Northwestern Medicine scientists have developed an AI tool called iSeg that not only matches doctors in accurately outlining lung tumors on CT scans but can also identify areas that some doctors may miss, reports a large new study.

The integration of artificial intelligence (AI) into biofunctional materials is transforming material design, synthesis, and optimization for medical applications. Machine learning and deep learning models now predict material properties (e.g., mechanical strength, degradation rate) with > 90% accuracy, dramatically reducing trial-and-error in scaffold and nanoparticle fabrication. AI-driven platforms accelerate surface functionalization strategies to enhance cell adhesion and drug loading, while generative models design stimuli-responsive hydrogels and smart polymers that mimic tissue mechanics. Case studies include rapid optimization of nanoparticle synthesis via Bayesian frameworks and the discovery of biodegradable stent materials through random forest screening. Despite remaining challenges in data quality and regulatory alignment, these advances underscore AI’s capacity to deliver high-performance, sustainable biomaterials and point toward an interdisciplinary roadmap for next-generation therapeutic solutions.

The Universitat Jaume I de Castelló (UJI) has published the second version of its Code of Good Practice in Research and Doctoral Studies (CBPID), a key tool in its commitment to responsible, high-quality research. With this update, one of the first carried out by a Spanish university, the UJI adapts its ethical framework to new challenges and needs that have arisen in recent years.

The new version of the code (the previous one was from 2022) expands and reinforces fundamental content with the addition of two new sections dedicated to the use of artificial intelligence in research and the institutional affiliation of research staff. The section on open science has also been expanded and relevant aspects surrounding the definition of conflicts of interest and the functioning of the UJI's Ethics and Integrity System have been updated.

Mouse models are central to drug development, including treatments for neurological disorders, such as Parkinson’s disease. In a new eNeuro publication, researchers from OIST’s Neuronal Rhythms in Movement Unit introduce a mouse motion capture method which provides very high-resolution measurement of mouse movement. Their marker-based approach avoids the need for extensive data processing with AI or machine learning, and can capture high-quality data on complex movement, such as running or climbing.

Aluminum (Al) exhibits excellent electrical conductivity, mechanical ductility, and good chemical compatibility with high-ionic-conductivity electrolytes. This makes it more suitable as an anode material for all-solid-state lithium batteries (ASSLBs) compared to the overly reactive metallic lithium anode and the mechanically weak silicon anode. This study finds that the pre-lithiated Al anode demonstrates outstanding interfacial stability with the Li₆PS₅Cl (LPSCl) electrolyte, maintaining stable cycling for over 1200 h under conditions of deep charge–discharge. This paper combines the pre-lithiated Al anode with a high-nickel cathode, LiNi_0.8Co_0.1Mn_0.1O₂, paired with the highly ionic conductive LPSCl electrolyte, to design an ASSLB with high energy density and stability. Using anode pre-lithiation techniques, along with dual-reinforcement technology between the electrolyte and the cathode active material, the ASSLB achieves stable cycling for 1000 cycles at a 0.2C rate, with a capacity retention rate of up to 82.2%. At a critical negative-to-positive ratio of 1.1, the battery’s specific energy reaches up to 375 Wh kg^-1, and it maintains over 85.9% of its capacity after 100 charge–discharge cycles. This work provides a new approach and an excellent solution for developing low-cost, high-stability all-solid-state batteries.