Panel Discussion

Abstract List

19 Sep 2026 15:05 15:25
Chin LinTaiwan Speaker Deep Learning Electrocardiography as a Non-Invasive Window into Kidney Function and Electrolyte DisturbancesRecent advances in deep learning have transformed the standard 12-lead electrocardiogram (ECG) from a tool for rhythm interpretation into a scalable physiologic sensor capable of detecting systemic disease. In this lecture, we will review the development and clinical translation of AI-enabled ECG models for dyskalemia detection and renal-function estimation. Using large real-world cohorts, convolutional and attention-based neural networks have demonstrated high accuracy for identifying moderate-to-severe hyperkalemia and hypokalemia directly from ECG waveforms, frequently preceding laboratory confirmation and outperforming clinician interpretation. Beyond electrolyte detection, AI-ECG signatures were also associated with adverse outcomes, cardiovascular risk, and future chronic kidney disease progression, even among patients with apparently normal laboratory findings. We will further discuss pragmatic deployment studies showing how real-time AI-ECG alerts integrated into emergency department workflows can accelerate treatment decisions for life-threatening hyperkalemia. Finally, the talk will address how signal-based AI can complement EHR-based prediction models in critical-care nephrology, including issues of interpretability, calibration drift, implementation, and multimodal foundation-model integration.
Kianoush KashaniUnited States Speaker Practical Steps to Train (and Become) an AI-Era PhysicianAI in Critical Care Nephrology — State of the Art and the Path from Algorithm to BedsideCritical Care Nephrology: Literature ReviewChat GPT Applications in Critical Care Nephrology
Nattachai SrisawatThailand Speaker Precision Sepsis-AKI — Biomarkers, AI and Phenotyping in the Asia-PacificSepsis-associated AKI in the Asia-Pacific carries a burden that global datasets underrepresent. In the SEA-AKI prospective multicentre ICU study, AKI was very common, with stage 3 reaching 28.9%, and infectious disease among the independent risk factors for AKI development. In the InSEA-RRT registry — 2,315 critically ill patients with stage 3 AKI across 24 hospitals in Southeast Asia and India — 47% died during hospitalization, with major adverse kidney events tracked to two years. These cohorts expose the limits of treating sepsis-AKI as one disease. This lecture examines how damage and stress biomarkers, machine-learning models trained on regional rather than imported data, and sub-phenotyping by trajectory and host response can move us from a single creatinine-based label toward actionable patient groups. Emphasis will be on what is deployable in resource-variable settings, where biomarker access is uneven and registry infrastructure is often the practical starting point for precision medicine.Timing of DialysisWhen to start kidney replacement therapy (KRT) in AKI remains one of the most frequent bedside decisions in critical care, and the trial evidence has settled less than it first appears. ELAIN favoured early initiation, AKIKI and IDEAL-ICU did not, STARRT-AKI showed no survival benefit from an accelerated strategy with more dialysis dependence at 90 days, and AKIKI-2 found that delaying further offered no advantage and may cause harm. This session works through illustrative cases rather than trial summaries: the patient with rising creatinine but no urgent indication, refractory hyperkalaemia or acidosis demanding immediate treatment, the fluid-overloaded patient with worsening oxygenation, and the diuretic-responsive patient in whom watchful waiting proves correct. Each case is used to separate absolute indications from the discretionary zone where trials apply. Emphasis will be on practical decision aids — urine output trajectory, furosemide stress testing, fluid balance, and organ-support burden — and on when not starting is the better decision.Acute PD vs Acute HD: Which Is the Right Choice?For AKI requiring kidney replacement therapy outside well-resourced ICUs, the practical question is which modality can be started safely tonight. Acute peritoneal dialysis (PD) needs no vascular access, anticoagulation, water treatment, or machine, but concerns persist about clearance and ultrafiltration control. Our multicentre randomized trial assigned 157 patients with AKI to lower-dosage acute PD (18–24 L/day) or intermittent hemodialysis three times weekly. Sepsis caused 68% of AKI. Twenty-eight-day mortality was 50% versus 49% (risk difference 0.6%), meeting the prespecified noninferiority margin, with comparable dialysis-free survival and seven-day fluid balance. Complications diverged rather than favoured one modality: intradialytic hypotension was more frequent with hemodialysis, hypokalemia with PD. This lecture translates these findings into practice — where lower-dosage PD is a legitimate first choice, where it is not, how to prescribe and monitor it, and how modality availability shapes AKI preparedness in resource-limited settings.
Sejoong KimSouth Korea Speaker Korean Big-Data Experience in AKI and CRRT OutcomesSouth Korea has established a robust nationwide health data infrastructure, enabling large-scale analyses of acute kidney injury (AKI) and continuous renal replacement therapy (CRRT). Leveraging the Health Insurance Review and Assessment Service (HIRA) and National Health Insurance Service (NHIS) databases, Korean researchers have characterized AKI incidence, risk factors, and short- and long-term outcomes across diverse clinical settings. Studies utilizing CRRT data have identified predictors of mortality, renal recovery, and progression to chronic kidney disease. These big-data approaches provide critical real-world evidence, informing clinical practice and guiding future interventional strategies in critically ill patients with AKI.Precision Volume Management in CRRT: Insights from Bioimpedance and BiomarkersOptimal fluid balance is critical in critically ill patients undergoing continuous renal replacement therapy (CRRT), yet accurate volume assessment remains challenging. Bioimpedance analysis (BIA) offers a non-invasive, objective method to quantify fluid overload and guide individualized ultrafiltration targets. Complementing BIA, emerging biomarkers provide dynamic, real-time insights into volume status and end-organ perfusion. Integrating these tools into a precision medicine framework may optimize fluid removal strategies, reduce complications, and improve survival outcomes in CRRT-dependent patients. Prospective validation of this combined approach is warranted.