dispersed cobalt nanoparticles on nitrogen-doped carbon as the cathode electrocatalyst.

To endow microbial fuel cells (MFCs) with low cost, long-term stability and high-power output, a novel cobalt-based cathode electrocatalyst (Nano-Co@NC) is synthesized from a polygonal metal-organic framework ZIF-67. After calcining the resultant ZIF-67, the as-synthesized Nano-Co@NC is characteristic of cobalt nanoparticles (Nano-Co) embedded in nitrogen-doped carbon (NC) that inherits the morphology of ZIF-67 with a large surface area. The Nano-Co particles that are highly dispersed and firmly fixed on NC not only ensure electrocatalytic activity of Nano-Co@NC toward the oxygen reduction reaction on the cathode, but also inhibit the growth of non-electrogenic bacteria on the anode. Consequently, the MFC using Nano-Co@NC as the cathode electrocatalyst demonstrates excellent performance, delivering a comparable initial power density and exhibiting far better durability than that using Pt/C (20 wt%) as the cathode electrocatalyst. The low cost and the excellent performance of Nano-Co@NC make it promising for MFCs to be used in practice.

A facile synthesis of FeS/Fe3C nanoparticles highly dispersed on in situ grown N-doped CNTs as cathode electrocatalysts for microbial fuel cells.

A novel composite of iron sulfide, iron carbide and nitrogen carbides (Nano-FeS/Fe3C@NCNTs) as a cathode electrocatalyst for microbial fuel cells (MFCs) is synthesized by a one-pot solid state reaction, which yields a unique configuration of FeS/Fe3C nanoparticles highly dispersed on in situ grown nitrogen-doped carbon nanotubes (NCNTs). The highly dispersed FeS/Fe3C nanoparticles possess large active sites, while the NCNTs provide an electronically conductive network. Consequently, the resultant Nano-FeS/Fe3C@NCNTs exhibit excellent electrocatalytic activity towards the oxygen reduction reaction (ORR), with a half-wave potential close to that of Pt/C (about 0.88 V vs. RHE), and enable MFCs to deliver a power density of 1.28 W m-2 after two weeks' operation, which is higher than that of MFCs with Pt/C as the cathode electrocatalyst (1.02 W m-2). Theoretical calculations and experimental data demonstrate that there is a synergistic effect between Fe3C and FeS in Nano-FeS/Fe3C@NCNTs. Fe3C presents a strong attraction and electron-donating tendency to oxygen molecules, serving as the main active component, while FeS reduces charge transfer resistance by transferring electrons to Fe3C, synergistically improving the kinetics of the ORR and power density of MFCs.

The clinical presentation and treatment of an invasive conjunctival squamous spindle cell carcinoma.

Ocular surface squamous neoplasia represents neoplastic epithelial abnormalities of conjunctiva and cornea, ranging from squamous dysplasia to invasive squamous cell carcinoma and is both sight- and life-threatening. Squamous spindle cell carcinoma (SSCC) of conjunctiva is a rare variant with distinct behavior which is thought to be more locally aggressive. We describe an 83-year-old woman with a progressively enlarging huge SSCC in her right eye over the past 2 years. The tumor bulged out with local invasion into intraocular and orbital cavities. Wide excision of the tumor with frozen section control was performed. After surgery, topical 0.03% mitomycin C was given as adjuvant therapy. At 40-month follow-up, the lesion site showed no evidence of local recurrence. This case provides a valuable and complete experience of the clinical presentation for the progression and treatment of this rare disease.

Multiparametric MRI Evaluation of VI-RADS for Bladder Tumors Located at the Ureteral Orifice.

Background The Vesical Imaging Reporting and Data System (VI-RADS) based on multiparametric MRI scans standardizes preoperative bladder cancer staging. However, limitations have been reported for VI-RADS, particularly for ureteral orifice tumors. Purpose To investigate the diagnostic performance and interobserver agreement of VI-RADS in evaluating muscle invasion for bladder tumors located at the ureteral orifice. Materials and Methods In this retrospective study, patients with histopathologically confirmed bladder cancer occurring at the ureteral orifice from January 2012 to November 2021 were analyzed. Two blinded radiologists independently scored multiparametric MRI scans according to VI-RADS. Interobserver agreement of the VI-RADS scores was evaluated with weighted κ analysis. Receiver operating characteristic curve analysis was used to evaluate the diagnostic performance of the VI-RADS scores in the prediction of muscle invasion. Results A total of 78 patients (mean age, 67 years ± 7 [SD]; age range, 46-90 years; 67 men) were included in the final analysis: 25 with non-muscle-invasive bladder cancer and 53 with muscle-invasive bladder cancer (MIBCa). At consensus reading, one (1%) case was scored as VI-RADS 1, 27 cases (35%) were scored as VI-RADS 2, six (8%) were scored as VI-RADS 3, 10 (13%) were scored as VI-RADS 4, and 34 (44%) were scored as VI-RADS 5. On comparison of the VI-RADS score with histopathologic findings, it was confirmed that the presence of muscle invasion was 0% (zero of one) for VI-RADS 1, 15% (four of 27) for VI-RADS 2, 83% (five of six) for VI-RADS 3, 100% (10 of 10) for VI-RADS 4, and 100% (34 of 34) for VI-RADS 5. The area under the receiver operating characteristic curve of VI-RADS in the detection of MIBCa was 0.96 (95% CI: 0.92, 1.00). Conclusion The Vesical Imaging Reporting and Data System could be used to accurately predict muscle invasion for bladder tumors occurring at the ureteral orifice. © RSNA, 2022 Online supplemental material is available for this article.

Amide Proton Transfer-weighted MRI in Predicting Histologic Grade of Bladder Cancer.

Background Bladder cancer is classified into high and low grades with different clinical treatments and prognoses. Thus, accurate preoperative evaluation of the histologic grade through imaging techniques is essential. Purpose To investigate the potential of amide proton transfer-weighted (APTw) MRI in evaluating the grade of bladder cancer and to evaluate whether APTw MRI can add value to diffusion-weighted imaging (DWI) at MRI. Materials and Methods In this single-center prospective study, participants with pathologic analysis-confirmed bladder cancer with no previous treatment, lesions larger than 10 mm, and adequate MRI quality were enrolled from July 2020 to September 2021 in a university teaching hospital. All participants underwent preoperative multiparametric MRI, including APTw MRI and DWI. The mean APTw and apparent diffusion coefficient (ADC) values of the primary tumor were measured independently by two radiologists. Receiver operating characteristic curves were generated to evaluate the diagnostic performance of these quantitative parameters. Results In total, 83 participants (mean age, 64 years ± 13 [SD]; 72 men) were evaluated: 51 with high-grade and 32 with low-grade bladder cancer. High-grade bladder cancer showed higher APTw values (6% [IQR, 4%-12%] vs 2% [IQR, 1%-3%]; P < .001) and lower ADC values (0.92 × 10-3 mm2/sec ± 0.17 vs 1.21 × 10-3 mm2/sec ± 0.25; P < .001) than low-grade bladder cancer. The area under the receiver operating characteristic curve (AUC) of APTw and ADC for differentiating low- and high-grade bladder cancer was similar (0.84 for both; P = .94). Moreover, the combination of the two techniques improved the diagnostic performance (AUC, 0.93; all P = .01). Conclusion The combination of amide proton transfer-weighted and diffusion-weighted MRI has the potential to improve the histologic characterization of bladder cancer by differentiating low- from high-grade cancers. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Milot in this issue. An earlier incorrect version appeared online. This article was corrected on July 7, 2022.

Intrinsic epigenetic control of angiogenesis in induced pluripotent stem cell-derived endothelium regulates vascular regeneration.

Human-induced pluripotent stem cell-derived endothelial cells (iECs) provide opportunities to study vascular development and regeneration, develop cardiovascular therapeutics, and engineer model systems for drug screening. The differentiation and characterization of iECs are well established; however, the mechanisms governing their angiogenic phenotype remain unknown. Here, we aimed to determine the angiogenic phenotype of iECs and the regulatory mechanism controlling their regenerative capacity. In a comparative study with HUVECs, we show that iECs increased expression of vascular endothelial growth factor receptor 2 (VEGFR2) mediates their highly angiogenic phenotype via regulation of glycolysis enzymes, filopodia formation, VEGF mediated migration, and robust sprouting. We find that the elevated expression of VEGFR2 is epigenetically regulated via intrinsic acetylation of histone 3 at lysine 27 by histone acetyltransferase P300. Utilizing a zebrafish xenograft model, we demonstrate that the ability of iECs to promote the regeneration of the amputated fin can be modulated by P300 activity. These findings demonstrate how the innate epigenetic status of iECs regulates their phenotype with implications for their therapeutic potential.

Hypoxia-induced blood-brain barrier dysfunction is prevented by pericyte-conditioned media via attenuated actomyosin contractility and claudin-5 stabilization.

The blood-brain barrier (BBB) regulates molecular and cellular entry from the cerebrovasculature into the surrounding brain parenchyma. Many diseases of the brain are associated with dysfunction of the BBB, where hypoxia is a common stressor. However, the contribution of hypoxia to BBB dysfunction is challenging to study due to the complexity of the brain microenvironment. In this study, we used a BBB model with brain microvascular endothelial cells and pericytes differentiated from iPSCs to investigate the effect of hypoxia on barrier function. We found that hypoxia-induced barrier dysfunction is dependent upon increased actomyosin contractility and is associated with increased fibronectin fibrillogenesis. We propose a role for actomyosin contractility in mediating hypoxia-induced barrier dysfunction through modulation of junctional claudin-5. Our findings suggest pericytes may protect brain microvascular endothelial cells from hypoxic stresses and that pericyte-derived factors could be candidates for treatment of pathological barrier-forming tissues.

Deep Segmentation Feature-Based Radiomics Improves Recurrence Prediction of Hepatocellular Carcinoma.

Objective and Impact Statement. This study developed and validated a deep semantic segmentation feature-based radiomics (DSFR) model based on preoperative contrast-enhanced computed tomography (CECT) combined with clinical information to predict early recurrence (ER) of single hepatocellular carcinoma (HCC) after curative resection. ER prediction is of great significance to the therapeutic decision-making and surveillance strategy of HCC. Introduction. ER prediction is important for HCC. However, it cannot currently be adequately determined. Methods. Totally, 208 patients with single HCC after curative resection were retrospectively recruited into a model-development cohort (n=180) and an independent validation cohort (n=28). DSFR models based on different CT phases were developed. The optimal DSFR model was incorporated with clinical information to establish a DSFR-C model. An integrated nomogram based on the Cox regression was established. The DSFR signature was used to stratify high- and low-risk ER groups. Results. A portal phase-based DSFR model was selected as the optimal model (area under receiver operating characteristic curve (AUC): development cohort, 0.740; validation cohort, 0.717). The DSFR-C model achieved AUCs of 0.782 and 0.744 in the development and validation cohorts, respectively. In the development and validation cohorts, the integrated nomogram achieved C-index of 0.748 and 0.741 and time-dependent AUCs of 0.823 and 0.822, respectively, for recurrence-free survival (RFS) prediction. The RFS difference between the risk groups was statistically significant (P<0.0001 and P=0.045 in the development and validation cohorts, respectively). Conclusion. CECT-based DSFR can predict ER in single HCC after curative resection, and its combination with clinical information further improved the performance for ER prediction.

A narrative review on machine learning in diagnosis and prognosis prediction for tongue squamous cell carcinoma.

Background: Tongue squamous cell carcinoma (TSCC) is the most common subtype of oral cavity squamous cell carcinoma (OCSCC), and it also has the worst prognosis. It is crucial to find an effective way to solve the challenges in diagnosis and prognosis prediction for TSCC. Machine learning (ML) has been widely used in medical research and has shown good performance. It can be used for feature extraction, feature selection, model construction, etc. Radiomics and deep learning (DL), the new components of ML, have also been utilized to explore the relationship between image features and diseases. The current study aimed to highlight the importance of ML as a potential method for addressing the challenges in diagnosis and prognosis prediction of TSCC by reviewing studies on ML in TSCC. Methods: The studies on ML in TSCC in PubMed, Scopus, Web of Science, and China National Knowledge Infrastructure published between the dates of inception of these databases and April 30, 2022, were reviewed. Key Content and Findings: ML (including radiomics and DL) which was used in diagnosis and prognosis prediction for TSCC, has shown promising performance. Conclusions: Despite its limitations, ML is still a potential approach that can help to deal with the challenges in diagnosis and prognosis prediction for TSCC. Nevertheless, more efforts are needed to enhance the usefulness of ML in this field.

Prediction of Early Treatment Response to Initial Conventional Transarterial Chemoembolization Therapy for Hepatocellular Carcinoma by Machine-Learning Model Based on Computed Tomography.

PURPOSE: The treatment response to initial conventional transarterial chemoembolization (cTACE) is essential for the prognosis of patients with hepatocellular carcinoma (HCC). This study explored and verified the feasibility of machine-learning models based on clinical data and contrast-enhanced computed tomography (CT) image findings to predict early responses of HCC patients after initial cTACE treatment. PATIENTS AND METHODS: Overall, 110 consecutive unresectable HCC patients who were treated with cTACE for the first time were retrospectively enrolled. Clinical data and imaging features based on contrast-enhanced CT were collected for the selection of characteristics. Treatment responses were evaluated based on the modified Response Evaluation Criteria in Solid Tumors (mRECIST) by postoperative CT examination within 2 months after the procedure. Python (version 3.70) was used to develop machine learning models. Least absolute shrinkage and selection operator (LASSO) algorithm was applied to select features with the impact on predicting treatment response after the first TACE procedure. Six machine learning algorithms were used to build predictive models, including XGBoost, decision tree, support vector machine, random forest, k-nearest neighbor, and fully convolutional networks, and their performances were compared using receiver operator characteristic (ROC) curves to determine the best performing model. RESULTS: Following TACE, 31 patients (28.2%) were described as responsive to TACE, while 72 patients (71.8%) were nonresponsive to TACE. Portal vein tumor thrombosis type, albumin level, and distribution of tumors within the liver were selected for predictive model building. Among the models, the RF model showed the best performance, with area under the curve (AUC), accuracy, sensitivity, and specificity of 0.802, 0.784, 0.904, and 0.480, respectively. CONCLUSION: Machine learning models can provide an accurate prediction of the early response of initial TACE treatment for HCC, which can help in individualizing clinical decision-making and modification of further treatment strategies for patients with unresectable HCC.

Multi-Parametric MRI-Based Radiomics Models for Predicting Molecular Subtype and Androgen Receptor Expression in Breast Cancer.

OBJECTIVE: To investigate whether radiomics features extracted from multi-parametric MRI combining machine learning approach can predict molecular subtype and androgen receptor (AR) expression of breast cancer in a non-invasive way. MATERIALS AND METHODS: Patients diagnosed with clinical T2-4 stage breast cancer from March 2016 to July 2020 were retrospectively enrolled. The molecular subtypes and AR expression in pre-treatment biopsy specimens were assessed. A total of 4,198 radiomics features were extracted from the pre-biopsy multi-parametric MRI (including dynamic contrast-enhancement T1-weighted images, fat-suppressed T2-weighted images, and apparent diffusion coefficient map) of each patient. We applied several feature selection strategies including the least absolute shrinkage and selection operator (LASSO), and recursive feature elimination (RFE), the maximum relevance minimum redundancy (mRMR), Boruta and Pearson correlation analysis, to select the most optimal features. We then built 120 diagnostic models using distinct classification algorithms and feature sets divided by MRI sequences and selection strategies to predict molecular subtype and AR expression of breast cancer in the testing dataset of leave-one-out cross-validation (LOOCV). The performances of binary classification models were assessed via the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). And the performances of multiclass classification models were assessed via AUC, overall accuracy, precision, recall rate, and F1-score. RESULTS: A total of 162 patients (mean age, 46.91 ± 10.08 years) were enrolled in this study; 30 were low-AR expression and 132 were high-AR expression. HR+/HER2- cancers were diagnosed in 56 cases (34.6%), HER2+ cancers in 81 cases (50.0%), and TNBC in 25 patients (15.4%). There was no significant difference in clinicopathologic characteristics between low-AR and high-AR groups (P > 0.05), except the menopausal status, ER, PR, HER2, and Ki-67 index (P = 0.043, <0.001, <0.001, 0.015, and 0.006, respectively). No significant difference in clinicopathologic characteristics was observed among three molecular subtypes except the AR status and Ki-67 (P = <0.001 and 0.012, respectively). The Multilayer Perceptron (MLP) showed the best performance in discriminating AR expression, with an AUC of 0.907 and an accuracy of 85.8% in the testing dataset. The highest performances were obtained for discriminating TNBC vs. non-TNBC (AUC: 0.965, accuracy: 92.6%), HER2+ vs. HER2- (AUC: 0.840, accuracy: 79.0%), and HR+/HER2- vs. others (AUC: 0.860, accuracy: 82.1%) using MLP as well. The micro-AUC of MLP multiclass classification model was 0.896, and the overall accuracy was 0.735. CONCLUSIONS: Multi-parametric MRI-based radiomics combining with machine learning approaches provide a promising method to predict the molecular subtype and AR expression of breast cancer non-invasively.