Surgery for brain cancer is a major problem in neurosurgery. The diffuse infiltration into the surrounding normal brain by these tumors makes their accurate identification by the naked eye difficult. ...Since surgery is the common treatment for brain cancer, an accurate radical resection of the tumor leads to improved survival rates for patients. However, the identification of the tumor boundaries during surgery is challenging. Hyperspectral imaging is a non-contact, non-ionizing and non-invasive technique suitable for medical diagnosis. This study presents the development of a novel classification method taking into account the spatial and spectral characteristics of the hyperspectral images to help neurosurgeons to accurately determine the tumor boundaries in surgical-time during the resection, avoiding excessive excision of normal tissue or unintentionally leaving residual tumor. The algorithm proposed in this study to approach an efficient solution consists of a hybrid framework that combines both supervised and unsupervised machine learning methods. Firstly, a supervised pixel-wise classification using a Support Vector Machine classifier is performed. The generated classification map is spatially homogenized using a one-band representation of the HS cube, employing the Fixed Reference t-Stochastic Neighbors Embedding dimensional reduction algorithm, and performing a K-Nearest Neighbors filtering. The information generated by the supervised stage is combined with a segmentation map obtained via unsupervised clustering employing a Hierarchical K-Means algorithm. The fusion is performed using a majority voting approach that associates each cluster with a certain class. To evaluate the proposed approach, five hyperspectral images of surface of the brain affected by glioblastoma tumor in vivo from five different patients have been used. The final classification maps obtained have been analyzed and validated by specialists. These preliminary results are promising, obtaining an accurate delineation of the tumor area.
With the development of deep representation learning, the domain of reinforcement learning (RL) has become a powerful learning framework now capable of learning complex policies in high dimensional ...environments. This review summarises deep reinforcement learning (DRL) algorithms and provides a taxonomy of automated driving tasks where (D)RL methods have been employed, while addressing key computational challenges in real world deployment of autonomous driving agents. It also delineates adjacent domains such as behavior cloning, imitation learning, inverse reinforcement learning that are related but are not classical RL algorithms. The role of simulators in training agents, methods to validate, test and robustify existing solutions in RL are discussed.
Plant transformation remains the most sought-after technology for functional genomics and crop genetic improvement, especially for introducing specific new traits and to modify or recombine already ...existing traits. Along with many other agricultural technologies, the global production of genetically engineered crops has steadily grown since they were first introduced 25 years ago. Since the first transfer of DNA into plant cells using Agrobacterium tumefaciens, different transformation methods have enabled rapid advances in molecular breeding approaches to bring crop varieties with novel traits to the market that would be difficult or not possible to achieve with conventional breeding methods. Today, transformation to produce genetically engineered crops is the fastest and most widely adopted technology in agriculture. The rapidly increasing number of sequenced plant genomes and information from functional genomics data to understand gene function, together with novel gene cloning and tissue culture methods, is further accelerating crop improvement and trait development. These advances are welcome and needed to make crops more resilient to climate change and to secure their yield for feeding the increasing human population. Despite the success, transformation remains a bottleneck because many plant species and crop genotypes are recalcitrant to established tissue culture and regeneration conditions, or they show poor transformability. Improvements are possible using morphogenetic transcriptional regulators, but their broader applicability remains to be tested. Advances in genome editing techniques and direct, non-tissue culture-based transformation methods offer alternative approaches to enhance varietal development in other recalcitrant crops. Here, we review recent developments in plant transformation and regeneration, and discuss opportunities for new breeding technologies in agriculture.
Videos represent the primary source of information for surveillance applications. Video material is often available in large quantities but in most cases it contains little or no annotation for ...supervised learning. This article reviews the state-of-the-art deep learning based methods for video anomaly detection and categorizes them based on the type of model and criteria of detection. We also perform simple studies to understand the different approaches and provide the criteria of evaluation for spatio-temporal anomaly detection.
ABSTRACT
Modern artificial intelligence (AI) tools built on high‐dimensional patient data are reshaping oncology care, helping to improve goal‐concordant care, decrease cancer mortality rates, and ...increase workflow efficiency and scope of care. However, data‐related concerns and human biases that seep into algorithms during development and post‐deployment phases affect performance in real‐world settings, limiting the utility and safety of AI technology in oncology clinics. To this end, the authors review the current potential and limitations of predictive AI for cancer diagnosis and prognostication as well as of generative AI, specifically modern chatbots, which interfaces with patients and clinicians. They conclude the review with a discussion on ongoing challenges and regulatory opportunities in the field.
The authors cover emerging use cases of predictive and generative artificial intelligence in three broad use cases: diagnosis, prognostication, and patient communication. They discuss limitations including lack of explainability, bias, and performance drift and strategies to overcome them.
Product quality depends upon symbiotic inter-dependencies among material, design, tooling and process parameters. Such multifarious relationships make it very difficult to avoid defects in metal ...casting—still the most economical process to produce large parts with intricate features. To get the quality ‘right first time and every time,’ we need to transform this art into science by leveraging the latest information technologies. Four relevant R&D works carried out by the author and associates over the last three decades are described in this article: (1) user-friendly simulation software to predict casting defects; (2) intelligent design and optimization of methoding and tooling; (3) cloud-based utilities for ubiquitous access; and (4) smart foundry for process data capture and analytics. These are laying the path for new technologies to reduce the difference between designed and manufactured products. Such advanced and inter-disciplinary projects are also changing the mind-set about metal casting and attracting the next generation of researchers.
Many engineers are drawn to technologybased start-up ventures for professional and personal fulfillment. Success in entrepreneurship, however, is not easy. A survey conducted at the Indian Institute ...of Technology (IIT) Bombay of 326 start-up founders, aspiring entrepreneurs, mentors, and other stakeholders showed a critical need for training in business finance, design thinking, team management, and marketing and sales. Another survey of 28 start-ups founded by alumni of the Desai Sethi (DS) School of Entrepreneurship in Mumbai, India, highlighted the role of structured mentoring programs.
Geminiviruses cause damaging diseases in several important crop species. However, limited progress has been made in developing crop varieties resistant to these highly diverse DNA viruses. Recently, ...the bacterial CRISPR/Cas9 system has been transferred to plants to target and confer immunity to geminiviruses. In this study, we use CRISPR-Cas9 interference in the staple food crop cassava with the aim of engineering resistance to African cassava mosaic virus, a member of a widespread and important family (Geminiviridae) of plant-pathogenic DNA viruses.
Our results show that the CRISPR system fails to confer effective resistance to the virus during glasshouse inoculations. Further, we find that between 33 and 48% of edited virus genomes evolve a conserved single-nucleotide mutation that confers resistance to CRISPR-Cas9 cleavage. We also find that in the model plant Nicotiana benthamiana the replication of the novel, mutant virus is dependent on the presence of the wild-type virus.
Our study highlights the risks associated with CRISPR-Cas9 virus immunity in eukaryotes given that the mutagenic nature of the system generates viral escapes in a short time period. Our in-depth analysis of virus populations also represents a template for future studies analyzing virus escape from anti-viral CRISPR transgenics. This is especially important for informing regulation of such actively mutagenic applications of CRISPR-Cas9 technology in agriculture.
Abstract Little is known about specific modes of death in patients with heart failure with preserved ejection fraction (HFpEF). Herein, the authors critically appraise the current state of data and ...offer potential future directions. They conducted a systematic review of 1,608 published HFpEF papers from January 1, 1985, to December 31, 2015, which yielded 8 randomized clinical trials and 24 epidemiological studies with mode-of-death data. Noncardiovascular modes of death represent an important competing risk in HFpEF. Although sudden death accounted for ∼25% to 30% of deaths in trials, its definition is nonspecific; it is unclear what proportion represents arrhythmic deaths. Moving forward, reporting and definitions of modes of death must be standardized and tailored to the HFpEF population. Broad-scale systematic autopsies and long-term rhythm monitoring may clarify the underlying pathology and mechanisms driving mortal events. There is an unmet need for a longitudinal multicenter, global registry of patients with HFpEF to map its natural history.