Sanghoon Jheon
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Sanghoon Jheon has a total of 53 co-authors. Here are the top co-authors by number of shared publications:
Byeongju Lee
Electronics and Telecommunications Research Institute (ETRI), Daejeon, Republic of Korea
1 shared publicationChang Ho Kim
Internal Medicine, School of Medicine, Kyungpook National University, Daegu, Republic of Korea
1 shared publication- Choon-Taek Lee
Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul National University College of Medicine Seoul National University Bundang Hospital 82, Gumi‐ro 173 Beon‐gil, Bundang‐gu Seongnam‐si Gyeonggi‐do 13620 Republic of Korea
1 shared publication - Dae-Sik Lee
Electronics and Telecommunications Research Institute (ETRI)
1 shared publication - Dal Rae Kim
Division of Pulmonology and Critical Care Medicine, Department of Internal Medicine and Respiratory Center, Seoul National University Bundang Hospital, Seongnam, Republic of Korea
1 shared publication Eun Jin Lee
Lung Cancer Center, Kyungpook National University Medical Center, Daegu, Republic of Korea
1 shared publication
Latest Publications
- Open Access
- 8 January 2026
- Published by Springer Nature in npj Digital Medicine
- Vol. 9 (1)
Advanced breath analysis through hierarchical deep convolutional neural network for multi-cancer screening
Breath analysis offers a promising noninvasive strategy for early cancer detection by capturing disease-specific volatile organic compound (VOC) signatures in exhaled breath. In this study, we developed a hierarchical deep convolutional neural network (HD-CNN)-based platform for dual-cancer classification using a multimodal gas sensor array. Clinical breath samples were collected from 206 participants, including 67 healthy controls (HC), 78 lung cancer (LC), and 61 gastric cancer (GC) patients. The sensor array, composed of semiconductor metal oxide (SMO), electrochemical (EC), and photoionization detector (PID) sensors, generated time-resolved signals that were converted into 2D response maps for classification. The HD-CNN model employed a two-stage structure: a coarse classifier to distinguish HC from cancer patients (CP), followed by a fine classifier to separate LC and GC. In 5-fold cross-validation, the HD-CNN achieved classification accuracies of 82.1% (HC), 84.0% (LC), and 88.1% (GC), and average AUCs of 0.89, 0.92, and 0.89, respectively. Compared to a 1D CNN, the HD-CNN demonstrated superior class separability, increased prediction confidence, and overall enhanced performance. Additionally, we evaluated multiple coarse–fine configurations and found that isolating HC in the first stage resulted in the highest overall performance. These results support the utility of hierarchical learning and multimodal sensing for robust and scalable breath-based multi-cancer screening.
- 8 January 2026
- Published by Springer Nature in npj Digital Medicine
- Vol. 9 (1)
Open AccessAdvanced breath analysis through hierarchical deep convolutional neural network for multi-cancer screening
Breath analysis offers a promising noninvasive strategy for early cancer detection by capturing disease-specific volatile organic compound (VOC) signatures in exhaled breath. In this study, we developed a hierarchical deep convolutional neural network (HD-CNN)-based platform for dual-cancer classification using a multimodal gas sensor array. Clinical breath samples were collected from 206 participants, including 67 healthy controls (HC), 78 lung cancer (LC), and 61 gastric cancer (GC) patients. The sensor array, composed of semiconductor metal oxide (SMO), electrochemical (EC), and photoionization detector (PID) sensors, generated time-resolved signals that were converted into 2D response maps for classification. The HD-CNN model employed a two-stage structure: a coarse classifier to distinguish HC from cancer patients (CP), followed by a fine classifier to separate LC and GC. In 5-fold cross-validation, the HD-CNN achieved classification accuracies of 82.1% (HC), 84.0% (LC), and 88.1% (GC), and average AUCs of 0.89, 0.92, and 0.89, respectively. Compared to a 1D CNN, the HD-CNN demonstrated superior class separability, increased prediction confidence, and overall enhanced performance. Additionally, we evaluated multiple coarse–fine configurations and found that isolating HC in the first stage resulted in the highest overall performance. These results support the utility of hierarchical learning and multimodal sensing for robust and scalable breath-based multi-cancer screening.
- Open Access
- 1 January 2025
- Published by Wiley in Canadian Journal of Infectious Diseases and Medical Microbiology
- Vol. 2025 (1)
High‐Accuracy Long‐Read Sequencing of Mycobacterium tuberculosis PSNK363 Isolated From the Democratic People’s Republic of Korea
Long‐read sequencing is a valuable technique for high‐precision genome analysis. Despite the widespread use of the Mycobacterium tuberculosis H37Rv genome sequence as a reference for genetic variation analysis, its suitability for comparing clinical strains is limited. Therefore, we constructed the first known whole genome of a clinical M. tuberculosis strain, PSNK363, isolated from the Democratic People’s Republic of Korea, using high‐quality high‐fidelity (HiFi) read sequencing and compared its genetic variations to those of H37Rv. PSNK363 was cultured to obtain genomic DNA, which was subjected to de novo whole‐genome assembly using PacBio Sequel II with long‐read HiFi sequencing. The sequences were compared to the reference genome H37Rv. HiFi long‐read sequencing of M. tuberculosis PSNK363, with an accuracy of 99.99%, revealed a single circular chromosome of 4,422,110 bp, which is 10,578 bp longer than the H37Rv chromosome. The assembly had an average G + C content of 65.6%, 4079 protein‐coding sequences, 53 tRNA genes, and 3 rRNA genes. Most genes (72.7%) were assigned as putative functions, whereas the remaining 27.3% were annotated as hypothetical. Comparison with H37Rv revealed a large inversion in the PSNK363 genome, which contains most of the deletion and insertion variants. M. tuberculosis PSNK363 had a longer genome sequence, more protein‐coding genes, and a larger inversion region than H37Rv. High‐accuracy whole‐genome sequencing of PSNK363 holds the potential for enriching virulence databases and identifying informative loci for drug resistance analysis in M. tuberculosis isolates in the Democratic People’s Republic of Korea.
- 1 January 2025
- Published by Wiley in Canadian Journal of Infectious Diseases and Medical Microbiology
- Vol. 2025 (1)
Open AccessHigh‐Accuracy Long‐Read Sequencing of Mycobacterium tuberculosis PSNK363 Isolated From the Democratic People’s Republic of Korea
Long‐read sequencing is a valuable technique for high‐precision genome analysis. Despite the widespread use of the Mycobacterium tuberculosis H37Rv genome sequence as a reference for genetic variation analysis, its suitability for comparing clinical strains is limited. Therefore, we constructed the first known whole genome of a clinical M. tuberculosis strain, PSNK363, isolated from the Democratic People’s Republic of Korea, using high‐quality high‐fidelity (HiFi) read sequencing and compared its genetic variations to those of H37Rv. PSNK363 was cultured to obtain genomic DNA, which was subjected to de novo whole‐genome assembly using PacBio Sequel II with long‐read HiFi sequencing. The sequences were compared to the reference genome H37Rv. HiFi long‐read sequencing of M. tuberculosis PSNK363, with an accuracy of 99.99%, revealed a single circular chromosome of 4,422,110 bp, which is 10,578 bp longer than the H37Rv chromosome. The assembly had an average G + C content of 65.6%, 4079 protein‐coding sequences, 53 tRNA genes, and 3 rRNA genes. Most genes (72.7%) were assigned as putative functions, whereas the remaining 27.3% were annotated as hypothetical. Comparison with H37Rv revealed a large inversion in the PSNK363 genome, which contains most of the deletion and insertion variants. M. tuberculosis PSNK363 had a longer genome sequence, more protein‐coding genes, and a larger inversion region than H37Rv. High‐accuracy whole‐genome sequencing of PSNK363 holds the potential for enriching virulence databases and identifying informative loci for drug resistance analysis in M. tuberculosis isolates in the Democratic People’s Republic of Korea.
- Open Access
- 19 February 2018
- Published by MDPI AG in Journal of Clinical Medicine
- Vol. 7 (2) , 33
Long-Term Oncologic Outcomes, Opioid Use, and Complications after Esophageal Cancer Surgery
Effective and adequate opioid use and prevention of postoperative complications are important for enhanced recovery after surgery. We examined the effects of postoperative opioid use and postoperative complications on overall survival and recurrence-free survival after esophageal cancer surgery. This retrospective cohort study analyzed the records of patients diagnosed with esophageal cancer who underwent the Ivor Lewis operation between January 2005 and December 2011. We collected data on total opioid use for 8 days postoperatively, as well as information on postoperative complications (Clavien-Dindo classification). One hundred and twenty-one patients were included in the final analysis. Total opioid use was not significantly associated with overall survival (p = 0.520) and recurrence-free survival (p = 0.818). In contrast, the hazard ratio of postoperative overall survival was significantly higher with respect to Clavien-Dindo classification 1–2 (hazard ratio: 2.009, p = 0.046), 3a–3b (hazard ratio: 5.759, p < 0.001), and 4a–5 (hazard ratio: 3.982, p = 0.020) complications compared to no complications. Additionally, the hazard ratio of the recurrence-free survival was significantly higher in class 1–2 complications (hazard ratio: 2.336, p = 0.028) compared to none. Our study demonstrates that postoperative opioid use is not associated with survival and recurrence-free survival after esophageal cancer surgery, while postoperative complications may increase the hazard ratio for survival and recurrence-free survival.
- 19 February 2018
- Published by MDPI AG in Journal of Clinical Medicine
- Vol. 7 (2) , 33
Open AccessLong-Term Oncologic Outcomes, Opioid Use, and Complications after Esophageal Cancer Surgery
Effective and adequate opioid use and prevention of postoperative complications are important for enhanced recovery after surgery. We examined the effects of postoperative opioid use and postoperative complications on overall survival and recurrence-free survival after esophageal cancer surgery. This retrospective cohort study analyzed the records of patients diagnosed with esophageal cancer who underwent the Ivor Lewis operation between January 2005 and December 2011. We collected data on total opioid use for 8 days postoperatively, as well as information on postoperative complications (Clavien-Dindo classification). One hundred and twenty-one patients were included in the final analysis. Total opioid use was not significantly associated with overall survival (p = 0.520) and recurrence-free survival (p = 0.818). In contrast, the hazard ratio of postoperative overall survival was significantly higher with respect to Clavien-Dindo classification 1–2 (hazard ratio: 2.009, p = 0.046), 3a–3b (hazard ratio: 5.759, p < 0.001), and 4a–5 (hazard ratio: 3.982, p = 0.020) complications compared to no complications. Additionally, the hazard ratio of the recurrence-free survival was significantly higher in class 1–2 complications (hazard ratio: 2.336, p = 0.028) compared to none. Our study demonstrates that postoperative opioid use is not associated with survival and recurrence-free survival after esophageal cancer surgery, while postoperative complications may increase the hazard ratio for survival and recurrence-free survival.
- 1 August 2016
- Published by AME Publishing Company in Journal of Thoracic Disease
- Vol. 8 (S8) , S637-S641
Hospital information systems: experience at the fully digitized Seoul National University Bundang Hospital
The different levels of health information technology (IT) adoption and its integration into hospital workflow can affect the maximization of the benefits of using of health IT. We aimed at sharing our experiences and the journey to the successful adoption of health IT over 13 years at a tertiary university hospital in South Korea. The integrated system of comprehensive applications for direct care, support care, and smart care has been implemented with the latest IT and a rich user information platform, achieving the fully digitized hospital. The users experience design methodology, barcode and radio-frequency identification (RFID) technologies, smartphone and mobile technologies, and data analytics were integrated into hospital workflow. Applications for user-centered electronic medical record (EMR) and clinical decision support (CDS), closed loop medication administration (CLMA), mobile EMR and dashboard system for care coordination, clinical data warehouse (CDW) system, and patient engagement solutions were designed and developed to improve quality of care, work efficiency, and patient safety. We believe that comprehensive electronic health record systems and patient-centered smart hospital applications will go a long way in ensuring seamless patient care and experience.
- 1 August 2016
- Published by AME Publishing Company in Journal of Thoracic Disease
- Vol. 8 (S8) , S637-S641
Hospital information systems: experience at the fully digitized Seoul National University Bundang Hospital
The different levels of health information technology (IT) adoption and its integration into hospital workflow can affect the maximization of the benefits of using of health IT. We aimed at sharing our experiences and the journey to the successful adoption of health IT over 13 years at a tertiary university hospital in South Korea. The integrated system of comprehensive applications for direct care, support care, and smart care has been implemented with the latest IT and a rich user information platform, achieving the fully digitized hospital. The users experience design methodology, barcode and radio-frequency identification (RFID) technologies, smartphone and mobile technologies, and data analytics were integrated into hospital workflow. Applications for user-centered electronic medical record (EMR) and clinical decision support (CDS), closed loop medication administration (CLMA), mobile EMR and dashboard system for care coordination, clinical data warehouse (CDW) system, and patient engagement solutions were designed and developed to improve quality of care, work efficiency, and patient safety. We believe that comprehensive electronic health record systems and patient-centered smart hospital applications will go a long way in ensuring seamless patient care and experience.
- Open Access
- 9 January 2015
- Published by Impact Journals, LLC in Oncotarget
- Vol. 6 (6) , 4451-4466
RACK1is a candidate gene associated with the prognosis of patients with early stage non-small cell lung cancer
RACK1 is a candidate gene associated with the prognosis of patients with early stage non-small cell lung cancer RACK1 is a candidate gene associated with the prognosis of patients with early stage non-small cell lung cancer
- 9 January 2015
- Published by Impact Journals, LLC in Oncotarget
- Vol. 6 (6) , 4451-4466
Open AccessRACK1is a candidate gene associated with the prognosis of patients with early stage non-small cell lung cancer
RACK1 is a candidate gene associated with the prognosis of patients with early stage non-small cell lung cancer RACK1 is a candidate gene associated with the prognosis of patients with early stage non-small cell lung cancer
- Open Access
- 8 March 2014
- Published by Springer Nature in Journal of Cardiothoracic Surgery
- Vol. 9 (1) , 44
Clinical application of single incision thoracoscopic surgery: early experience of 264 cases
Single incision thoracoscopic surgery (SITS) is recognized as a difficult procedure and surgeons hesitate to perform this technique. We describe our experience of SITS and determine whether SITS can be a routine approach in minimally invasive surgery.
- 8 March 2014
- Published by Springer Nature in Journal of Cardiothoracic Surgery
- Vol. 9 (1) , 44
Open AccessClinical application of single incision thoracoscopic surgery: early experience of 264 cases
Single incision thoracoscopic surgery (SITS) is recognized as a difficult procedure and surgeons hesitate to perform this technique. We describe our experience of SITS and determine whether SITS can be a routine approach in minimally invasive surgery.
- Open Access
- 1 January 2012
- Published by Springer Nature in Respiratory Research
- Vol. 13 (1) , 106
The association of down-regulated toll-like receptor 4 expression with airflow limitation and emphysema in smokers
An association between innate immunity including Toll-like receptors (TLRs) and COPD is reported recently; TLR4 deficiency in lung can cause emphysema in animals, which is not evident in humans. We analyzed the association of TLR4 expression, airflow limitation and emphysema in smokers.
- 1 January 2012
- Published by Springer Nature in Respiratory Research
- Vol. 13 (1) , 106
Open AccessThe association of down-regulated toll-like receptor 4 expression with airflow limitation and emphysema in smokers
An association between innate immunity including Toll-like receptors (TLRs) and COPD is reported recently; TLR4 deficiency in lung can cause emphysema in animals, which is not evident in humans. We analyzed the association of TLR4 expression, airflow limitation and emphysema in smokers.