1.Detection of Cervical Foraminal Stenosis from Oblique Radiograph Using Convolutional Neural Network Algorithm
Jihie KIM ; Jae Jun YANG ; Jaeha SONG ; SeongWoon JO ; YoungHoon KIM ; Jiho PARK ; Jin Bog LEE ; Gun Woo LEE ; Sehan PARK
Yonsei Medical Journal 2024;65(7):389-396
Purpose:
This study was conducted to develop a convolutional neural network (CNN) algorithm that can diagnose cervical foraminal stenosis using oblique radiographs and evaluate its accuracy.
Materials and Methods:
A total of 997 patients who underwent cervical MRI and cervical oblique radiographs within a 3-month interval were included. Oblique radiographs were labeled as “foraminal stenosis” or “no foraminal stenosis” according to whether foraminal stenosis was present in the C2–T1 levels based on MRI evaluation as ground truth. The CNN model involved data augmentation, image preprocessing, and transfer learning using DenseNet161. Visualization of the location of the CNN model was performed using gradient-weight class activation mapping (Grad-CAM).
Results:
The area under the curve (AUC) of the receiver operating characteristic curve based on DenseNet161 was 0.889 (95% confidence interval, 0.851–0.927). The F1 score, accuracy, precision, and recall were 88.5%, 84.6%, 88.1%, and 88.5%, respectively.The accuracy of the proposed CNN model was significantly higher than that of two orthopedic surgeons (64.0%, p<0.001; 58.0%, p<0.001). Grad-CAM analysis demonstrated that the CNN model most frequently focused on the foramen location for the determination of foraminal stenosis, although disc space was also frequently taken into consideration.
Conclusion
A CNN algorithm that can detect neural foraminal stenosis in cervical oblique radiographs was developed. The AUC, F1 score, and accuracy were 0.889, 88.5%, and 84.6%, respectively. With the current CNN model, cervical oblique radiography could be a more effective screening tool for neural foraminal stenosis.
2.Application of surface-guided radiation therapy in prostate cancer: comparative analysis of differences with skin marking-guided patient setup
Jaeha LEE ; Yeon Joo KIM ; Youngmoon GOH ; Eunyeong YANG ; Ha Un KIM ; Si Yeol SONG ; Young Seok KIM
Radiation Oncology Journal 2023;41(3):172-177
Purpose:
Surface-guided radiation therapy is an image-guided method using optical surface imaging that has recently been adopted for patient setup and motion monitoring during treatment. We aimed to determine whether the surface guide setup is accurate and efficient compared to the skin-marking guide in prostate cancer treatment.
Materials and Methods:
The skin-marking setup was performed, and vertical, longitudinal, and lateral couch values (labeled as "M") were recorded. Subsequently, the surface-guided setup was conducted, and couch values (labeled as "S") were recorded. After performing cone-beam computed tomography (CBCT), the final couch values was recorded (labeled as "C"), and the shift value was calculated (labeled as "Gap (M-S)," "Gap (M-C)," "Gap (S-C)") and then compared. Additionally, the setup times for the skin marking and surface guides were also compared.
Results:
One hundred and twenty-five patients were analyzed, totaling 2,735 treatment fractions. Gap (M-S) showed minimal differences in the vertical, longitudinal, and lateral averages (-0.03 cm, 0.07 cm, and 0.06 cm, respectively). Gap (M-C) and Gap (S-C) exhibited a mean difference of 0.04 cm (p = 0.03) in the vertical direction, a mean difference of 0.35 cm (p = 0.52) in the longitudinal direction, and a mean difference of 0.11 cm (p = 0.91) in the lateral direction. There was no correlation between shift values and patient characteristics. The average setup time of the skin-marking guide was 6.72 minutes, and 7.53 minutes for the surface guide.
Conclusion
There was no statistically significant difference between the surface and skin-marking guides regarding final CBCT shift values and no correlation between translational shift values and patient characteristics. We also observed minimal difference in setup time between the two methods. Therefore, the surface guide can be considered an accurate and time-efficient alternative to skin-marking guides.
3.Outcome of dose-escalated intensity-modulated radiotherapy for limited disease small cell lung cancer
Eunyeong YANG ; Young Seob SHIN ; Ji Hyeon JOO ; Wonsik CHOI ; Su Ssan KIM ; Eun Kyung CHOI ; Jaeha LEE ; Si Yeol SONG
Radiation Oncology Journal 2023;41(3):199-208
Purpose:
An optimal once-daily radiotherapy (RT) regimen is under investigation for definitive concurrent chemoradiotherapy (CCRT) in limited disease small cell lung cancer (LD-SCLC). We compared the efficacy and safety of dose escalation with intensity-modulated radiotherapy (IMRT).
Materials and Methods:
Between January 2016 and March 2021, patients treated with definitive CCRT for LD-SCLC with IMRT were retrospectively reviewed. Patients who received a total dose <50 Gy or those with a history of thoracic RT or surgery were excluded. The patients were divided into two groups (standard and dose-escalated) based on the total biologically effective dose (BED, α/β = 10) of 70 Gy. The chemotherapeutic regimen comprised four cycles of etoposide and cisplatin.
Results:
One hundred and twenty-two patients were analyzed and the median follow-up was 27.8 months (range, 4.4 to 76.9 months). The median age of the patients was 63 years (range, 35 to 78 years) and the majority had a history of smoking (86.0%). The 1- and 3-year overall survival rates of the escalated dose group were significantly higher than those of the standard group (93.5% and 50.5% vs. 76.7% and 33.3%, respectively; p = 0.008), as were the 1- and 3-year freedom from in-field failure rates (91.4% and 66.5% vs. 73.8% and 46.9%, respectively; p = 0.018). The incidence of grade 2 or higher acute and late pneumonitis was not significantly different between the two groups (p = 0.062, 0.185).
Conclusion
Dose-escalated once-daily CCRT with IMRT led to improved locoregional control and survival, with no increase in toxicity.
4.Characteristic Risk Factors Associated with Planned versus Impulsive Suicide Attempters.
Jaeha KIM ; Kang Sook LEE ; Dai Jin KIM ; Seung Chul HONG ; Kyoung Ho CHOI ; Youngmin OH ; Sheng Min WANG ; Hae Kook LEE ; Yong Sil KWEON ; Chung Tai LEE ; Kyoung Uk LEE
Clinical Psychopharmacology and Neuroscience 2015;13(3):308-315
OBJECTIVE: The present study aimed to investigate predictors for planned suicide attempters. METHODS: This study included 1,003 patients who attempted suicide and visited emergency department. They were divided into two groups, planned suicide attempters (SAs; n=133 [13.3%]) and impulsive SAs (n=870, [86.7%]), and the demographic variables, clinical characteristics, factors related to suicide, and psychiatric resources of the groups were compared. RESULTS: Major depressive disorder and substance use disorders were more common among planned SAs than among impulsive SAs. Additionally, the planned SAs were older, more likely to be divorced, separated or widowed, and more likely to have comorbid medical illnesses, severe depression, higher suicidality, and self-blaming tendencies than the impulsive SAs. Financial problems and physical illnesses were more common in planned SAs but interpersonal conflicts were more frequent in impulsive SAs. Planned SAs had fewer previous suicide attempts but these were more serious suicide attempts. The presence of the hope to die, a written will, and suicidal ideation of a repetitive, intense, and continuous nature were predictive of planned SAs. CONCLUSION: The present findings demonstrated that planned SAs had more severe psychopathology and medical illnesses than impulsive SAs. Therefore, screening for depression, substance use disorders, and suicidal plans among old and medically ill patients may be important for preventing suicide attempts.
Depression
;
Depressive Disorder, Major
;
Divorce
;
Emergency Service, Hospital
;
Hope
;
Humans
;
Impulsive Behavior
;
Mass Screening
;
Psychopathology
;
Risk Factors*
;
Substance-Related Disorders
;
Suicidal Ideation
;
Suicide*
;
Suicide, Attempted
;
Widowhood