Jae Eun Ko

Dr. Jae Eun Ko


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 February 2025 in Sensors
Verified
Journal Article
Open access
PDF file
JKJae Eun Ko
Jae Eun Ko
SKSeungHui Kim
SeungHui Kim
JSJae Ho Sul
Jae Ho Sul
SKSung Min Kim
Sung Min Kim
Background: Human activity recognition (HAR) plays a pivotal role in digital healthcare, enabling applications such as exercise monitoring and elderly care. However, traditional HAR methods relying on accelerometer data often require complex preprocessing steps, including noise reduction and manual feature extraction. Deep learning-based human activity recognition (HAR) using one-dimensional accelerometer data often suffers from noise and limited feature extraction. Transforming time-series signals into two-dimensional representations has shown potential for enhancing feature extraction and reducing noise. However, existing methods relying on single-feature inputs or extensive preprocessing face limitations in robustness and accuracy. Methods: This study proposes a multi-input, two-dimensional CNN architecture using three distinct data reconstruction methods. By fusing features from reconstructed images, the model enhances feature extraction capabilities. This method was validated on a custom HAR dataset without requiring complex preprocessing steps. Results: The proposed method outperformed models using single-reconstruction methods or raw one-dimensional data. Compared to a one-dimensional baseline, it achieved 16.64%, 13.53%, and 16.3% improvements in accuracy, precision, and recall, respectively. We tested across various levels of noise, and the proposed model consistently demonstrated greater robustness than the time-series-based approach. Fusing features from three inputs effectively captured latent patterns and variations in accelerometer data. Conclusions: This study demonstrates that HAR can be effectively improved using a multi-input CNN approach with reconstructed data. This method offers a practical and efficient solution, streamlining feature extraction and enhancing performance, making it suitable for real-world applications.

7 January 2025 in Diagnostics
Verified
Journal Article
Open access
PDF file
KLKyoung Yoon Lim
Kyoung Yoon Lim
JKJae Eun Ko
Jae Eun Ko
YHYoo Na Hwang
Yoo Na Hwang
SLSang Goo Lee
Sang Goo Lee
SKSung Min Kim
Sung Min Kim
Background: Liver cancer has a high mortality rate worldwide, and clinicians segment liver vessels in CT images before surgical procedures. However, liver vessels have a complex structure, and the segmentation process is conducted manually, so it is time-consuming and labor-intensive. Consequently, it would be extremely useful to develop a deep learning-based automatic liver vessel segmentation method. Method: As a segmentation method, UNet is widely used as a baseline, and a multi-scale block or attention module has been introduced to extract context information. In recent machine learning efforts, not only has the global context extraction been improved by introducing Transformer, but a method to reinforce the edge area has been proposed. However, the data preprocessing step still commonly uses general augmentation methods, such as flip, rotation, and mirroring, so it does not perform robustly on images of varying brightness or contrast levels. We propose a method of applying image augmentation with different HU windowing values. In addition, to minimize the false negative area, we propose TransRAUNet, which introduces a reverse attention module (RAM) that can focus edge information to the baseline TransUNet. The proposed architecture solves context loss for small vessels by applying edge module (RAM) in the upsampling phase. It can also generate semantic feature maps that allows it to learn edge, global context, and detail location by combining high-level edge and low-level context features. Results: In the 3Dricadb dataset, the proposed model achieved a DSC of 0.948 and a sensitivity of 0.944 in liver vessel segmentation. This study demonstrated that the proposed augmentation method is effective and robust by comparisons with the model without augmentation and with the general augmentation method. Additionally, an ablation study showed that RAM has improved segmentation performance compared to TransUNet. Compared to prevailing state-of-the-art methods, the proposed model showed the best performance for liver vessel segmentation. Conclusions: TransRAUnet is expected to serve as a navigation aid for liver resection surgery through accurate liver vessel and tumor segmentation.

14 February 2025 in Sensors
Verified
Journal Article
Open access
PDF file
SKSeungHui Kim
SeungHui Kim
JKJae Eun Ko
Jae Eun Ko
SBSeungbin Baek
Seungbin Baek
DKdaechang kim
daechang kim
SKSung Min Kim
Sung Min Kim
Reduced body movement and weakened musculoskeletal function as a result of aging increase the risk of falls and serious physical injuries requiring medical attention. To solve this problem, a fall prevention algorithm using an acceleration sensor has been developed, and research is being conducted to enable continuous monitoring using a Holter electrocardiograph. In this study, we implemented a multi-input model that can detect and classify movements, including falls, utilizing the baroreflex characteristics of the heart’s potential energy changes due to movement, measured with an electrocardiogram with a three-axis acceleration sensor and a Holter electrocardiograph. Patterns were identified from the various movement characteristics of acceleration sensor data using a deep learning model consisting of CNN-LSTM, and heart rate variability (HRV) data were analyzed using a wide learning model to provide additional weight values for fall classification. Finally, a multi-input model using wide and deep learning was proposed to enhance the accuracy of fall classification. The results show that the HRV increased in fall case except in two motion types, while it decreased when standing up from a chair, indicating the application of the baroreflex characteristics reflecting the heart’s potential energy. Compared to the classification model using conventional HRV and ACC, a higher accuracy was achieved in the multi-input model using ACC-HRV data, and a precision, recall, and F1 score of 0.91 was measured, indicating improved performance. This is expected to have a positive impact on fall prevention by improving the accuracy of fall classification in the elderly for 15 different movements.

31 December 2025 in Bioengineering
Verified
Journal Article
Open access
PDF file
DKDaechang Kim
Daechang Kim
JKJae Eun Ko
Jae Eun Ko
SKSung Min Kim
Sung Min Kim
This study aims to propose a non-pharmacological approach to pain relief by analyzing changes in electrocardiogram (ECG) parameters following transcutaneous microcurrent stimulation generated according to the pulse train characteristics of intensity and frequency. Therefore, we analyze and interpret stimulation methods that induce parasympathetic nervous system (PNS) activity, which is the clinical basis for pain relief. There were 14 male participants, with a height of 176.08 ± 7.05 cm, a weight of 77.07 ± 10.32 Kg, and an age of 26.35 ± 1.71 years, and 10 female participants, with a height of 160.6 ± 5.88 cm, a weight of 52.9 ± 9.03 Kg, and an age of 24 ± 1.61 years. The microcurrent stimulation patch was attached to the left wrist. In order to observe the PNS induction effect of the measured electrocardiograms, time and frequency domains were analyzed and additional nonlinear analysis was performed. Data measurements had a rest period of more than 1 h depending on the intensity, and more than 1 day depending on the frequency to ensure sufficient stabilization time. Although physiological changes were shown differently in various pulse trains, among them, after 7 Vpp microcurrent stimulation at 1 Hz, the values of the square root of the mean squared differences of successive R-R intervals and instantaneous RR interval variability, which indicate PNS activity in the subjects, significantly increased from 41.31 ± 34.13, 29.23 ± 24.14 ms to 65.09 ± 32.46, 44.56 ± 37.92 ms (p < 0.05). Activation of PNS, which can relieve pain, was confirmed only in the 7 Vpp with 1 Hz stimulation. This suggests that microcurrent stimulation can relieve pain in a non-pharmacological way by inducing activation of PNS.

28 October 2026 in Journal of Medical Internet Research
Verified
Journal Article
Open access
JKJae Eun Ko
Jae Eun Ko
JHJi Woo Hwang
Ji Woo Hwang
SKSung Min Kim
Sung Min Kim
JKJi Yean Kwon
Ji Yean Kwon
Background: Generative AI (GenAI) is being deployed rapidly in medicine, and evidence of its clinical safety and efficacy depends on transparent reporting of randomized controlled trials (RCTs). The CONSORT-AI (Consolidated Standards of Reporting Trials–AI) extension is a benchmark for AI trial reporting, but it was drafted for conventional AI, and its applicability to GenAI has not been examined. Objective: This study aimed to assess item-level adherence to the CONSORT-AI among RCTs evaluating GenAI for medical purposes and to distinguish deficiencies attributable to nonadherence from those attributable to items that do not correspond to how GenAI interventions are specified or delivered. Methods: PubMed, CENTRAL (Cochrane Central Register of Controlled Trials), and Embase were searched for peer-reviewed, English-language RCTs published from 2022 to April 3, 2026, supplemented by backward and forward citation searching, in which patients, health care professionals, or laypeople received a GenAI intervention for a medical purpose; non-GenAI, nonmedical, nonrandomized, and non-peer-reviewed reports were excluded. Two reviewers independently screened studies, extracted data, and rated CONSORT-AI items as reported, partially reported, not reported, or not applicable using signaling questions, with disagreements resolved by a third reviewer (Cohen κ=0.81 and prevalence-adjusted, bias-adjusted κ=0.84). Risk of bias was not assessed, as the target was reporting adherence. Adherence was the proportion of reports rated “reported” per item; item-level adherence was compared across 9 characteristics using Fisher exact and Freeman-Halton tests (2-sided α=.05). Results: From 16,485 retrieved records, 74 reports from 73 RCTs (24,150 participants; median 100, IQR 50.0‐247.0) were included. Large language models (LLMs) accounted for 98.6% (73/74) of interventions, with 60.8% (45/74) using GPT-family models. Item-level reporting adherence ranged from 0% (0/72) to 98.6% (73/74). High adherence occurred in general metadata and output specification items (range 59-73, 79.7%‐98.6%). Conversely, reporting was poor for input and technical execution domains, input data eligibility (3/73, 4.1%), input acquisition/selection (9/74, 12.2%), algorithm versioning (17/74, 23.0%), trial setting integration (21/74, 28.4%), human-AI interaction (25/74, 33.8%), performance error (27/74, 36.5%), and open access/code (31/74, 41.9%). Input quality handling (Item 5[iii]) was completely unreported (0/72). Only 9 of 74 (12.2%) reports explicitly cited CONSORT-AI. While citing reports showed higher compliance for interaction and data sharing, major technical gaps persisted regardless of citation. Conclusions: This review provides the first empirical audit distinguishing general nonadherence from structural mismatches where deterministic CONSORT-AI items fail to capture GenAI mechanics. Differing from traditional AI evaluations, it reveals that key reproducibility drivers—system prompts, inference settings, and access modalities—are omitted from current standards. These empirical findings highlight critical reporting gaps in medical GenAI trials and identify domains for a tailored extension to current reporting standards, although the evidence is limited by the use of nonvalidated signaling questions. Establishing GenAI-specific trial guidelines will ensure that clinical decisions, regulatory approvals, and patient safety safeguards are grounded in reproducible and trustworthy evidence.