Background: In staged excision procedures, it is difficult to estimate the number of excisions that will be required and the extent of scar lengthening. The purpose of this study was to investigate the effects of the size, shape, and resection amount of lesions on the outcomes of staged excision through an animal experiment. Methods: In total, 20 ellipses with five different designs (n=4) were evaluated on pig skin. The experiment consisted of two groups: group 1 had excisions of the same length, but with different widths, while group 2 had excisions of the same size, but with different amounts of resection. The size of the lesions and the amount of resection were analyzed in terms of the ratio of length (long axis) and width (short axis) (S/L ratio). Results: In the first group, initial ellipses measuring 5×4, 5×3, and 5×2 cm increased in size to 9.25±0.07 (185%), 8.55±0.07 (171%), and 8.10±0.14 cm (162%), respectively. In the second group, in which all ellipses measured 5×3 cm, those with a resection amount of 5×1.5, 5×2, and 5×2 cm with a fish fin grew to 8.75±0.15 (175%), 8.55±0.07 (171%), and 8.60±0.17 cm (172%), respectively. In group 1, the larger the S/L ratio, the longer the final length. In group 2, a greater resection amount was associated with a shorter final length. Conclusions We believe that the measurements of this study in terms of shape, size, and excision amount will be reasonable predictive references for staged excision procedures.

Objective: Urolithiasis is obstructive uropathy that can progress to acute pyelonephritis by retrograde urinary tract infection. This study aimed to assess the clinical features of acute pyelonephritis with urolithiasis based on the computed tomography (CT) findings. Methods: The medical records of patients who visited the emergency room were reviewed retrospectively from January 1, 2018, to January 31, 2021. This study investigated 109 patients (over 14 years old) diagnosed with acute pyelonephritis by kidney computed tomography. In the patient with acute pyelonephritis with urolithiasis, there were some findings like wedge-shaped hypodensity, kidney enlargement, perinephric fat stranding, pelvicalyceal wall thickness and enhancement, Gerota’s fascia thickness, and delayed excretion of contrast. The subjects were classified into two groups according to the number of CT findings: less than three CT findings group (group 1) and three or more CT findings group (group 2). Results: A higher incidence of hypotension, tachycardia, and sepsis was observed in group 2 than in group 1 (all P<0.05). Furthermore, the high sensitivity C-reactive protein level was also significantly higher than group 1. The number and location of urinary stones had no significant relationship with the CT findings. Conclusion The characteristics of acute pyelonephritis with urolithiasis differ for each group. A significant correlation was observed between the number of CT findings and the prevalence of sepsis. Predicting the patients’ clinical characteristics and the presence of sepsis will be helpful in treatment.

The use of a balloon guide catheter (BGC) in the endovascular management of acute ischemic stroke is known to improve the efficacy and efficiency of the procedure by reducing the risk of distal embolization. During the procedure, the balloon of the catheter causes a temporary arrest of cerebral blood flow. However, failure of the balloon to deflate during the BGC procedure can result in catastrophic complications, including aggravated hypoxic damage.
This paper aims to share the resolution and methodological analysis of our experience with BGC balloon deflation failure, which was confirmed by a reproducible experiment under similar conditions.

Objectives: Developing large language models (LLMs) in biomedicine requires access to high-quality training and alignment tuning datasets. However, publicly available Korean medical preference datasets are scarce, hindering the advancement of Korean medical LLMs. This study constructs and evaluates the efficacy of the Korean Medical Preference Dataset (KoMeP), an alignment tuning dataset constructed with an automated pipeline, minimizing the high costs of human annotation. Methods: KoMeP was generated using the DAHL score, an automated hallucination evaluation metric. Five LLMs (Dolly-v2-3B, MPT-7B, GPT-4o, Qwen-2-7B, Llama-3-8B) produced responses to 8,573 biomedical examination questions, from which 5,551 preference pairs were extracted. Each pair consisted of a “chosen” response and a “rejected” response, as determined by their DAHL scores. The dataset was evaluated when trained through two different alignment tuning methods, direct preference optimization (DPO) and odds ratio preference optimization (ORPO) respectively across five different models. The KorMedMCQA benchmark was employed to assess the effectiveness of alignment tuning. Results: Models trained with DPO consistently improved KorMedMCQA performance; notably, Llama-3.1-8B showed a 43.96% increase. In contrast, ORPO training produced inconsistent results. Additionally, English-to-Korean transfer learning proved effective, particularly for English-centric models like Gemma-2, whereas Korean-to-English transfer learning achieved limited success. Instruction tuning with KoMeP yielded mixed outcomes, which suggests challenges in dataset formatting. Conclusions KoMeP is the first publicly available Korean medical preference dataset and significantly improves alignment tuning performance in LLMs. The DPO method outperforms ORPO in alignment tuning. Future work should focus on expanding KoMeP, developing a Korean-native dataset, and refining alignment tuning methods to produce safer and more reliable Korean medical LLMs.

Objectives: Developing large language models (LLMs) in biomedicine requires access to high-quality training and alignment tuning datasets. However, publicly available Korean medical preference datasets are scarce, hindering the advancement of Korean medical LLMs. This study constructs and evaluates the efficacy of the Korean Medical Preference Dataset (KoMeP), an alignment tuning dataset constructed with an automated pipeline, minimizing the high costs of human annotation. Methods: KoMeP was generated using the DAHL score, an automated hallucination evaluation metric. Five LLMs (Dolly-v2-3B, MPT-7B, GPT-4o, Qwen-2-7B, Llama-3-8B) produced responses to 8,573 biomedical examination questions, from which 5,551 preference pairs were extracted. Each pair consisted of a “chosen” response and a “rejected” response, as determined by their DAHL scores. The dataset was evaluated when trained through two different alignment tuning methods, direct preference optimization (DPO) and odds ratio preference optimization (ORPO) respectively across five different models. The KorMedMCQA benchmark was employed to assess the effectiveness of alignment tuning. Results: Models trained with DPO consistently improved KorMedMCQA performance; notably, Llama-3.1-8B showed a 43.96% increase. In contrast, ORPO training produced inconsistent results. Additionally, English-to-Korean transfer learning proved effective, particularly for English-centric models like Gemma-2, whereas Korean-to-English transfer learning achieved limited success. Instruction tuning with KoMeP yielded mixed outcomes, which suggests challenges in dataset formatting. Conclusions KoMeP is the first publicly available Korean medical preference dataset and significantly improves alignment tuning performance in LLMs. The DPO method outperforms ORPO in alignment tuning. Future work should focus on expanding KoMeP, developing a Korean-native dataset, and refining alignment tuning methods to produce safer and more reliable Korean medical LLMs.

Objectives: Developing large language models (LLMs) in biomedicine requires access to high-quality training and alignment tuning datasets. However, publicly available Korean medical preference datasets are scarce, hindering the advancement of Korean medical LLMs. This study constructs and evaluates the efficacy of the Korean Medical Preference Dataset (KoMeP), an alignment tuning dataset constructed with an automated pipeline, minimizing the high costs of human annotation. Methods: KoMeP was generated using the DAHL score, an automated hallucination evaluation metric. Five LLMs (Dolly-v2-3B, MPT-7B, GPT-4o, Qwen-2-7B, Llama-3-8B) produced responses to 8,573 biomedical examination questions, from which 5,551 preference pairs were extracted. Each pair consisted of a “chosen” response and a “rejected” response, as determined by their DAHL scores. The dataset was evaluated when trained through two different alignment tuning methods, direct preference optimization (DPO) and odds ratio preference optimization (ORPO) respectively across five different models. The KorMedMCQA benchmark was employed to assess the effectiveness of alignment tuning. Results: Models trained with DPO consistently improved KorMedMCQA performance; notably, Llama-3.1-8B showed a 43.96% increase. In contrast, ORPO training produced inconsistent results. Additionally, English-to-Korean transfer learning proved effective, particularly for English-centric models like Gemma-2, whereas Korean-to-English transfer learning achieved limited success. Instruction tuning with KoMeP yielded mixed outcomes, which suggests challenges in dataset formatting. Conclusions KoMeP is the first publicly available Korean medical preference dataset and significantly improves alignment tuning performance in LLMs. The DPO method outperforms ORPO in alignment tuning. Future work should focus on expanding KoMeP, developing a Korean-native dataset, and refining alignment tuning methods to produce safer and more reliable Korean medical LLMs.

OBJECTIVE: We developed a biosignal telemonitoring system which is based on HL7 and MFER standard. METHODS: For the communication of waveform data (ECG, EEG etc.) we adopted MFER(Medical waveform description Format Encoding Rules) standard and HL7(Health Level 7). MFER is a standard for encoding waveform biosignal such as ECG, EEG and so on. HL7 is a standard for electronic data communication between two different information systems. RESULTS: The telemornitoring system consists of an HL7 interface gateway and a central repository. The HL7 interface gateway has three modules of an MFER analyzer, an HL7 message generator, and a file sender. The central repository is a central database combined with an HL7 archiver. CONCLUSION: Through this study we might conclude that the proposed system can be a promising model for telemonitoring system in healthcare environment.
Delivery of Health Care ; Electrocardiography ; Electroencephalography ; Information Systems

PURPOSE: Medical treatment of Peyronie's disease with tamoxifen has been initially proposed as acting upon the early phase of the disease. As recent reports show no significant benefit of tamoxifen, we review the long term results of tamoxifen treatment of Peyronie's disease. MATERIALS AND METHODS: Time to progression during tamoxifen treatment of patients showing acute disease and chronic disease was compared. The acute phase was identified by pain during erection. Progression was defined as enlargement of plaque size or appearance of calcification. RESULTS: The average treatment duration was 15.9+/-13.8 months (range: 3 to 48 months). The median time to progression was 7 months for acute patients and 20 months for chronic patients. Eighty percent of patients in the acute phase showed relief of pain; however, overall progression was 72.1% (78.0% for acute, 66.7% for chronic). Patient history, comorbidities, serum testosterone or initial plaque characteristics, and severity of curvature were not predictive of disease progression. CONCLUSIONS: Tamoxifen showed no significant benefit in slowing the progression of Peyronie's disease in the acute phase over the chronic phase. Peyronie's disease continued to progress, though at a dampened rate for patient's in the chronic phase.
Acute Disease ; Chronic Disease ; Comorbidity ; Humans ; Male ; Penile Induration ; Tamoxifen ; Testosterone

PURPOSE: To evaluate the efficacy of catheter-directed thrombolysis in treating symptomatic deep venous thrombosis (DVT) in lower limbs. METHODS: Between Jan. 1999 and Dec. 2002, 29 consecutive patients with DVT had received thrombolytic therapy. The male: female ratio was 6: 23 and the mean age was 50.3+/-13.5 years. The mean duration of symptom was 9.9+/-22.1 days. Catheter-directed infusions of urokinase were administrated via ipsilateral popliteal veins and the angioplasty and stent placement performed after the thrombolytic procedure. The mean dosage of urokinase and duration of thrombolysis were 2, 435, 000+/-887, 000 units and mean duration of thrombolysis was 36.8+/-17.9 hours. Oral medication of warfarin continued at least six months or more. To evaluate the venous patency, duplex ultrasonography or CT venography were performed. RESULTS: Lysis was complete in 17 patients (58.6%, all acute DVT), partial in 11 (37.9%), with only one patient failing. Iliac vein stenosis had shown in 16 patient after thrombdysis. Which were treated with balloon angioplasty and stent insertion. As a postprocedural complication, vaginal bleeding occurred in two patients; one was treated with transfusion but the other stopped without treatment. CONCLUSION: Catheter-directed thrombolysis with urokinase is effective for the treatment of DVT in lower limbs. However further study will be reguired to evaluate the relationship between the incidence of postthrombotic syndrome and thrombolytic therapy alone.
Angioplasty ; Angioplasty, Balloon ; Constriction, Pathologic ; Female ; Humans ; Iliac Vein ; Incidence ; Lower Extremity ; Male ; Phlebography ; Popliteal Vein ; Postthrombotic Syndrome ; Stents ; Thrombolytic Therapy ; Ultrasonography ; Urokinase-Type Plasminogen Activator* ; Uterine Hemorrhage ; Venous Thrombosis* ; Warfarin

PURPOSE: Popliteal artery entrapment syndrome (PAES) is rare but major cause of non-atheromatous popliteal arterial insufficiency in young. Because of its rareness, it is often neglected or misdiagnosed as thrombosis or embolism. Consequently surgeons would lose the appropriate time of treatment. METHOD: We reviewed 11 cases of PAES from 1994 to 2002 regarding to clinical characteristics, image findings, management and their results. RESULT: Two of 11 patients had bilateral involvement. All patients were male and aged 12 to 45 year old (mean; 32.1). Intermittent claudication was presented as initial symptom in all. One had toe gangrene. Conventional arteriography (11 cases) was used as initial diagnostic method. CT (7 cases) and MR (4 cases) angiography were also used to make diagnosis. Type II PAES were most common in 7 limbs. 11 limbs of 10 patients underwent operation. One was managed conservatively because of advanced liver cirrhosis. Resection of medial head of gastrocnemius and popliteal arterial bypass were performed in 7 limbs. One myectomy with femoroposterotibial bypass, one femoropopliteal bypass without myectomy, and myectomy with patch angioplasty were performed. Postoperative complication occurred in two limbs. One had occlusion of graft, another had occluded segment of endarterectomised popliteal artery. Primary graft patency at 6 mo, 1 yr and 3 yr were 81% 81%, 81% respectively. CONCLUSION: In young patients with claudication who have localized lesion at popliteal artery, clinicians should pay attention to rule out PAES. Accurate diagnosis can be achieved by CT or MR angiography. Early surgical correction is recommended to minimize surgical procedure and reduce complication of the disease.
Angiography ; Angioplasty ; Diagnosis ; Embolism ; Extremities ; Gangrene ; Head ; Humans ; Intermittent Claudication ; Liver Cirrhosis ; Male ; Middle Aged ; Popliteal Artery* ; Postoperative Complications ; Thrombosis ; Toes ; Transplants