Mixup-Inf-Net: A data augmentation algorithm for segmentation of new coronary pneumonia infections

In 2020, the global spread of Coronavirus Disease 2019 exposed entire world to a severe health crisis. This has limited fast and accurate screening of suspected cases due to equipment shortages and and harsh testing environments. The current diagnosis of suspected cases has benefited greatly from the use of radiographic brain imaging, also including X-ray and scintigraphy, as a crucial addition to screening tests for new coronary pneumonia disease. However, it is impractical to gather enormous volumes of data quickly, which makes it difficult for depth models to be trained. To solve these problems, we obtained a new dataset by data augmentation Mixup method for the used chest CT slices. It uses lung infection segmentation (Inf-Net [1]) in a deep network and adds a learning framework with semi-supervised to form a Mixup-Inf-Net semi-supervised learning framework model to identify COVID-19 infection area from chest CT slices. The system depends primarily on unlabeled data and merely a minimal amount of annotated data is required;therefore, the unlabeled data generated by Mixup provides good assistance. Our framework can be used to improve improve learning and performance. The SemiSeg dataset and the actual 3D CT images that we produced are used in a variety of tests, and the analysis shows that Mixup-Inf-Net semi-supervised outperforms most SOTA segmentation models learning framework model in this study, which also enhances segmentation performance. © 2023 SPIE.

Immunotherapy: ALLOGENEIC UMBILICAL CORD BLOOD REGULATORY T CELLS DECREASE INFLAMMATION AND PRESERVE ANTI-TUMOR ACTIVITY OF CAR T CELLS

Background & Aim: With larger accessibility and increased number of patients being treated with CART cell therapy, real-world toxicity continues to remain a significant challenge to its widespread adoption. We have previously shown that allogeneic umbilical cord blood derived (UCB) regulatory T cells (Tregs) can resolve uncontrolled inflammation and can treat acute and immune mediated lung injury in a xenogenic model as well as in patients suffering from COVID-19 acute respiratory distress syndrome. The unique properties of UCB Tregs including: i) lack of plasticity when exposed to inflammatory micro-environments;ii) no requirement for HLA matching;iii) long shelf life of cryopreserved Tregs;and iv) immediate product availability for on demand treatment, makes them an attractive source for treating acute inflammatory syndromes. Therefore, we hypothesized that add-on therapy with UCB derived Tregs may resolve uncontrolled inflammation responsible for CART cell therapy associated toxicity. Methods, Results & Conclusion(s): UCB Tregs were added in 1:1 ratio to CART cells, where no interference in their ability to kill CD19+ Raji cells, was detected at different ratios : 8:1 (80.4% vs. 81.5%);4:1 (62.0% vs. 66.2%);2:1 (50.1% vs. 54.7%);1:1 (35.4% vs. 44.1%) (Fig 1A). In a xenogenic B cell lymphoma model, multiple injections of Tregs were administered after CART injection (Fig 1B), which did not impact distribution of CD8+ T effector cells (Fig 1C) or CART cells cells (Fig 1D) in different organs. No decline in the CAR T levels was observed in the Tregs recipients (Fig 1E). Specifically, no difference in tumor burden was detected between the two arms (Fig 2A). No tumor was detected in CART+Tregs in liver (Fig 2B) or bone marrow (Fig 2C). A corresponding decrease in multiple inflammatory cytokines in peripheral blood was observed in CART+Tregs when compared to CART alone (Fig 2D). Here we show "proof of concept" for add-on therapy with Tregs to mitigate hyper-inflammatory state induced by CART cells without interference in their on-target anti-tumor activity. The timing of Tregs administration after CART cells have had sufficient time for forming synapse with tumor cells allows for preservation of their anti-tumor cytotoxicity, such that the infused Tregs home to the areas of tissue damage to bind to the resident antigen presenting cells which in turn collaborate with Tregs to resolve inflammation. Such differential distribution of cells allow for a Treg "cooling blanket" and lays ground for clinical study. [Figure presented]Copyright © 2023 International Society for Cell & Gene Therapy

Segmentation Algorithm for COVID-19 Infections using Joint Mix and Context Encoder

Automatic image segmentation is critical for medical image segmentation. For example, automatic segmentation of infection area of COVID-19 before and after diagnosis and treatment can help us automatically analyze the diagnosis and treatment effect. The existing algorithms do not solve the problems of insufficient data and insufficient feature extraction at the same time. In this paper, we propose a new data augmentation algorithm to handle the insufficient data problem, named Joint Mix;we utilize an improved U-Net with context encoder to enhance the feature extraction ability. Experiments in the segmentation of COVID-19 infection region using CT images demonstrate its effectiveness. © 2022 IEEE.

Where have you been-the usage of machine learning for capturing travel history

Background and objectives: Travel history has become an indispensable part of emergency department (ED) patient assessment due to the ongoing COVID-19 pandemic. The ability to highlight travel from free text notes may augment travel history that is not completely captured in structured data fields. We explore if named-entity recognition (NER), a natural language processing (NLP) technique, can be used to extract travel history from ED free text triage notes (FTN) using a widely available, off-the-shelf, open-source NLP tool. Methods: The FTN of 10,000 attendances at an ED were reviewed by a team of annotators, and the countries, regions, or cities of travel were extracted. The annotated notes were used to train the native, out-of-the-box NER model in SpaCy 3.0.5, which is based on a deep convolutional neural network. Predictions made by the trained model were evaluated on a previously unseen test set. Results: The NER model achieved F1 score of 97.64%, precision of 98.68%, and recall of 96.6% in capturing travel history. Conclusion: Machine learning can be used to accurately capture travel history from ED FTN.

Randomized, Multi-Center, Double-Blinded, Placebo Controlled Safety and Early Efficacy Trial of Cryopreserved Cord Blood Derived T-Regulatory Cell Infusions (CK0802) in the Treatment of COVID-19 Induced ARDS. (RESOLVE Trial)

Background. COVID19 associated moderate to severe acute respiratory distress syndrome (ARDS) is associated with high rates of morbidity and mortality. Immune dysfunction and hyper-inflammatory responses result in a vicious cycle of tissue inflammation and end organ damage. Based on the suggestion of early efficacy of adoptive therapy with allogeneic T regulatory cells in COVID19 ARDS (Gladstone et al., Ann Int Med 2020), Cellenkos ® initiated a randomized, placebo controlled, multi-center trial of multiple doses of CK0802 (allogeneic, off-the-shelf, cryopreserved, cord blood derived T regulatory cells) for treatment of moderate-to-severe COVID19-related ARDS patients. Study design. Multi-center, randomized, blinded, placebo controlled trial of CK0802 at two different doses (100 million cells and 300 million cells ) were compared to placebo. Each patient was randomized to receive the assigned product on days 0, 3 and 7 (Figure 1), without HLA matching. Enrollment was staggered for the first 6 active treatment patients with 7 days between each patient while monitoring for any safety signals. Subsequent patients were enrolled on a continuous basis. DSMB monitoring occurred after every cohort of 15 patients (5 controls;5 of each active treatment). Results are presented as median (with range) unless otherwise indicated. Primary Outcomes. The two co-primary outcomes were: • Dose Limiting Toxicity (DLT) = Regimen related grade 3, 4, or 5 toxicity within 48 hours of first infusion • S28 = [Alive and not intubated 28 days after the date of first infusion] = 28-day treatment success Secondary Outcomes. Secondary outcomes, recorded from first day of infusion up to 28 days later, included: i) time to extubation, ii) ventilator-free days;iii) organ failure-free days;iii) ICU free days;iv) PaO 2/FiO 2 between days 0 and 11;and v) 28-day all-cause mortality Covariates. Patient covariates recorded at enrollment included: i) age, ii) gender iii) on vasopressors;iii) on hemodialysis;iv) duration of intubation prior to enrollment. Study Conduct. The multicenter study (n=5 centers) was activated in October 2020 and enrollment completed in March 2021. Results. Forty-five patients were enrolled (60% male, median age 60 [range 21-85], 46.7% Caucasian race). At baseline 13% were on hemodialysis;62% on vasopressors;SOFA score=8 (6-13);PaO 2=85 mmHg (45-133);FiO 2=60% (40-100);PEEP=10 cmH 2O (5-18) with a median duration of intubation of 48 hrs (0-120) prior to enrollment. Patient were intubated a median of 72 [0-144] hours prior to infusion. Sixty percent of patients were alive and extubated at day 28. Median time to extubation from first infusion was 10.5 [2-46] days and median ventilator free days at day 28 was 12 [0-26]days. No treatment related SAEs were reported. Time to extubation from first infusion was 10.5 days (2-46) and at day 28 the ventilator free days were 12 (0-26). The estimated day 28 overall survival was 78.6% with the following breakdown according to the co-variates: i) age>60 yrs =77.5% vs. age<60yrs=79.9%;ii) female=85.7% vs male=73.7%;iii) on vasopressor=65.8% vs. 77.8%;iv) on hemodialysis=75% vs. 79%. Duration of intubation to enrollment had no impact on 28d survival. At baseline, 14 pts were positive for both HLA I and HLA II antibodies (Abs);3 pts positive for HLA I Ab only, and 9 positive for HLA II Ab only. In 20 paired samples collected on day 0 and day 28, HLA I Ab and HLA II Ab seroconversion was observed in 4 and 1 pt, respectively. Discussion This is the first clinical trial to examine safety and early efficacy of multiple doses of allogenic, off-the-shelf, cryopreserved, T regulatory cells for the treatment of COVID-19-related ARDS. Full data analysis of treatment groups (placebo;CK0802-100 million;CK0802-300 million) is ongoing and will be presented at the conference. Additional data to be presented will include: 3- and 6- month QOL, mental health, and cognitive index analyses, and paired Biomarker analysis. [Formula presented] Disclosures: Hari: Janssen: Honoraria, Membership on an ntity's Board of Directors or advisory committees, Other, Research Funding, Speakers Bureau;GSK: Consultancy, Membership on an entity's Board of Directors or advisory committees, Other, Research Funding, Speakers Bureau;Celgene-BMS: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Other, Research Funding, Speakers Bureau;Millenium: Membership on an entity's Board of Directors or advisory committees, Research Funding;Sanofi: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau;Takeda: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Other, Research Funding, Speakers Bureau;Adaptive Biotech: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau;Amgen: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Other, Research Funding, Speakers Bureau;Karyopharm: Consultancy;Oncopeptides: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Sadeghi: Cellenkos Inc.: Current Employment. Parmar: Cellenkos Inc.: Current holder of individual stocks in a privately-held company, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties, Research Funding. Mukherjee: Vor Biopharma: Current equity holder in publicly-traded company, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties: coinventor on issued and pending patent applications licensed to Vor Biopharma. S.M. has equity ownership and is on the Scientific Advisory Board of Vor Biopharma., Research Funding.

WHEN FACE RECOGNITION MEETS OCCLUSION: A NEW BENCHMARK

The existing face recognition datasets usually lack occlusion samples, which hinders the development of face recognition. Especially during the COVID-19 coronavirus epidemic, wearing a mask has become an effective means of preventing the virus spread. Traditional CNN-based face recognition models trained on existing datasets are almost ineffective for heavy occlusion. To this end, we pioneer a simulated occlusion face recognition dataset. In particular, we first collect a variety of glasses and masks as occlusion, and randomly combine the occlusion attributes (occlusion objects, textures,and colors) to achieve a large number of more realistic occlusion types. We then cover them in the proper position of the face image with the normal occlusion habit. Furthermore, we reasonably combine original normal face images and occluded face images to form our final dataset, termed as Webface-OCC. It covers 804,704 face images of 10,575 subjects, with diverse occlusion types to ensure its diversity and stability. Extensive experiments on public datasets show that the ArcFace retrained by our dataset significantly outperforms the state-of-the-arts. Webface-OCC is available at https://github.com/Baojin-Huang/Webface-OCC.

Adoptive Therapy with Cord Blood T Regulatory Cells Enhances Anti-Myeloma Efficacy of T Cell Based Immunotherapies

Background. Inflammatory tumor microenvironment leads to T cell exhaustion in multiple myeloma leading to treatment failure and relapse. Specifically, T cell based therapies including bispecific antibodies and chimeric antigen receptor (CAR) T cell are associated with the additional side effects of non-specific T cell activation and cytokine release syndrome. Adoptive therapy with allogeneic cord blood (CB) T regulatory (Treg) cell therapy has been shown to be safe with clinical efficacy in a wide range of diseases including graft vs. host disease (GvHD), inflammatory bone marrow failures and COVID-19 induced acute respiratory distress syndrome. Furthermore, combination of Tregs with donor lymphocyte infusion (DLI) has led to resolution of leukemia relapse without GvHD flare up. We hypothesize that co-administration of Tregs with adoptive T cell based therapy will improve myeloma outcomes. Methods. 3x106 GFP-labeled MM.1S cells were injected into NSG mice followed by 5x106 CD3+ T conventional (Tcon) cells on day 14. In a subset of the Tcon treated mice, 1x107 CB Treg cells were injected on day 47, 54 and 61. Mice were followed every other day for weight and GvHD score. Non-invasive bioluminescent imaging (BLI) was performed serially. Weekly blood draw was performed for cell analysis and cytokine assays. At the time of euthanasia, blood, spleen and bone marrow were harvested for histopathology and flow analysis. In a subsequent experiment, intra-peritoneal injection of the bi-specific antibody against CD3 and BCMA (BCMA-BiTE) was administered in the xenogenic myeloma model in the presence or absence of CB Treg cells. Pan T cells were injected into all mice to facilitate the anti-tumor action of BiTE. Results. Both Tcon and Tcon+Treg recipients maintained their body weight compared to myeloma alone or myeloma + Treg arm (Figure A). All mice showed evidence of tumor growth by day 20 (Figure A). Widespread MM.1S cell growth in the myeloma only mice at day 27 was demonstrated by BLI whereas no measurable tumor growth was evident in Tcon recipients or Tcon+Treg recipients. By day 69, Tcon only mice were significantly increased tumor growth compared to Tcon+Treg recipients (Figure B). While circulating multiple myeloma cells were detected in myeloma alone and myeloma+Treg arm, no such evidence was detectable in the Tcon or Tcon+Treg recipients. However, upon euthanasia, extramedullary relapse of myeloma as retroperitoneal mass was detected in Tcon recipient (Figure C). Addition of Treg + BiTE led to a similar degree of tumor control compared to BiTE alone treated mice, however, a significant weight loss was observed in this arm (Figure D) with a corresponding high GvHD score (Figure E). Furthermore, addition of CB Treg cells led to decrease of T cell exhaustion phenotypic markers (data not shown). Conclusion. We are the first to show that CB Treg cells can be administered in combination with the T-cell based immunotherapies directed against myeloma. Such a strategy should be examined in the clinical setting. [Formula presented] Disclosures: Nishimoto: Bayer Yakuhin, Ltd:: Research Funding;Janssen Pharmaceutical K.K.:: Research Funding. Sadeghi: Cellenkos Inc.: Current Employment. Shah: GSK, Amgen, Indapta Therapeutics, Sanofi, BMS, CareDx, Kite, Karyopharm: Consultancy;BMS, Janssen, Bluebird Bio, Sutro Biopharma, Teneobio, Poseida, Nektar: Research Funding. Patel: Nektar: Consultancy, Research Funding;Celgene: Consultancy, Research Funding;Cellectis: Research Funding;Takeda: Consultancy, Research Funding;Janssen: Consultancy, Research Funding;Oncopeptides: Consultancy;Poseida: Research Funding;Precision Biosciences: Research Funding;Bristol Myers Squibb: Consultancy, Research Funding. Parmar: Cellenkos Inc.: Current equity holder in private company, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties, Research Funding.

Adoptive Cord Blood T Regulatory Cell Therapy Leads to Resolution of Inflammation and Decreased Proteinuria in Lupus Nephritis

Background. Systemic lupus erythematosus (SLE) is associated with widespread inflammation with multi-organ involvement where renal failure is the most dreaded and fatal complication. Adoptive therapy with cord blood (CB) derived T regulatory cells has been shown to improve outcomes in disease driven by inflammation including graft versus host disease (GvHD), inflammatory bone marrow disorder, COVID-19 induced acute respiratory distress syndrome. We hypothesize that CB Treg therapy can treat lupus nephritis. Methods. The suppressive abilities of CB Tregs expressing CD4+CD25highCD127lowFoxP3+ were assayed by human cytokines assay kits (IL-10, IFN-γ, IP-10, TNF-α, IL-6, and IL-17A) in the cell culture supernatants. For examining the efficacy of CB Tregs in vivo, SLE xenograft model was created with female Rag2-/-γc-/- mice transplanted with 3x106 human SLE-peripheral blood mononuclear cells (PBMCs) by intravenous injection on day 0. The mice were allowed to develop disease and on day 30 post-transplant, they were divided into 2 groups: i) control and ii) treatment. 1x107 ex vivo-expanded, cryopreserved, allogeneic, non-HLA matchedCB Tregs were injected into SLE xenografts intravenously once per week for 4 weeks through the tail vein. Serial blood draws were performed for the phenotypic analysis, cytokine assay and anti-double stranded (ds)DNA IgG antibody analysis. Serial examination of the urine samples was performed for creatinine and albumin quantification. Histopathologic examination of the harvested organs was performed at the time of planned euthanasia at 13 weeks. Results. Co-culture of CB Tregs with the pathogenic SLE-PBMCs decreased the secretion of inflammatory cytokines including IFN-γ, IP-10, TNF-α, IL-6, and IL-17A (Figure A) with a reciprocal increase in the secretion of the anti-inflammatory IL-10 cytokine (Figure B). Adoptive therapy with CB Treg cells led to a significant decrease in circulating CD8+ effector T cells and an increase in CD4+ helper T cells (Figure C). CB Treg recipients showed preserved weight gain (Figure D), lower GvHD score (Figure E) and improved overall survival (Figure F). A significant decrease in proteinuria at 9 weeks post-transplant (Figure G) correlated with a decrease in anti-dsDNA IgG Ab levels (Figure H) and soluble CD40 ligand levels (Figure I). Histopathological results from two index cases from each arm (Figure J) demonstrated that CB Treg recipients show reduced T-cells (CD3+) (Figure K) and B-cells (CD20+) (Figure L) in the kidneys, as well as a decrease in the lymphoid infiltration into glomeruli and renal parenchyma as compared to the control arm (Figure M). Conclusion. We are the first to demonstrate the benefit of allogeneic CB Treg cell therapy for treatment of lupus nephritis. We propose to examine such a strategy in the clinical setting. [Formula presented] Disclosures: Nishimoto: Janssen Pharmaceutical K.K.:: Research Funding;Bayer Yakuhin, Ltd:: Research Funding. Sadeghi: Cellenkos Inc.: Current Employment. Parmar: Cellenkos Inc.: Current equity holder in private company, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties, Research Funding.