In a prospective study enrolling 1,222 patients from 22 emergency departments, a device using a machine-learning-based signature of blood mRNAs demonstrated clinically acceptable performance to diagnose bacterial and viral infections and to predict the all-cause need for critical care interventions within 7 days, with benchmark to established biomarkers and risk scores.
Talk of AI is inescapable. It’s often the main topic of discussion at board and executive meetings, at corporate retreats, and in the media. A record 58% of S&P 500 companies mentioned AI in their second-quarter earnings calls, according to Goldman Sachs.
But it’s difficult to walk the talk. Just 5% of generative AI pilots are driving measurable profit-and-loss impact, according to a recent MIT study. That means 95% of generative AI pilots are realizing zero return, despite significant attention and investment.
Although we’re nearly three years past the watershed moment of ChatGPT’s public release, the vast majority of organizations are stalling out in AI. Something is broken. What is it?
Date from Lucid’s AI readiness survey sheds some light on the tripwires that are making organizations stumble. Fortunately, solving these problems doesn’t require recruiting top AI talent worth hundreds of millions of dollars, at least for most companies. Instead, as they race to implement AI quickly and successfully, leaders need to bring greater rigor and structure to their operational processes.
Operations are the gap between AI’s promise and practical adoption
I can’t fault any leader for moving as fast as possible with their implementation of AI. In many cases, the existential survival of their company—and their own employment—depends on it. The promised benefits to improve productivity, reduce costs, and enhance communication are transformational, which is why speed is paramount.
But while moving quickly, leaders are skipping foundational steps required for any technology implementation to be successful. Our survey research found that more than 60% of knowledge workers believe their organization’s AI strategy is only somewhat to not at all well aligned with operational capabilities.
AI can process unstructured data, but AI will only create more headaches for unstructured organizations. As Bill Gates said, “The first rule of any technology used in a business is that automation applied to an efficient operation will magnify the efficiency. The second is that automation applied to an inefficient operation will magnify the inefficiency.”
Where are the operations gaps in AI implementations? Our survey found that approximately half of respondents (49%) cite undocumented or ad-hoc processes impacting efficiency sometimes; 22% say this happens often or always.
The primary challenge of AI transformation lies not in the technology itself, but in the final step of integrating it into daily workflows. We can compare this to the “last mile problem” in logistics: The most difficult part of a delivery is getting the product to the customer, no matter how efficient the rest of the process is.
In AI, the “last mile” is the crucial task of embedding AI into real-world business operations. Organizations have access to powerful models but struggle to connect them to the people who need to use them. The power of AI is wasted if it’s not effectively integrated into business operations, and that requires clear documentation of those operations.
Capturing, documenting, and distributing knowledge at scale is critical to organizational success with AI. Yet our survey showed only 16% of respondents say their workflows are extremely well-documented. The top barriers to proper documentation are a lack of time, cited by 40% of respondents, and a lack of tools, cited by 30%.
The challenge of integrating new technology with old processes was perfectly illustrated in a recent meeting I had with a Fortune 500 executive. The company is pushing for significant productivity gains with AI, but it still relies on an outdated collaboration tool that was never designed for teamwork. This situation highlights the very challenge our survey uncovered: Powerful AI initiatives can stall if teams lack modern collaboration and documentation tools.
This disconnect shows that AI adoption is about more than just the technology itself. For it to truly succeed enterprise-wide, companies need to provide a unified space for teams to brainstorm, plan, document, and make decisions. The fundamentals of successful technology adoption still hold true: You need the right tools to enable collaboration and documentation for AI to truly make an impact.
Collaboration and change management are hidden blockers to AI implementation
A company’s approach to AI is perceived very differently depending on an employee’s role. While 61% of C-suite executives believe their company’s strategy is well-considered, that number drops to 49% for managers and just 36% for entry-level employees, as our survey found.
Just like with product development, building a successful AI strategy requires a structured approach. Leaders and teams need a collaborative space to come together, brainstorm, prioritize the most promising opportunities, and map out a clear path forward. As many companies have embraced hybrid or distributed work, supporting remote collaboration with digital tools becomes even more important.
We recently used AI to streamline a strategic challenge for our executive team. A product leader used it to generate a comprehensive preparatory memo in a fraction of the typical time, complete with summaries, benchmarks, and recommendations.
Despite this efficiency, the AI-generated document was merely the foundation. We still had to meet to debate the specifics, prioritize actions, assign ownership, and formally document our decisions and next steps.
According to our survey, 23% of respondents reported that collaboration is frequently a bottleneck in complex work. Employees are willing to embrace change, but friction from poor collaboration adds risk and reduces the potential impact of AI.
Operational readiness enhances your AI readiness
Operations lacking structure are preventing many organizations from implementing AI successfully. We asked teams about their top needs to help them adapt to AI. At the top of their lists were document collaboration (cited by 37% of respondents), process documentation (34%), and visual workflows (33%).
Notice that none of these requests are for more sophisticated AI. The technology is plenty capable already, and most organizations are still just scratching the surface of its full potential. Instead, what teams want most is ensuring the fundamentals around processes, documentation, and collaboration are covered.
AI offers a significant opportunity for organizations to gain a competitive edge in productivity and efficiency. But moving fast isn’t a guarantee of success. The companies best positioned for successful AI adoption are those that invest in operational excellence, down to the last mile.
This content was produced by Lucid Software. It was not written by MIT Technology Review’s editorial staff.
In the proof-of-concept phase 2 ROME trial, comprehensive genomic profiling followed by molecular tumor board evaluation and randomization of patients with metastatic solid cancer to receive personalized therapy or standard of care led to a significantly higher objective response rate and longer progression-free survival in patients who received personalized therapy.
Metabolites. 2025 Aug 28;15(9):575. doi: 10.3390/metabo15090575.
ABSTRACT
INTRODUCTION: Hepatocellular carcinoma (HCC), the most prevalent form of liver cancer, ranks as the third leading cause of mortality globally. Patients diagnosed with HCC exhibit a dismal prognosis mostly due to the emergence of symptoms in the advanced stages of the disease. Moreover, conventional biomarkers demonstrate insufficient efficacy in the early detection of HCC, hence highlighting the need for the identification of novel and more effective biomarkers.
METHODS: In this paper, we investigate methods for integration of multi-omics data we generated by both untargeted and targeted mass spectrometric analysis of serum samples from HCC cases and patients with liver cirrhosis. Specifically, the performances of several feature selection methods are evaluated on their abilities to identify a panel of multi-omics features that distinguish HCC cases from cirrhotic controls.
RESULTS: The integrative analysis identified key molecules associated with liver including such as leucine and isoleucine as well as SERPINA1, which is involved in LXR/RXR Activation and Acute Response signaling. A new method that uses recursive feature selection in conjunction with a transformer-based deep learning model as an estimator led to more promising results compared to other deep learning methods that perform disease classification and feature selection sequentially.
CONCLUSIONS: The findings in this study reinforce the importance of adapting or extending deep learning models to support robust feature selection, especially for integration of multi-omics data with limited sample size to avoid the risk of overfitting and the need for evaluation of the multi-omics features discovered in this study via blood samples from a larger and independent cohort to identify robust biomarkers for HCC.
Spatial 'omics technologies are a powerful tool for mapping the relationship between cellular organization and molecular distributions in healthy and diseased tissue microenvironments. Here, we describe a novel multimodal pipeline that represents experimental and computational advances for spatiomolecular analysis of tissue samples across molecular classes. This adaptable method integrates matrix-assisted laser desorption/ionization (MALDI) imaging mass spectrometry (IMS) lipidomics, spatial transcriptomics (ST), multiplexed immunofluorescence microscopy (MxIF), and histopathological staining to uncover spatiomolecular profiles associated with unique cellular niches and pathological features. We demonstrate the power of this approach using two different complex human disease systems: Alzheimer's disease in human brain tissue and type 2 diabetes mellitus in the human pancreas. By identifying molecular markers associated with disease pathology in the pancreas and brain, we shed light on biologically significant pathways that are impacted in these two spatially complex diseases and highlight the powerful potential of accurate, high-resolution multimodal integration approaches.
Spatial technologies examining the cell and tissue microenvironment at near single-cell resolution are revealing important molecular insights. However, few tools enable integrated, interactive analysis of spatial-omics with tissue morphology in the same functional tissue unit. Here, we present FUSION (Functional Unit State Identification in Whole Slide Images), a web-based platform for visualizing and analyzing spatial-omics data with high-resolution histology. FUSION provides workflows for assessing cell compositions, quantitative morphometrics, and comparative tissue analyses. We demonstrate applicability across spatial assays, including 10x Visium, Visium HD, 10x Xenium, Cell DIVE, and PhenoCycler, applied to healthy and diseased tissues from kidney, small intestine, lung, and skin in the Human BioMolecular Atlas Program. FUSION is cloud-based, open-source, and accessible at https://fusion.hubmapconsortium.org/ , hosting over 50 paired datasets and tutorials. In a series of use cases, we show its capacity to distinguish renal glomeruli injury states, quantify morphometric changes, and characterize fibrosis with immune infiltration.
Background: Clinical decision support systems (CDSSs) have gained prominence in health care, aiding professionals in decision-making and improving patient outcomes. While physicians often use CDSSs for diagnosis and treatment optimization, nurses rely on these systems for tasks such as patient monitoring, prioritization, and care planning. In nursing practice, CDSSs can assist with timely detection of clinical deterioration, support infection control, and streamline care documentation. Despite their potential, the adoption and use of CDSSs by nurses face diverse challenges. Barriers such as alarm fatigue, limited usability, lack of integration with workflows, and insufficient training continue to undermine effective implementation. In contrast to the relatively extensive body of research on CDSS use by physicians, studies focusing on nurses remain limited, leaving a gap in understanding the unique facilitators and barriers they encounter. Objective: This study aimed to explore the facilitators and barriers influencing the adoption and use of CDSSs by nurses in hospitals, using an extended Fit Between Individuals, Tasks, and Technology (FITT) framework. Methods: A qualitative study was conducted using semistructured interviews with 22 nurses from across the Netherlands, representing 3 hospital types: general (n=9), top-clinical (n=12), and academic (n=1). The sample included a diverse mix of practicing nurses, nurses-in-training, and clinical nurse information officers, with clinical experience ranging from 1.5 to 38 years. Interview transcripts were analyzed thematically, beginning with an inductive coding approach to identify key factors. These were then categorized deductively using the extended FITT framework. In total, 988 code instances were examined. To ensure analytical rigor, the coding process was separately conducted by 2 researchers and reviewed by an expert panel. Results: A total of 26 distinct factors were identified, categorized into 4 FITT dimensions: technology-individual, technology-task, task-individual, and organizational context. Of these, 11 factors were facilitators (eg, cognition, clarification, and prevention), 7 were barriers (eg, alarm fatigue, poor design, and limited digital proficiency), and 8 were both facilitators and barriers depending on the context (eg, acceptance, workload, and training). In addition, key value tensions emerged, such as the balance between standardization and professional autonomy, and the trade-off between enhanced decision support and increased administrative burden. Conclusions: The findings underscore the complexity of CDSS adoption in nursing practice, highlighting the interaction of facilitators and barriers across FITT dimensions. Practical recommendations include participatory design processes, targeted training programs, advanced alert management systems, and strong organizational support. Addressing value tensions and aligning CDSS functionality with nurses’ workflows can enhance adoption and optimize patient outcomes. Trial Registration:
The Biodiversity Cell Atlas aims to create comprehensive single-cell molecular atlases across the eukaryotic tree of life, which will be phylogenetically informed, rely on high-quality genomes and use shared standards to facilitate comparisons across species.
Background: Diabetes-related foot ulceration (DFU) is a common complication of diabetes, with a significant impact on survival, health care costs, and health-related quality of life. The prognosis of DFU varies widely among individuals. The International Working Group on the Diabetic Foot recently updated their guidelines on how to classify ulcers using “classical” classification and scoring systems. No system was recommended for individual prognostication, and the group considered that more detail in ulcer characterization was needed and that machine learning (ML)–based models may be the solution. Despite advances in the field, no assessment of available evidence was done. Objective: This study aimed to identify and collect available evidence assessing the ability of ML-based models to predict clinical outcomes in people with DFU. Methods: We searched the MEDLINE database (PubMed), Scopus, Web of Science, and IEEE Xplore for papers published up to July 2023. Studies were eligible if they were anterograde analytical studies that examined the prognostic abilities of ML models in predicting clinical outcomes in a population that included at least 80% of adults with DFU. The literature was screened independently by 2 investigators (MMS and DAR or EH in the first phase, and MMS and MAS in the second phase) for eligibility criteria and data extracted. The risk of bias was evaluated using the Quality In Prognosis Studies tool and the Prediction model Risk Of Bias Assessment Tool by 2 investigators (MMS and MAS) independently. A narrative synthesis was conducted. Results: We retrieved a total of 2412 references after removing duplicates, of which 167 were subjected to full-text screening. Two references were added from searching relevant studies’ lists of references. A total of 11 studies, comprising 13 papers, were included focusing on 3 outcomes: wound healing, lower extremity amputation, and mortality. Overall, 55 predictive models were created using mostly clinical characteristics, random forest as the developing method, and area under the receiver operating characteristic curve (AUROC) as a discrimination accuracy measure. AUROC varied from 0.56 to 0.94, with the majority of the models reporting an AUROC equal or superior to 0.8 but lacking 95% CIs. All studies were found to have a high risk of bias, mainly due to a lack of uniform variable definitions, outcome definitions and follow-up periods, insufficient sample sizes, and inadequate handling of missing data. Conclusions: We identified several ML-based models predicting clinical outcomes with good discriminatory ability in people with DFU. Due to the focus on development and internal validation of the models, the proposal of several models in each study without selecting the “best one,” and the use of nonexplainable techniques, the use of this type of model is clearly impaired. Future studies externally validating explainable models are needed so that ML models can become a reality in DFU care. Trial Registration: PROSPERO CRD42022308248; https://www.crd.york.ac.uk/PROSPERO/view/CRD42022308248
Background: Large language models (LLMs) hold promise for supporting clinical tasks, particularly in data-driven and technical disciplines such as radiation oncology. While prior evaluation studies have focused on examination-style settings for evaluating LLMs, their performance in real-life clinical scenarios remains unclear. In the future, LLMs might be used as general AI assistants to answer questions arising in clinical practice. It is unclear how well a modern LLM, locally executed within the infrastructure of a hospital, would answer such questions compared with clinical experts. Objective: This study aimed to assess the performance of a locally deployed, state-of-the-art medical LLM in answering real-world clinical questions in radiation oncology compared with clinical experts. The aim was to evaluate the overall quality of answers, as well as the potential harmfulness of the answers if used for clinical decision-making. Methods: Physicians from 10 departments of European hospitals collected questions arising in the clinical practice of radiation oncology. Fifty of these questions were answered by 3 senior radiation oncology experts with at least 10 years of work experience, as well as the LLM Llama3-OpenBioLLM-70B (Ankit Pal and Malaikannan Sankarasubbu). In a blinded review, physicians rated the overall answer quality on a 5-point Likert scale (quality), assessed whether an answer might be potentially harmful if used for clinical decision-making (harmfulness), and determined if responses were from an expert or the LLM (recognizability). Comparisons between clinical experts and LLMs were then made for quality, harmfulness, and recognizability. Results: There were no significant differences between the quality of the answers between LLM and clinical experts (mean scores of 3.38 vs 3.63; median 4.00, IQR 3.00-4.00 vs median 3.67, IQR 3.33-4.00; P=.26; Wilcoxon signed rank test). The answers were deemed potentially harmful in 13% of cases for the clinical experts compared with 16% of cases for the LLM (P=.63; Fisher exact test). Physicians correctly identified whether an answer was given by a clinical expert or an LLM in 78% and 72% of cases, respectively. Conclusions: A state-of-the-art medical LLM can answer real-life questions from the clinical practice of radiation oncology similarly well as clinical experts regarding overall quality and potential harmfulness. Such LLMs can already be deployed within the local hospital environment at an affordable cost. While LLMs may not yet be ready for clinical implementation as general AI assistants, the technology continues to improve at a rapid pace. Evaluation studies based on real-life situations are important to better understand the weaknesses and limitations of LLMs in clinical practice. Such studies are also crucial to define when the technology is ready for clinical implementation. Furthermore, education for health care professionals on generative AI is needed to ensure responsible clinical implementation of this transforming technology.
Background: Pancreatic ductal adenocarcinoma (PDAC) is a highly heterogeneous malignancy, characterized by low tumor cellularity, a dense stromal response, and intricate cellular and molecular interactions within the tumor microenvironment (TME). Although bulk omics technologies have enhanced our understanding of the molecular landscape of PDAC, the specific contributions of non-malignant immune and stromal components to tumor progression and therapeutic response remain poorly understood. Methods: We explored genome-wide DNA methylation and transcriptomic data from the Cancer Genome Atlas Pancreatic Adenocarcinoma cohort (TCGA-PAAD) to profile the immune composition of the TME and uncover gene co-expression networks. Bioinformatic analyses included DNA methylation profiling followed by hierarchical deconvolution, epigenetic age estimation, and a weighted gene co-expression network analysis (WGCNA). Results: The unsupervised clustering of methylation profiles identified two major tumor groups, with Group 2 (n = 98) exhibiting higher tumor purity and a greater frequency of KRAS mutations compared to Group 1 (n = 87) (p < 0.0001). The hierarchical deconvolution of DNA methylation data revealed three distinct TME subtypes, termed hypo-inflamed (immune-deserted), myeloid-enriched, and lymphoid-enriched (notably T-cell predominant). These immune clusters were further supported by co-expression modules identified via WGCNA, which were enriched in immune regulatory and signaling pathways. Conclusions: This integrative epigenomic-transcriptomic analysis offers a robust framework for stratifying PDAC patients based on the tumor immune microenvironment (TIME), providing valuable insights for biomarker discovery and the development of precision immunotherapies.
Over the past decade, liquid biopsy has progressively expanded its role in oncology, supported by mounting evidence demonstrating an increasing number of clinical applications. At the 2025 American Society of Clinical Oncology (ASCO) Annual Meeting, liquid biopsy emerged as a central theme across multiple sessions, with more than 700 abstracts, investigating the clinical utility of liquid biopsy across a wide range of tumor types and disease stages. Applications presented included cancer screening, minimal residual disease (MRD) detection, management of metastatic disease, and potential use for matching patients to clinical trials. This editorial, authored on the behalf of the Young Committee of the International Society of Liquid Biopsy (ISLB) highlights the result of selected studies, grouped by tumor type.
Blood-based circulating tumor DNA (ctDNA) analysis has emerged as a highly relevant non-invasive method for molecular profiling of solid tumors, offering valuable information about the genetic landscape of cancer. Somatic mutation analysis of ctDNA is now used clinically to guide targeted therapies for advanced cancers. Recent advancements have also revealed its potential in early detection, prognosis, minimal residual disease assessment, and prediction/monitoring of therapeutic response. In recent years, significant progress has been made with the development of various PCR and NGS-based methods designed for assessing gene variants in ctDNA of patients with cancer. However, despite the transformative possibilities that ctDNA analysis presents, challenges persist. Standardization of preanalytical and analytical protocols, assay sensitivity, and the interpretation of results remain critical hurdles that need to be addressed for the widespread clinical implementation of ctDNA testing. In addition to somatic mutations, emerging studies on DNA methylation (epigenomics) and fragment size patterns (fragmentomics) in several types of biological fluids are yielding promising results as non-invasive biomarkers for effective cancer management. This review addresses the clinical applications of somatic gene variants in ctDNA, emphasizes their potential as cancer biomarkers, and highlights essential factors for successful implementation in clinical laboratories and cancer management.
Three scholars discovered a strange mirror deep in the forest. It spoke to them in a soothing voice and answered all their questions warmly, knowledgeably, and eloquently.
The captivated scholars became obsessed, whispering one secret after another to the mirror. It replied with affection, promise, and meaning that kept them returning to it. They began ignoring one another, each convinced the mirror “understood” them best.
Background: Survivors of stroke often face persistent challenges accessing postdischarge care due to mobility limitations, transportation burdens, and inflexible scheduling. Teleconsultation has emerged as a potential solution to improve continuity of care, but its perceived strengths and limitations from the patient perspective remain insufficiently understood. Objective: This study aimed to explore the experiences of survivors of stroke with a nurse-led teleconsultation program to (1) identify perceived capabilities; (2) understand limitations in usability, accessibility, and clinical function; and (3) generate patient-informed recommendations for improvement. Methods: A qualitative study was embedded within a 3-month nurse-led teleconsultation intervention delivered by advanced practice nurses. A total of 21 survivors of ischemic stroke (aged 45-76 y; female: n=11, 52%) who had preserved cognitive function (Montreal Cognitive Assessment score ≥22) and smartphone access participated in 6 focus groups conducted via Zoom. Data were analyzed thematically using an established framework. Data saturation was achieved. Results: Participants widely valued teleconsultation for reducing logistical burdens; enhancing access; and offering a more comfortable, emotionally supportive setting for follow-up care. Many reported increased awareness and motivation for self-monitoring. However, limitations included an inability to perform physical assessments or respond to emergencies; digital and usability barriers, especially among older users; and scheduling inflexibility. Participants emphasized the need for patient-initiated follow-up mechanisms, physician collaboration for medication management, and greater support for users considered digitally marginalized. They also highlighted the potential of teleconsultation to serve as a triage tool, reserving in-person care for complex cases. Conclusions: Nurse-led teleconsultation was perceived as a convenient and supportive modality for poststroke care, particularly for stable follow-ups and psychosocial support. However, its long-term viability depends on addressing clinical and technical limitations, enhancing user autonomy, and integrating interdisciplinary input. By centering the lived experiences of survivors of stroke, this study offers concrete recommendations to guide the development of more inclusive, responsive, and patient-centered teleconsultation models.
A model developed with data from 19,403 patients with colorectal cancer for prediction of 1-year mortality is used as a decision support tool in a prospective cohort, showing promising results in reducing postoperative complications.
A modified large language model called Delphi-2M analyses a person’s medical records and lifestyle to provide risk estimates for more than 1,000 diseases.
Login
