Citations

Background: Lung cancer ranks as the leading cause of cancer-related mortality worldwide. The complexity of tumor delineation, crucial for radiation therapy, requires expertise often unavailable in resource-limited settings. Artificial Intelligence (AI), particularly with advancements in deep learning (DL) and natural language processing (NLP), offers potential solutions yet is challenged by high false positive rates. Purpose: The Oncology Contouring Copilot (OCC) system is developed to leverage oncologist expertise for precise tumor contouring using textual descriptions, aiming to increase the efficiency of oncological workflows by combining the strengths of AI with human oversight. Methods: Our OCC system initially identifies nodule candidates from CT scans. Employing Language Vision Models (LVMs) like GPT-4V, OCC then effectively reduces false positives with clinical descriptive texts, merging textual and visual data to automate tumor delineation, designed to elevate the quality of oncology care by incorporating knowledge from experienced domain experts. Results: The deployment of the OCC system resulted in a 35.0% reduction in the false discovery rate, a 72.4% decrease in false positives per scan, and an F1-score of 0.652 across our dataset for unbiased evaluation. Conclusions: OCC represents a significant advance in oncology care, particularly through the use of the latest LVMs, improving contouring results by (1) streamlining oncology treatment workflows by optimizing tumor delineation and reducing manual processes; (2) offering a scalable and intuitive framework to reduce false positives in radiotherapy planning using LVMs; (3) introducing novel medical language vision prompt techniques to minimize LVM hallucinations with ablation study; and (4) conducting a comparative analysis of LVMs, highlighting their potential in addressing medical language vision challenges.

Abstract Parkinson’s disease (PD) is the fastest growing neurodegenerative disease in the world1. Gastrointestinal (GI) dysfunction can occur decades before motor impairments and in up to 80% of individuals living with PD2,3,4. We investigated peripheral relationships that may underlie environmental contributions to mechanisms along the gut-blood axis that contribute to PD pathogenesis. Single-cell multiomic spatial molecular imaging (SMI) of colonic tissue localized inflammatory injury within epithelial cells that appear to be associated with iron mishandling in both inflammatory bowel disease (IBD) and PD biosamples. We found that both the single-cell SMI of RNA and protein revealed parallel cross-modal dysregulation in the gut epithelium, in both IBD and PD biosamples. These data are accompanied by plasma (PD) and stool (IBD) protein depletion of CCL22. Our findings suggest iron mishandling along the gut barrier likely contributes to systemic inflammation, which may be the catalyst that primes circulating immune cells to body-first PD pathogenesis.

Abstract Motivation Compositional data are frequently encountered in many disciplines, such as in next-generation sequencing experiments widely used in biomedical studies. Regression analysis with compositional data as either responses or predictors has been well studied. However, when both responses and predictors are compositional, the inventory of analysis tools is surprisingly limited, especially in the high-dimensional setting. Among the few existing methods, most of them rely on a log-ratio transformation to move compositional data from the simplex to real numbers. Yet, a serious weakness of these methods is their failure to handle the substantial fraction of zeroes observed in data collected from next-generation sequencing experiments. Results To investigate associations between two high-dimensional multi-omics compositions, we propose a composition-on-composition (COC) regression analysis method which does not require log-ratio transformations and hence can handle zeroes in the data. To account for high dimensionality, we estimate regression coefficients using a penalized estimation equation approach. Finally, inference procedures for COC regression are also proposed. Superior performance of COC is demonstrated through both comprehensive numerical simulations and case studies. Availability and implementation Source R codes to implement COC method is available at https://github.com/nrios4/COC.

Background/Objectives: The accurate delineation of primary tumors (GTVp) and metastatic lymph nodes (GTVn) in head and neck (HN) cancers is essential for effective radiation treatment planning, yet remains a challenging and laborious task. This study aims to develop a deep-learning-based auto-segmentation (DLAS) model trained on external datasets with false-positive elimination using clinical diagnosis reports. Methods: The DLAS model was trained on a multi-institutional public dataset with 882 cases. Forty-four institutional cases were randomly selected as the external testing dataset. DLAS-generated GTVp and GTVn were validated against clinical diagnosis reports to identify false-positive and false-negative segmentation errors using two large language models: ChatGPT-4 and Llama-3. False-positive ruling out was conducted by matching the centroids of AI-generated contours with the slice locations or anatomical regions described in the reports. Performance was evaluated using the Dice similarity coefficient (DSC), 95th percentile Hausdorff distance (HD95), and tumor detection precision. Results: ChatGPT-4 outperformed Llama-3 in accurately extracting tumor locations from the diagnostic reports. False-positive contours were identified in 15 out of 44 cases. The DSCmean of the DLAS contours for GTVp and GTVn increased from 0.68 to 0.75 and from 0.69 to 0.75, respectively, after the ruling-out process. Notably, the average HD95 value for GTVn decreased from 18.81 mm to 5.2 mm. Post ruling out, the model achieved 100% precision for GTVp and GTVn when compared with the results of physician-determined contours. Conclusions: The false-positive ruling-out approach based on diagnostic reports effectively enhances the precision of DLAS in the HN region. The model accurately identifies the tumor location and detects all false-negative errors.

Abstract This study benchmarks the robustness and resilience of computational deconvolution methods for estimating cell-type proportions in bulk tissues, with a focus on comparing reference-based and reference-free methods. Robustness is evaluated by generating in silico pseudo-bulk tissue RNA sequencing data from cell-level gene expression profiles derived from four different tissue types, with simulated cellular composition at varying levels of heterogeneity. To assess resilience, we intentionally alter single-cell RNA profiles to create pseudo-bulk tissue RNA-seq data. Deconvolution estimates are compared with ground truth using Pearson’s correlation coefficient, root mean squared deviation, and mean absolute deviation. The results show that reference-based methods are more robust when reliable reference data are available, whereas reference-free methods excel in scenarios lacking suitable reference data. Furthermore, variations in cell-level transcriptomic profiles and cell composition have emerged as critical factors influencing the performance of deconvolution methods. This study provides significant insights into the factors affecting bulk tissue deconvolution performance, which are essential for guiding users and advancing the development of more powerful and reliable algorithms in the future.

With their unique therapeutic properties and amenability to engineering modifications, exosomes are considered a promising tool for treating peripheral nerve injuries. Through genetic engineering or physicochemical modifications, the targeting ability and functionality of exosomes can be further enhanced, enabling efficient delivery of therapeutic molecules and significantly boosting their therapeutic potential. This review aims to explore exosome-based therapies in peripheral nerve repair and their mechanisms and summarizes recent advances in the roles of exosomes in neuroprotection, axonal regeneration, vascular regeneration, and inflammatory modulation. Exosomes act as natural messengers between cells, carrying genetic materials, cytokines, and growth factors that can modulate the local microenvironment and foster peripheral nerve regeneration. Studies have shown that exosomes derived from Schwann cells, macrophages, neurons, and mesenchymal stem cells can all promote axonal regeneration and modulate inflammatory responses through various mechanisms. Compared with cell therapy, exosome therapy offers greater stability and safety, with the ability to traverse biological barriers, presenting new possibilities for regenerative medicine and precision therapy. However, challenges such as the heterogeneity and complexity of exosomes, uncertainty of in vivo distribution and metabolic pathways, and the limitations of targeting remain to be addressed. Future research will focus on the engineering modifications of exosomes to enhance their therapeutic efficiency and targeting precision.

Abstract Renal artery stenosis (RAS) is the leading cause of secondary hypertension worldwide. However, current medical and surgical treatment modalities provide minimal benefits for kidney injury. Recent preclinical RAS models have demonstrated promising potential of human mesenchymal stem cells (MSC) and their daughter extracellular vesicles (EV) in improving murine renal function and attenuating inflammation. However, the extent and mechanisms underlying immune rejection of xenogeneic MSCs or EVs are yet undetermined. Therefore, adipose tissue was harvested from adult healthy patients. Adipose-derived MSCs were extracted and cultured, and EVs were isolated from their supernatants via ultra-centrifugation. Then, mice randomly assigned to RAS or sham surgery were divided into 6 groups: sham surgery, RAS, sham + MSC, RAS + MSC, sham + EV, and RAS + EV. Two weeks after intra-aortic injection of MSCs (5 × 105) or EVs (20 µg protein), we compared the intrarenal T-cell and macrophage accumulation, splenic B-cell numbers, circulating cytokines and anti-human antibodies levels among the groups. MSCs and EVs did not influence intrarenal immune cell infiltrations. However, MSCs significantly increased circulating anti-human antibodies. In the spleen, RAS + EV mice showed higher memory IgM+ B-cells but reduced CD19+ B-cells compared to RAS + MSC. In vitro T-cell recall assay showed that both MSCs and EVs exhibited reduced IFN-γ release upon re-stimulation, indicating an immunosuppressive effect. Therefore, xenogeneic MSCs induced a greater humoral response in mice, while EVs triggered a splenic cellular response, but neither elicits discernible kidney rejection. Our results provide key insights into the immunomodulatory mechanisms of MSCs and EVs and immune mechanisms underlying xenograft rejection.

Abstract Molecular mechanisms underlying disruptions in brain glucose uptake and metabolism, linked with cognitive decline in Alzheimer’s disease (AD) patients, are only partially understood. This study investigated how soluble amyloid beta (sAβ) peptides affect glucose transport at the blood-brain barrier (BBB), the primary portal for glucose entry into the brain. We demonstrated that [ 18 F]-fluorodeoxyglucose ( 18 FDG) uptake is reduced in sAβ overproducing APP,PS1 transgenic mice compared to wild-type mice. Moreover, the influx rate of 18 FDG decreased in sAβ40 or sAβ42 pre-infused mice, highlighting the inhibitory effect of sAβ peptides on glucose transport at the BBB. Consistently, the expression of GLUT1, the primary glucose transporter at the BBB, is reduced in polarized human cerebral microvascular endothelial cell (hCMEC/D3) monolayers upon exposure to sAβ peptides and in Aβ-laden cerebral vasculature in vivo. The study further examined the influence of sAβ on the insulin-AKT pathway, known to regulate glucose uptake through modulation of thioredoxin-interacting protein (TXNIP) expression. Results showed that sAβ peptides suppress AKT phosphorylation and reduce GLUT1 expression by upregulating TXNIP levels in hCMEC/D3 monolayers. Co-incubation of resveratrol with sAβ peptides reduced TXNIP expression and rectified reductions in GLUT1 expression. In summary, toxic sAβ impairs BBB glucose transport by disrupting the insulin/AKT/TXNIP axis.

Summary The eosinophilic esophagitis (EoE) Endoscopic Reference Score (EREFS) measures endoscopic severity. While a score of ≤2 has been proposed for endoscopic response, it is unknown whether achieving this threshold results in clinically important outcomes. We aimed to determine whether an EREFS response to topical steroids (tCS) is associated with a decreased need for future esophageal dilation. In this retrospective cohort study, we included patients with a new diagnosis of EoE who underwent esophageal dilation, were then treated with tCS, and had at least two follow-up endoscopies. Endoscopic response was defined as EREFS ≤2. Histological and global symptom responses were recorded. We compared patients with and without endoscopic response and calculated the number of subsequent dilations in each group. Of 113 EoE patients, 55 (49%) had endoscopic response post-tCS. Compared to endoscopic non-responders, responders were older (43.4 vs 34.4 years; p = 0.001) and had lower EREFS (4.4 ± 1.6 vs. 6.3 ± 1.5; p < 0.001) at baseline. EREFS responders required fewer dilations on their initial post-treatment endoscopy (65% vs 90%; p = 0.002) and had greater symptom (92% vs 64%; p = 0.005) and histologic responses (82% vs 24% <15 eos/hpf; p < 0.001). Over the median follow-up time of 1106 days, EREFS responders required fewer dilations than non-responders (4.5 ± 2.9 vs 6.2 ± 4.5; p = 0.03). Endoscopic responders to tCS required fewer esophageal dilations compared to non-responders. Responders also had better symptom and histologic responses. These results provide evidence that endoscopic response is associated with important clinical outcomes including a reduction in future esophageal dilations.

Abstract Parkinson’s disease (PD) is the fastest growing neurodegenerative disease in the world 1 . Gastrointestinal (GI) dysfunction can occur decades before motor impairments and in up to 80% of individuals living with PD 2,3,4 . We investigated peripheral relationships that may underlie mechanisms along the gut-blood axis that contribute to PD pathogenesis. Single-cell multiomic spatial molecular imaging (SMI) of colonic tissue localized inflammatory injury within epithelial cells that appear to be associated with iron mishandling in both inflammatory bowel disease (IBD) and PD biosamples. We found that both the single-cell SMI of RNA and protein revealed parallel cross-modal dysregulation in the gut epithelium, in both IBD and PD biosamples. These data are accompanied by plasma (PD) and stool (IBD) protein depletion of CCL22. Our findings suggest iron mishandling along the gut barrier likely contributes to systemic inflammation, which may be the catalyst that primes circulating immune cells to body-first PD pathogenesis.