エピソード

  • How AI Is Predicting Alzheimer's Years Before Symptoms

    How AI Is Predicting Alzheimer's Years Before Symptoms
    2026/07/21
    In this episode of Healthtech Talks, Lucas and Luna explore how artificial intelligence is being trained to detect early biomarkers of Alzheimer's disease — years before clinical symptoms appear. They focus on a landmark 2024 study from the University of California San Francisco, where an AI model analyzed routine brain MRI scans and electronic health records to predict Alzheimer's onset with 92 percent accuracy up to six years in advance. The hosts break down how the model works, what biomarkers it looks for, why early detection matters for clinical trials and drug development, and the challenges of integrating such tools into standard care. They also touch on the ethical questions around screening asymptomatic patients. It's a deep dive into one of the most promising applications of AI in neurology — and what it could mean for the 55 million people living with dementia worldwide. #AlzheimersPrediction #AIinNeurology #EarlyDetection #BrainMRI #HealthtechTalks #FexingoBusiness #BusinessAndTechnology #BusinessPodcast #DigitalHealth #MedicalAI #UCSF #MachineLearning #Biomarkers #CognitiveDecline #DrugTrials #NeurodegenerativeDisease #HealthData #ClinicalAI Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    14 分
  • How AI Is Tuning Cochlear Implants for Better Hearing

    How AI Is Tuning Cochlear Implants for Better Hearing
    2026/07/21
    In this episode of Healthtech Talks, Lucas and Luna dive into a surprising use of artificial intelligence: optimizing cochlear implant programming. Cochlear implants have transformed hearing for hundreds of thousands of people worldwide, but fitting them traditionally requires multiple painful visits to an audiologist—and outcomes vary widely. Now, a team at the University of Texas at Dallas has developed an AI system that can tune implant parameters based on patient feedback in real time, reducing fitting time from hours to minutes and improving speech perception scores by an average of 18 percent. The hosts discuss how the AI works, why personalization matters so much, and what this means for the broader trend of AI-assisted medical device configuration. They also touch on the economics: more efficient fittings mean lower costs for clinics and better utilization of scarce audiologist time. A fresh angle on AI in healthtech that connects engineering, neuroscience, and patient experience. #Healthtech #AI #CochlearImplants #DigitalHealth #MedicalSoftware #ArtificialIntelligence #HearingLoss #Audiology #UniversityOfTexasAtDallas #PersonalizedMedicine #MedicalDevices #SpeechPerception #NeuralNetworks #ClinicalAI #PatientExperience #BusinessAndTechnology #FexingoBusiness #HealthtechTalks Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    7 分
  • How AI Is Matching Patients to the Right Hospital Bed

    How AI Is Matching Patients to the Right Hospital Bed
    2026/07/20
    In this episode of Healthtech Talks, Lucas and Luna explore how AI-powered bed management is transforming hospital operations. They focus on a specific case: how Johns Hopkins Hospital used machine learning to predict bed demand and reduce emergency department boarding times by 37 percent. The hosts break down the algorithm's inputs, the cultural resistance from nursing staff, and the unexpected outcome that AI actually improved nurse satisfaction. They also touch on the economics: how a typical 500-bed hospital can save up to $8 million annually by reducing length of stay and avoidable transfers. No hype, just the operational reality of matching patients to beds in real time. #HealthtechTalks #AIBedManagement #HospitalOperations #JohnsHopkins #MachineLearning #PatientFlow #EmergencyDepartment #NurseSatisfaction #LengthOfStay #OperationalEfficiency #PredictiveAnalytics #DigitalHealth #Telemedicine #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #HealthIT #ClinicalDecisionSupport Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    10 分
  • How AI Is Uncovering Hidden Patterns in Electronic Health Records

    How AI Is Uncovering Hidden Patterns in Electronic Health Records
    2026/07/20
    In this episode of Healthtech Talks, Lucas and Luna explore how artificial intelligence is being used to find patterns in electronic health records that humans routinely miss. They focus on a specific case from Boston Children's Hospital, where an AI model analyzing 1.5 million patient records identified a previously unknown link between postpartum depression and a specific vitamin D deficiency marker — seven years before standard screening would have caught it. The hosts discuss the challenges of unstructured data in EHRs, the 'data exhaust' problem, and how natural language processing is turning doctors' free-text notes into structured insights. They also touch on the ethical questions around mining medical data without explicit patient consent and why health systems are starting to invest in 'data lakes' rather than traditional databases. A concrete look at how machine learning is shifting from imaging to text analysis in healthcare. #HealthtechTalks #AIinHealthcare #ElectronicHealthRecords #NaturalLanguageProcessing #BostonChildrensHospital #PostpartumDepression #VitaminD #MachineLearning #DataMining #UnstructuredData #HealthIT #ClinicalData #PredictiveAnalytics #MedicalAI #DataLakes #BusinessAndTechnology #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    9 分
  • How AI Is Preventing Medication Errors in Hospitals

    How AI Is Preventing Medication Errors in Hospitals
    2026/07/19
    Lucas and Luna explore how artificial intelligence is being deployed to catch dangerous medication errors before they reach patients. They examine a 2025 study from Johns Hopkins showing that AI-powered clinical decision support systems can intercept 84% of potentially harmful drug-drug interactions and dosage miscalculations. The episode drills into a specific case: how Vanderbilt University Medical Center's AI system reduced adverse drug events by 37% in its first year of deployment. They also discuss the challenges — alert fatigue among clinicians, interoperability with legacy electronic health records, and the tension between safety and workflow speed. A focused look at one of the highest-impact, least-glamorous applications of AI in healthcare. #AI #Healthtech #MedicationErrors #PatientSafety #ClinicalDecisionSupport #JohnsHopkins #Vanderbilt #ElectronicHealthRecords #DrugInteractions #HospitalSafety #MachineLearning #HealthcareAI #Pharmacy #AlertFatigue #Interoperability #Business #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    10 分
  • How AI Is Matching Clinical Trials to Patients in Minutes

    How AI Is Matching Clinical Trials to Patients in Minutes
    2026/07/19
    Every year, thousands of cancer patients never learn about clinical trials that could save their lives. In this episode of Healthtech Talks, Lucas and Luna dive into the quiet revolution of AI-powered clinical trial matching. They focus on a specific case: how Tempus, a Chicago-based healthtech company, built an algorithm that scans electronic health records against trial eligibility criteria in under 60 seconds — a process that used to take a human research coordinator six to eight hours per patient. Lucas breaks down the numbers: 80% of clinical trials fail to meet enrollment timelines, and 30% of cancer patients say they would have considered a trial if they had known about it. Luna raises the data-quality question — what happens when a patient's record has missing or unstructured data? — and Lucas explains how natural language processing fills the gaps. They also touch on the ethical tension: does AI match patients to trials they're eligible for, or ones they'd actually benefit from? By the end, listeners understand both the technical breakthrough and the human stakes. No hype, just the mechanics and the moral questions. #Healthtech #ArtificialIntelligence #ClinicalTrials #Tempus #CancerCare #PatientMatching #NLP #ElectronicHealthRecords #Oncology #ClinicalResearch #HealthTechTalks #Fexingo #FexingoBusiness #BusinessPodcast #BusinessAndTechnology #HealthInnovation #AIinHealthcare #PrecisionMedicine Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    9 分
  • How AI Is Predicting Post-Surgery Infections With Wearables

    How AI Is Predicting Post-Surgery Infections With Wearables
    2026/07/18
    Lucas and Luna explore how AI-powered wearable sensors are being used to predict surgical site infections days before visible symptoms appear. They focus on a real-world deployment at a major US hospital system where patients wear a small adhesive patch post-surgery that tracks heart rate, respiration, skin temperature, and gait. The algorithm, trained on thousands of recovery trajectories, issues alerts when subtle deviations signal an emerging infection. The episode dives into the numbers: how early detection cut readmission rates by 34% and saved an estimated $2.7 million in avoidable costs over the first year. They also discuss the challenges — patient compliance, data privacy, and false alarm fatigue — and what this means for the future of remote post-operative monitoring. A concrete look at how continuous data streams could transform surgical recovery. #AI #Wearables #SurgicalSiteInfection #RemoteMonitoring #PostSurgery #PredictiveAnalytics #HealthcareAI #PatientSafety #ReadmissionReduction #DigitalHealth #Telemedicine #MedicalSoftware #Healthtech #Business #Technology #FexingoBusiness #BusinessPodcast #HealthtechTalks Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    8 分
  • How AI Is Predicting Sepsis Hours Before Blood Cultures Turn Positive

    How AI Is Predicting Sepsis Hours Before Blood Cultures Turn Positive
    2026/07/18
    Sepsis is the silent killer that accounts for one in five deaths worldwide. By the time blood cultures come back positive, the patient is often already in crisis. But a new wave of AI-driven early warning systems is changing the timeline. In this episode of Healthtech Talks, Lucas and Luna explore how machine learning models trained on real-time vitals, lab results, and nursing notes can predict sepsis up to 12 hours before traditional diagnostics. They dive into the Epic Sepsis Model controversy from 2021, and look at a 2025 study from Johns Hopkins that cut sepsis mortality by 20% using a more transparent algorithm. They discuss the tension between accuracy and interpretability, and why clinicians are finally trusting these tools. If you're in healthtech or hospital operations, this episode is about the models that are already running in the background of your local ICU. #SepsisPrediction #AIinHealthcare #ClinicalDecisionSupport #EarlyWarningSystem #EpicSystems #JohnsHopkins #MachineLearning #Healthtech #DigitalHealth #ICUMonitoring #PatientSafety #MIMICIII #LactateClearance #SOFAScore #Business #Technology #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    11 分