Lesion-aware cross-phase attention network for renal tumor subtype classification on multi-phase CT scans.

Multi-phase computed tomography (CT) has been widely used for the preoperative diagnosis of kidney cancer due to its non-invasive nature and ability to characterize renal lesions. However, since enhancement patterns of renal lesions across CT phases are different even for the same lesion type, the visual assessment by radiologists suffers from inter-observer variability in clinical practice. Although deep learning-based approaches have been recently explored for differential diagnosis of kidney cancer, they do not explicitly model the relationships between CT phases in the network design, limiting the diagnostic performance. In this paper, we propose a novel lesion-aware cross-phase attention network (LACPANet) that can effectively capture temporal dependencies of renal lesions across CT phases to accurately classify the lesions into five major pathological subtypes from time-series multi-phase CT images. We introduce a 3D inter-phase lesion-aware attention mechanism to learn effective 3D lesion features that are used to estimate attention weights describing the inter-phase relations of the enhancement patterns. We also present a multi-scale attention scheme to capture and aggregate temporal patterns of lesion features at different spatial scales for further improvement. Extensive experiments on multi-phase CT scans of kidney cancer patients from the collected dataset demonstrate that our LACPANet outperforms state-of-the-art approaches in diagnostic accuracy.

Cross-domain Denoising for Low-dose Multi-frame Spiral Computed Tomography.

Computed tomography (CT) has been used worldwide as a non-invasive test to assist in diagnosis. However, the ionizing nature of X-ray exposure raises concerns about potential health risks such as cancer. The desire for lower radiation doses has driven researchers to improve reconstruction quality. Although previous studies on low-dose computed tomography (LDCT) denoising have demonstrated the effectiveness of learning-based methods, most were developed on the simulated data. However, the real-world scenario differs significantly from the simulation domain, especially when using the multi-slice spiral scanner geometry. This paper proposes a two-stage method for the commercially available multi-slice spiral CT scanners that better exploits the complete reconstruction pipeline for LDCT denoising across different domains. Our approach makes good use of the high redundancy of multi-slice projections and the volumetric reconstructions while leveraging the over-smoothing issue in conventional cascaded frameworks caused by aggressive denoising. The dedicated design also provides a more explicit interpretation of the data flow. Extensive experiments on various datasets showed that the proposed method could remove up to 70% of noise without compromised spatial resolution, while subjective evaluations by two experienced radiologists further supported its superior performance against state-of-the-art methods in clinical practice. Code is available at https://github.com/YCL92/TMD-LDCT.

Role of UPF1-LIN28A interaction during early differentiation of pluripotent stem cells.

UPF1 and LIN28A are RNA-binding proteins involved in post-transcriptional regulation and stem cell differentiation. Most studies on UPF1 and LIN28A have focused on the molecular mechanisms of differentiated cells and stem cell differentiation, respectively. We reveal that LIN28A directly interacts with UPF1 before UPF1-UPF2 complexing, thereby reducing UPF1 phosphorylation and inhibiting nonsense-mediated mRNA decay (NMD). We identify the interacting domains of UPF1 and LIN28A; moreover, we develop a peptide that impairs UPF1-LIN28A interaction and augments NMD efficiency. Transcriptome analysis of human pluripotent stem cells (hPSCs) confirms that the levels of NMD targets are significantly regulated by both UPF1 and LIN28A. Inhibiting the UPF1-LIN28A interaction using a CPP-conjugated peptide promotes spontaneous differentiation by repressing the pluripotency of hPSCs during proliferation. Furthermore, the UPF1-LIN28A interaction specifically regulates transcripts involved in ectodermal differentiation. Our study reveals that transcriptome regulation via the UPF1-LIN28A interaction in hPSCs determines cell fate.

Pretransplant BKV-IgG serostatus and BKV-specific ELISPOT assays to predict BKV infection after kidney transplantation.

Introduction: Polyomavirus (BKV) infection can lead to major complications and damage to the graft in kidney transplant recipients (KTRs). We investigated whether pretransplant BK serostatus and BK-specific cell-mediated immunity (CMI) predicts post-transplant BK infection. Methods: A total of 93 donor-recipient pairs who underwent kidney transplantation (KT) and 44 healthy controls were examined. Assessment of donor and recipient BKV serostatus and BKV-CMI in recipients was performed prior to transplantation using BKV-IgG ELISA and BKV-specific IFN-g ELISPOT assays against five BK viral antigens (LT, St, VP1, VP2, and VP3). BK viremia was diagnosed when blood BKV-DNA of 104 copies/mL or more was detected during follow-up periods. Results: Anti-BKV IgG antibody was detected in 74 (79.6%) of 93 KTRs and in 68 (73.1%) of 93 KT donors. A greater percentage of KTRs who received allograft from donors with high levels of anti-BKV IgG had posttransplant BK viremia (+) than KTRs from donors with low anti-BKV IgG (25.5% [12/47] vs. 4.3% [2/46], respectively; P = 0.007). Pretransplant total BKV-ELISPOT results were lower in BK viremia (+) patients than in patients without viremia (-) 20.5 [range 9.9-63.6] vs. 72.0 [43.2 - 110.8]; P = 0. 027). The sensitivity and specificity of the total BKV-ELISPOT assay (cut-off ≤ 53 spots/3×105 cells) for prediction of posttransplant BK viremia were 71.4 (95% CI: 41.9-91.6) and 54.4 (42.8-65.7), respectively. The combination of high donor BKV-IgG, low recipient BKV-IgG, and low total BKV-ELISPOT results improved specificity to 91.1%. Discussion: Our study highlights the importance of pretransplant BKV-IgG serostatus and BKV-specific CMI in predicting posttransplant BKV infection in KTRs. The combination of high donor BKV-IgG, low recipient BKV-IgG, and low total BKV-ELISPOT results predicted BK viremia after KT. Pretransplant identification of patients at highrisk for BK viremia could enable timely interventions and improve clinical outcomes of KTRs.

Deep Learning-Based Evaluation of Ultrasound Images for Benign Skin Tumors.

In this study, a combined convolutional neural network for the diagnosis of three benign skin tumors was designed, and its effectiveness was verified through quantitative and statistical analysis. To this end, 698 sonographic images were taken and diagnosed at the Department of Dermatology at Severance Hospital in Seoul, Korea, between 10 November 2017 and 17 January 2020. Through an empirical process, a convolutional neural network combining two structures, which consist of a residual structure and an attention-gated structure, was designed. Five-fold cross-validation was applied, and the train set for each fold was augmented by the Fast AutoAugment technique. As a result of training, for three benign skin tumors, an average accuracy of 95.87%, an average sensitivity of 90.10%, and an average specificity of 96.23% were derived. Also, through statistical analysis using a class activation map and physicians' findings, it was found that the judgment criteria of physicians and the trained combined convolutional neural network were similar. This study suggests that the model designed and trained in this study can be a diagnostic aid to assist physicians and enable more efficient and accurate diagnoses.

Risk-aware survival time prediction from whole slide pathological images.

Deep-learning-based survival prediction can assist doctors by providing additional information for diagnosis by estimating the risk or time of death. The former focuses on ranking deaths among patients based on the Cox model, whereas the latter directly predicts the survival time of each patient. However, it is observed that survival time prediction for the patients, particularly with close observation times, possibly has incorrect orders, leading to low prediction accuracy. Therefore, in this paper, we present a whole slide image (WSI)-based survival time prediction method that takes advantage of both the risk as well as time prediction. Specifically, we propose to combine these two approaches by extracting the risk prediction features and using them as guides for the survival time prediction. Considering the high resolution of WSIs, we extract tumor patches from WSIs using a pre-trained tumor classifier and apply the graph convolutional network to aggregate information across these patches effectively. Extensive experiments demonstrate that the proposed method significantly improves the time prediction accuracy when compared with direct prediction of the survival times without guidance and outperforms existing methods.

A Unified Multi-Phase CT Synthesis and Classification Framework for Kidney Cancer Diagnosis With Incomplete Data.

Multi-phase computed tomography (CT) is widely adopted for the diagnosis of kidney cancer due to the complementary information among phases. However, the complete set of multi-phase CT is often not available in practical clinical applications. In recent years, there have been some studies to generate the missing modality image from the available data. Nevertheless, the generated images are not guaranteed to be effective for the diagnosis task. In this paper, we propose a unified framework for kidney cancer diagnosis with incomplete multi-phase CT, which simultaneously recovers missing CT images and classifies cancer subtypes using the completed set of images. The advantage of our framework is that it encourages a synthesis model to explicitly learn to generate missing CT phases that are helpful for classifying cancer subtypes. We further incorporate lesion segmentation network into our framework to exploit lesion-level features for effective cancer classification in the whole CT volumes. The proposed framework is based on fully 3D convolutional neural networks to jointly optimize both synthesis and classification of 3D CT volumes. Extensive experiments on both in-house and external datasets demonstrate the effectiveness of our framework for the diagnosis with incomplete data compared with state-of-the-art baselines. In particular, cancer subtype classification using the completed CT data by our method achieves higher performance than the classification using the given incomplete data.

Deep learning for end-to-end kidney cancer diagnosis on multi-phase abdominal computed tomography.

In 2020, it is estimated that 73,750 kidney cancer cases were diagnosed, and 14,830 people died from cancer in the United States. Preoperative multi-phase abdominal computed tomography (CT) is often used for detecting lesions and classifying histologic subtypes of renal tumor to avoid unnecessary biopsy or surgery. However, there exists inter-observer variability due to subtle differences in the imaging features of tumor subtypes, which makes decisions on treatment challenging. While deep learning has been recently applied to the automated diagnosis of renal tumor, classification of a wide range of subtype classes has not been sufficiently studied yet. In this paper, we propose an end-to-end deep learning model for the differential diagnosis of five major histologic subtypes of renal tumors including both benign and malignant tumors on multi-phase CT. Our model is a unified framework to simultaneously identify lesions and classify subtypes for the diagnosis without manual intervention. We trained and tested the model using CT data from 308 patients who underwent nephrectomy for renal tumors. The model achieved an area under the curve (AUC) of 0.889, and outperformed radiologists for most subtypes. We further validated the model on an independent dataset of 184 patients from The Cancer Imaging Archive (TCIA). The AUC for this dataset was 0.855, and the model performed comparably to the radiologists. These results indicate that our model can achieve similar or better diagnostic performance than radiologists in differentiating a wide range of renal tumors on multi-phase CT.

Co-culture of functionally enriched cancer stem-like cells and cancer-associated fibroblasts for single-cell whole transcriptome analysis.

Considerable evidence suggests that breast cancer development and metastasis are driven by cancer stem-like cells (CSCs). Due to their unique role in tumor initiation, the interaction between CSCs and stromal cells is especially critical. In this work, we developed a platform to reliably isolate single cells in suspension and grow single-cell-derived spheres for functional enrichment of CSCs. The platform also allows adherent culture of stromal cells for cancer-stromal interaction. As a proof of concept, we grew SUM149 breast cancer cells and successfully formed single-cell-derived spheres. Cancer-associated fibroblasts (CAFs) as stromal cells were found to significantly enhance the formation and growth of cancer spheres, indicating elevated tumor-initiation potential. After on-chip culture for 14 days, we retrieved single-cell derived spheres with and without CAF co-culture for single-cell transcriptome sequencing. Whole transcriptome analysis highlights that CAF co-culture can boost cancer stemness especially ALDHhigh CSCs and alter epithelial/mesenchymal status. Single-cell resolution allows identification of individual CSCs and investigation of cancer cellular heterogeneity. Incorporating whole transcriptome sequencing data with public patient database, we discovered novel genes associated with cancer-CAF interaction and critical to patient survival. The preliminary works demonstrated a reliable platform for enrichment of CSCs and studies of cancer-stromal interaction.

Single-cell RNA-sequencing of migratory breast cancer cells: discovering genes associated with cancer metastasis.

Considerable evidence suggests breast cancer metastasis arises from cells undergoing epithelial-to-mesenchymal-transition (EMT) and cancer stem-like cells (CSCs). Using a microfluidic device that enriches migratory breast cancer cells with enhanced capacity for tumor formation and metastasis, we identified genes differentially expressed in migratory cells by high-throughput single-cell RNA-sequencing. Migratory cells exhibited overall signatures of EMT and CSCs with variable expression of marker genes, and they retained expression profiles of EMT over time. With single-cell resolution, we discovered intermediate EMT states and distinct epithelial and mesenchymal sub-populations of migratory cells, indicating breast cancer cells can migrate rapidly while retaining an epithelial state. Migratory cells showed differential profiles for regulators of oxidative stress, mitochondrial morphology, and the proteasome, revealing potential vulnerabilities and unexpected consequences of drugs. We also identified novel genes correlated with cell migration and outcomes in breast cancer as potential prognostic biomarkers and therapeutic targets to block migratory cells in metastasis.

Hydro-Seq enables contamination-free high-throughput single-cell RNA-sequencing for circulating tumor cells.

Molecular analysis of circulating tumor cells (CTCs) at single-cell resolution offers great promise for cancer diagnostics and therapeutics from simple liquid biopsy. Recent development of massively parallel single-cell RNA-sequencing (scRNA-seq) provides a powerful method to resolve the cellular heterogeneity from gene expression and pathway regulation analysis. However, the scarcity of CTCs and the massive contamination of blood cells limit the utility of currently available technologies. Here, we present Hydro-Seq, a scalable hydrodynamic scRNA-seq barcoding technique, for high-throughput CTC analysis. High cell-capture efficiency and contamination removal capability of Hydro-Seq enables successful scRNA-seq of 666 CTCs from 21 breast cancer patient samples at high throughput. We identify breast cancer drug targets for hormone and targeted therapies and tracked individual cells that express markers of cancer stem cells (CSCs) as well as of epithelial/mesenchymal cell state transitions. Transcriptome analysis of these cells provides insights into monitoring target therapeutics and processes underlying tumor metastasis.

Hydrogel micropost-based qPCR for multiplex detection of miRNAs associated with Alzheimer's disease.

Quantitative polymerase chain reaction (qPCR) renders profiling of genes of interest less time-consuming and cost-effective. Recently, multiplex profiling of miRNAs has enabled identifying or investigating predominant miRNAs for various diseases such as cancers and neurodegenerative diseases. Conventional multiplex qPCR technologies mostly use colorimetric measurements in solution phase, yet not only suffer from limited multiplexing capacity but also require target-screening processes due to non-specific binding between targets and primers. Here, we present hydrogel micropost-based qPCR for multiplex detection of miRNAs associated with Alzheimer's disease (AD). Our methodology promises two key advantages compared with the conventional solution-based PCR: 1) nearly no non-specific crosstalks between targets and primers, and 2) practically valuable multiplexing by spatial encoding within a single microchamber. Specifically, we immobilized hydrogel microposts (~ 400µm in diameter) within commercially available polycarbonate PCR chips by multi-step ultraviolet (UV, 365nm) exposure. We optimized this photoimmobilization for thermal cycles of PCR as well. Acrylated forward primers incorporated in polyethylene glycol diacrylate (PEGDA) posts played a crucial role to confine fluorescent signal of cDNA amplification within the PEGDA hydrogel. To demonstrate the potential of our platform, we successfully verified multiplex detection of five miRNAs, which were reported to be highly correlated with AD, from a complex buffer of human plasma.

Microparticle-based RT-qPCR for highly selective rare mutation detection.

The quantitative reverse transcription polymerase chain reaction (RT-qPCR) has become one of the most widely used methods in the detection of disease-specific RNAs. The RT-qPCR involves two separate steps, RT and qPCR. In this study, we suggest a new RT-qPCR protocol with the particles of primer-immobilized networks (PINs), performing capture, RT and amplification of a target RNA in one particle. The production of undesired cDNAs was dramatically suppressed by the specific capture of the target RNA within the particle. Afterward, RT and amplification processes are performed without loss of cDNAs as exchanging the reaction solution. The biomarker gene of chronic myeloid leukemia, Bcr-Abl fusion transcript, is detected in the sensitivity of single mutant leukemic cell mixed in 104 normal cell using this protocol with the excellent restraint of non-specific signal. This protocol that whole processes are performed in the particle in a row is preferred for the highly specific detection of target RNAs in complex sample.

Incidence and risk factors for early-onset hypertension after allogeneic hematopoietic stem cell transplantation in children.

BACKGROUND AND OBJECTIVES: Survivors of pediatric hematopoietic stem cell transplantation (HSCT) are at risk for developing hypertension. The objectives of this study are to evaluate the prevalence and risk factors of early onset hypertension during the engraftment period after HSCT. SUBJECTS AND METHODS: This is a retrospective study of 157 consecutive patients (mean age at HSCT: 9.1±5.1 years) who underwent HSCT for acute myeloid leukemia (n=47), acute lymphoblastic leukemia (n=43), severe aplastic anemia (n=41), and other reasons (n=26). Blood pressure data were collected at five time points: 0, 7, 14, 21, and 28 days after HSCT. Hypertension was defined as having systolic and/or diastolic blood pressure ≥95th percentile according to age, gender, and height. To analyze the risk factors related to hypertension, data, including patients' demographic and transplant characteristics, were reviewed. RESULTS: Hypertension developed in 59 patients (38%), among whom 12 (7.6%) required long term therapy. Thirty-two (54%) patients had systolic and diastolic, 8 (14%) had only systolic, and 19 (32%) had only diastolic hypertension. Younger age, acute graft-versus-host disease, sinusoidal obstruction syndrome, treatment with antifungal agent, and greater increase in serum creatinine (Cr) levels were associated with hypertension. Multivariate analysis showed that younger age at HSCT and greater increase in serum Cr level were independent risk factors for hypertension. CONCLUSION: Prevalence of hypertension during immediate post-HSCT period is high, especially in younger children. A greater increase in Cr after HSCT was significantly associated with hypertension. Further study is needed to elucidate long-term cardiovascular complications in pediatric HSCT survivors.

Cytogenetic characteristics and prognosis analysis in 231 myelodysplastic syndrome patients from a single institution.

We analyzed the clinical and hematologic data of 231 patients diagnosed with de novo myelodysplastic syndrome (MDS), identified cytogenetic characteristics, and evaluated the significance of prognostic systems. The median age was 51 years and the distribution of MDS subtypes demonstrated a markedly low incidence of MDS with deletion 5q (0.9%). The proportions of World Health Organization (WHO) categories differed according to patient age group. Refractory anemia with excess blasts-2 demonstrated the most significant trend toward increased frequency with advancing age. The incidence of abnormal karyotypes in our study was comparable to a previous study (50.2%), although with different patterns. The most frequent cytogenetic abnormality was +8 (34.5% of patients with abnormality), followed by 1q+ (17.2%), 5q- (15.5%), and 20q- (12.9%). Majority of +8, 1q+, -5/5q- and -7/7q- cases combined with additional cytogenetic abnormalities (60.0%, 75.0%, 88.5% and 100%, respectively). The median survival time was 49.5 months and 13.8% patients developed acute leukemia. WHO Prognostic Scoring System (WPSS) and age group were significant factors associated with overall survival. Otherwise, International Prognostic Scoring System was not included in the model. These results demonstrated the different cytogenetic features in Korean MDS patients compared to those of Western country. In addition, WPSS and age group are applicable to our patients as an effective and reliable prognostic model.

Differential blast counts obtained by automated blood cell analyzers.

BACKGROUND: Automated blood cell analyzers often read leukemic blasts as normal cells. In this study, we evaluated the 5-part differential patterns of blasts using automated analyzers to determine if they can differentiate among blast types. METHODS: Blood samples containing 10% or more blasts were collected from patients with acute leukemia (N=175). The 5-part differential count was conducted using DxH 800 (Beckman Coulter, USA) and XE-2100 analyzers (Sysmex Co., Japan), and the results were compared with manual differential counts, which was used as a reference method. RESULTS: The DxH 800 reported the 5-part white blood cell differential count in 98.9% of the cases. The XE-2100 provided an invalid automated differential count in 72% of the cases. Both analyzers counted most lymphoblasts as lymphocytes and most myeloblasts as monocytes. In 11 cases, the DxH 800 reported a 5-part differential count without a blast flag. CONCLUSIONS: Some automated analyzers are able to recognize and count blasts according to their characteristic cell types. Therefore, complete blood counts obtained automatically can provide valuable data for making provisional decisions regarding the lineage of leukemia cells before further investigation.

[Significance of Epstein-Barr virus DNA quantitation in donors of hematopoietic stem cell transplantation].

BACKGROUND: Epstein-Barr virus (EBV) is a well-known causative agent of various diseases including post-transplant lymphoproliferative disorders. Although the level of EBV viral load in donors is expected to have a direct effect on recipients after hematopoietic stem cell transplantation (HSCT), little has been studied providing a clear evidence for that. We performed EBV DNA quantitation in donors and analyzed the effect of donors' EBV viral load on the recipients after HSCT. METHODS: EBV DNA quantitation of peripheral blood in 94 healthy HSCT donors was performed by real-time PCR. We analyzed the distribution of EBV viral load in HSCT donors and EBV positivity in the recipients transplanted from donors who had detectable EBV. RESULTS: Fifteen HSCT donors (16%) showed positive results in EBV real-time quantitative PCR. EBV viral load was below 500 copies/mL in 5 donors and above 500 (680-11,300) copies/mL in 10 donors. Five of the recipients (33.3%) transplanted from these 15 donors showed positivity in EBV PCR after HSCT. All of the EBV PCR positive recipients were transplanted from donors with viral load of >1,000 copies/mL, and 5 (71%) of 7 donors with viral load of >1,000 copies/mL was associated with posttransplant EBV PCR positivity in the recipients. CONCLUSIONS: Higher levels of EBV viral load in donors appear to be associated with EBV transmission to recipients in HSCT. EBV real-time quantitative PCR may be needed for screening EBV DNA level in HSCT donors.