This paper proposes a new enhanced model architecture to perform classification of lumbar spine degeneration with DICOM images while using a hybrid approach, integrating EfficientNet and VGG19 together with custom-designed components. The proposed model is differentiated from traditional transfer learning methods as it incorporates a Pseudo-Newton Boosting layer along with a Sparsity-Induced Feature Reduction Layer that forms a multi-tiered framework, further improving feature selection and representation. The Pseudo-Newton Boosting layer makes smart variations of feature weights, with more detailed anatomical features, which are mostly left out in a transfer learning setup. In addition, the Sparsity-Induced Layer removes redundancy for learned features, producing lean yet robust representations for pathology in the lumbar spine. This architecture is novel as it overcomes the constraints in the traditional transfer learning approach, especially in the high-dimensional context of medical images, and achieves a significant performance boost, reaching a precision of 0.9, recall of 0.861, F1 score of 0.88, loss of 0.18, and an accuracy of 88.1%, compared to the baseline model, EfficientNet. This work will present the architectures, preprocessing pipeline, and experimental results. The results contribute to the development of automated diagnostic tools for medical images.

Large language models (LLMs) present a valuable technology for various applications in healthcare, but their tendency to hallucinate introduces unacceptable uncertainty in critical decision-making situations. Human-AI collaboration (HAIC) can mitigate this uncertainty by combining human and AI strengths for better outcomes. This paper presents a novel guided deferral system that provides intelligent guidance when AI defers cases to human decision-makers. We leverage LLMs' verbalisation capabilities and internal states to create this system, demonstrating that fine-tuning small-scale LLMs with data from large-scale LLMs greatly enhances performance while maintaining computational efficiency and data privacy. A pilot study showcases the effectiveness of our proposed deferral system.

The hard-and software that catalysed rapid developments in machine learning In late 2002 and early 2003, the release of the Radeon 9700 and GeForce FX video cards introduced a fully programmable graphics pipeline, extending and later replacing the existing fixed function pipelines. Unlike the fixed function pipeline, which allowed the user to only supply input matrices and parameters to built-in operations, the programmable pipeline introduced the execution of user-written shader programs on the GPU [Contributors, 2015]. This fundamental change allowed programmers and researchers to exploit the intrinsic parallelism of GPUs 2 years before Intel would introduce its first dual-core CPU. Within months of the availability of this new hardware and the accompanying APIs, researchers implemented linear algebra methods on GPUs and introduced programming frameworks to use GPUs for generalpurpose computations [Thompson et al., 2002, Krüger and Westermann, 2003]. This rapid development marked the dawn of general-purpose computing on graphics processing units (GPGPU). In a presentation at ICS '08, Harris presented the successes of GPGPU by highlighting a speed-up in molecular docking, N-body simulations, HD video stream transcoding, or image processing--applications in machine learning were not discussed. However, just one year later, the introduction of GPUs as general-purpose processors catalysed the deep learning explosion of the early 2010s by allowing deep learning algorithms pioneered by Alexey Ivakhnenko in 1971 to be run within practical time on widely available consumer hardware when Rajat et al. showed that GPUs outperform CPUs by an order of magnitude in large-scale deep unsupervised learning tasks [Ivakhnenko, 1971, Raina et al., 2009]. Hardware and energy requirements increase in machine learning research In 2010, Ciresan et al. [2010] introduced a multi-layer perceptron (MLP) with up to 12.11 million free parameters where forward and backward propagation were implemented on a GPU using NVIDIA's proprietary CUDA API introduced by Harris at ICS '08 two

Developed by a team at Cedars-Sinai, the novel AI technology can accurately forecast early signs of a heart attack, predicting which patients will experience a heart attack in five years based on the level and composition of plaque in arteries that supply the heart with blood. The findings of the team's research, which was funded by the National Heart, Lung, and Blood Institute, are published in The Lancet Digital Health. When plaque builds up, it can result in a narrowing of the arteries, making it more challenging for blood to be transported to the heart, which increases the chances of a heart attack. Traditionally, health professionals employ a medical test called coronary computed tomography angiography (CTA) to capture 3D images of the heart and arteries, which gives the doctors an estimate of how much a patient's arteries have narrowed. However, until now, there has not been an efficient, automated, and rapid method for measuring the plaque shown in the CTA images.

In research funded by the British Heart Foundation (BHF), the team developed the biomarker, or'fingerprint' – called the fat radiomic profile (FRP), using machine learning. The FRP reveals biological red flags in the perivascular space lining blood vessels which supply blood to the heart. Furthermore, the tool identifies inflammation, scarring, and changes to these blood vessels, which all indicate the chances of a heart attack in the future. Very often when an individual goes to the hospital with chest pain, a standard component of care is to have a coronary CT angiogram (CCTA). This is a scan of the coronary arteries to check for any narrowed or blocked segments.

Artificial intelligence could be used to predict those at risk of a fatal heart attack up to five years in advance, new research has found. Experts at the University of Oxford have developed a "fingerprint", or biomarker, using machine learning. When a patient is admitted to hospital with chest pain, it's standard procedure for a coronary CT angiogram (CCTA) to be performed. If no narrowing of the arteries is detected – about 75% of cases – then the patient is sent home – yet some of them suffer a heart attack in the future. There's currently no method routinely used by doctors to spot all underlying red flags of a future heart attack.

Coronary heart disease (CHD) is one of the leading causes of death in the UK; it's responsible for more than 66,000 deaths each year, and it's estimated that 2.3 million people in the UK are currently living with the disease. Imaging is an important part of diagnosing people and managing treatment with a range of cardiovascular diseases, and the NHS uses Echo, MRI and CT scans to do this. A CT scan is used to do a coronary CT angiography, which involves taking images of the heart blood vessels and using the CT scan with some dye contrast into the blood vessels. This enables doctors to see if blood vessels are narrowing, causing patients to have chest pain or angina. The test has been around for the last 10 or 15 years but has increased in usage as radiation doses have decreased dramatically.

Recently, the notion of accrual of arguments has received some attention from the argumentation community. Three principles for argument accrual have been identified as necessary to hold in argumentation frameworks. In this paper we propose an approach to model the accrual of arguments in the context of Defeasible Logic Programming, a logic programming approach to argumentation which has proven to be successful for many real-world applications. We will analyze the above mentioned principles in the context of our proposal, studying other interesting properties.