Jung-Hyun Lee

All Publications

• Neurological History Both Twinned and Queried by Generative Artificial Intelligence (P1-2.006) Lee, J., Choi, E., Castro, S., McDougal, R., Lytton, W. LIPPINCOTT WILLIAMS & WILKINS. 2025

  View details for DOI 10.1212/WNL.0000000000208430

  View details for Web of Science ID 001715411000071

• Stroke Etiology Prediction Utilizing GPT-4 Vasireddy, S., Lee, J., Castro, S., McDougal, R., Lytton, W., Levine, S. LIPPINCOTT WILLIAMS & WILKINS. 2025

  View details for DOI 10.1212/WNL.0000000000211572

  View details for Web of Science ID 001715455901081

• Localizing Lesions in the Peripheral Nervous System using Generative Artificial Intelligence Kim, M., Lee, J., Castro, S., McDougal, R., Lytton, W. LIPPINCOTT WILLIAMS & WILKINS. 2025

  View details for DOI 10.1212/WNL.0000000000210788

  View details for Web of Science ID 001715411004006

• Leveraging Generative Artificial Intelligence in Diagnosis of Thrombotic Microangiopathies: Focus on Thrombotic Thrombocytopenic Purpura Choi, E., Lee, J., Mcdougal, R., Lytton, W. W. ELSEVIER. 2024: 2616-2617

  View details for DOI 10.1182/blood-2024-194769

  View details for Web of Science ID 001411831900004

• Vitamin B12 Deficiency Manifesting As Pancytopenia, Lymphadenopathy, and Fever: A Clinical Mimic of Hematologic Malignancies. Cureus Choi, E., Galicia Garcia, G., Kishore Anna, K., Albuja Altamirano, M. F., Yip, L., Oh, J., Lee, J. H. 2024; 16 (7): e64676

  Abstract

  Pancytopenia is a complex medical condition characterized by decreased levels of red blood cells (RBCs), white blood cells (WBCs), and platelets (PLTs). It can arise from impaired production, peripheral destruction, or a combination of both. The causes of pancytopenia range from reversible factors like infections and medication reactions to irreversible conditions. Vitamin B12 deficiency is a notable reversible cause that can take years to manifest in adults due to stored reserves. However, deficiencies caused by impaired absorption, especially due to the lack of intrinsic factors (IFs), can lead to rapid deterioration within two to five years. A healthy 39-year-old male with an athletic lifestyle presented with symptoms such as dizziness, nausea, vomiting, palpitations, and fainting over a few days. These symptoms were preceded by weeks of persistent body aches, headaches, weakness, daily fevers, chills, and night sweats. Vital signs were stable. The physical examination revealed conjunctival pallor and lymphadenopathy in the submandibular and superficial cervical regions. Initial blood tests showed normocytic anemia (Hgb 4.9, MCV 80), leukopenia (2.99), thrombocytopenia (142), and elevated liver enzymes (AST 199, ALT 96, and total bilirubin of 2.04). The peripheral smear showed tear-drop cells and hypochromic cells. The initial impression was hematologic malignancy, including but not limited to leukemia, lymphoma, or myelofibrosis given clinical findings such as B-symptoms like night sweats, neck lymphadenopathy, and subjective daily fever, along with pancytopenia. The patient received a bolus of normal saline and a transfusion of two units of packed RBCs. CT scans of the chest, abdomen, and pelvis showed no adenopathy or splenomegaly. Although initial clinical assessment pointed toward a potential hematologic malignancy, comprehensive testing, including SPEP, reticulocyte count/fraction, serum folate, and serum vitamin B12, revealed only severe vitamin B12 deficiency, with a level of less than 150, with the presence of IF antibodies. Treatment involved intensive in-patient vitamin B12 injections followed by a detailed outpatient regimen. The patient completed a daily dose of vitamin B12 injections for seven consecutive days, followed by weekly injections for the next four weeks. Subsequent laboratory results demonstrated an increase in WBC count to 8.39, Hgb level to 13.2, and PLT count of 249, indicating a continued positive response to the vitamin B12 replacement therapy. In summary, pancytopenia poses a diagnostic challenge that demands careful evaluation of patient data and comprehensive testing. Vitamin B12 deficiency, which encompasses pernicious anemia (PA), is among the reversible factors to consider. This aspect holds significance before opting for more invasive measures like a bone marrow biopsy. Nutritional deficiencies need to be considered first as differentials in pancytopenia, even in the absence of typical signs of vitamin B12 deficiency (like macrocytosis and hypersegmented neutrophils) and in the presence of compelling clinical indications pointing to a hematologic malignancy.

  View details for DOI 10.7759/cureus.64676

  View details for PubMedID 39149666

  View details for PubMedCentralID PMC11326803

• GPT-4 Performance for Neurologic Localization. Neurology. Clinical practice Lee, J. H., Choi, E., McDougal, R., Lytton, W. W. 2024; 14 (3): e200293

  Abstract

  In health care, large language models such as Generative Pretrained Transformers (GPTs), trained on extensive text datasets, have potential applications in reducing health care disparities across regions and populations. Previous software developed for lesion localization has been limited in scope. This study aims to evaluate the capability of GPT-4 for lesion localization based on clinical presentation.GPT-4 was prompted using history and neurologic physical examination (H&P) from published cases of acute stroke followed by questions for clinical reasoning with answering for "single or multiple lesions," "side," and "brain region" using Zero-Shot Chain-of-Thought and Text Classification prompting. GPT-4 output on 3 separate trials for each of 46 cases was compared with imaging-based localization.GPT-4 successfully processed raw text from H&P to generate accurate neuroanatomical localization and detailed clinical reasoning. Performance metrics across trial-based analysis for specificity, sensitivity, precision, and F1-score were 0.87, 0.74, 0.75, and 0.74, respectively, for side; 0.94, 0.85, 0.84, and 0.85, respectively, for brain region. Class labels within the brain region were similarly high for all regions except the cerebellum and were also similar when considering all 3 trials to examine metrics by case. Errors were due to extrinsic causes-inadequate information in the published cases, and intrinsic causes-failures of logic or inadequate knowledge base.This study reveals capabilities of GPT-4 in the localization of acute stroke lesions, showing a potential future role as a clinical tool in neurology.

  View details for DOI 10.1212/CPJ.0000000000200293

  View details for PubMedID 38596779

  View details for PubMedCentralID PMC11003355

• Neurological history both twinned and queried by generative artificial intelligence. Frontiers in medicine Lee, J. H., Choi, E., Angulo, S. L., McDougal, R. A., Lytton, W. W. 2024; 11: 1496866

  Abstract

  We propose the use of GPT-4 to facilitate initial history-taking in neurology and other medical specialties. A large language model (LLM) could be utilized as a digital twin which could enhance queryable electronic medical record (EMR) systems and provide healthcare conversational agents (HCAs) to replace waiting-room questionnaires.In this observational pilot study, we presented verbatim history of present illness (HPI) narratives from published case reports of headache, stroke, and neurodegenerative diseases. Three standard GPT-4 models were designated Models P: patient digital twin; N: neurologist to query Model P; and S: supervisor to synthesize the N-P dialogue into a derived HPI and formulate the differential diagnosis. Given the random variability of GPT-4 output, each case was presented five separate times to check consistency and reliability.The study achieved an overall HPI content retrieval accuracy of 81%, with accuracies of 84% for headache, 82% for stroke, and 77% for neurodegenerative diseases. Retrieval accuracies for individual HPI components were as follows: 93% for chief complaints, 47% for associated symptoms and review of systems, 76% for relevant symptom details, and 94% for histories of past medical, surgical, allergies, social, and family factors. The ranking of case diagnoses in the differential diagnosis list averaged in the 89th percentile.Our tripartite LLM model demonstrated accuracy in extracting essential information from published case reports. Further validation with EMR HPIs, and then with direct patient care will be needed to move toward adaptation of enhanced diagnostic digital twins that incorporate real-time data from health-monitoring devices and self-monitoring assessments.

  View details for DOI 10.3389/fmed.2024.1496866

  View details for PubMedID 39895821

  View details for PubMedCentralID PMC11782252

• Autoimmune Encephalitis With Autoimmune Diabetes: A Case of Horror Autotoxicus. Cureus Alkaissi, H., Lee, J. H., McFarlane, S. I. 2023; 15 (1): e34268

  Abstract

  Diagnosing autoimmune encephalitis relies on clinical, radiological, and serological studies. Several autoantibodies have been implicated and recognized, with dozens of potential targets identified in the past 20 years. Despite that progress, some patients with encephalitis present a diagnostic dilemma with a seronegative status. The presence of other autoimmune diseases in a patient with encephalitis should provide a clue to the autoimmune nature of a developing neurological syndrome (cognitive, psychiatric, behavioral, and catatonia). In this report, we describe the case of a young man with type 1 diabetes mellitus who was diagnosed with seronegative autoimmune encephalitis after presenting with catatonia. We describe the lengthy clinical course, the various therapeutic trials, and his clinical outcome and response to B-cell depleting agent. This study also discusses the potential pathophysiologic pathways, providing a rationale for the diagnostic workup and therapeutic options for autoimmune encephalopathy in this case presentation.

  View details for DOI 10.7759/cureus.34268

  View details for PubMedID 36855486

  View details for PubMedCentralID PMC9968443